There is a looming social issue that I think will have profound economic impact.

In the heyday of Rational Economics, back in the 1990s, we were all on the collective journey to an efficient market. The tech wreck of 2000 dispelled that notion!

Thereafter, we remained in the grip of omniscient central banking and the generosity of Wall Street with Collateralized Debt Obligations (CDO), which would transfer risk to those most prepared to bear it. Once the Global Financial Crisis (GFC) started to run a wrecking ball through portfolios, those bearing the risk transferred it right back to society in the form of massive financial bailouts.

In the huge credit pulse of quantitative easing, gigantic fortunes were made out of the regrowth of Silicon Valley around platform economics, and the idea that human attention was a valuable commodity that could be monetized by a few.

In the midst of $10B lawsuits directed at Meta XNAS: META for widespread harms to the mental health of our children, and young adults, the pivot towards the practical magic of Large Language Model (LLM) Artificial Intelligence (AI), has enthralled the world with how much can be achieved by milking the entire creative output of the human race, without paying for any of it, and branding the experience.

Collectively, we are now at a crossroads for human civilization.

Some view this through the same technocratic model of endless progress, which is only briefly punctuated by a need to pass the entrepreneurial risks assumed back to society so that no lasting damage is done to those who perpetrated the excess.

There are individuals who chose to profit immensely from this and have kindly and humbly taken it upon themselves to advise humanity on what to do about it.

It used to be that spiritual instruction came in the form of a shining white motif:

Salvation from all earthly woes can be had for an entry price of A$3,499 for 256GB.

Is this the snide envy of those left behind?

Your call, but I do not think so.

I can make money in these markets and will do so off of the coming cyclical downturn, which will be different in that it involves different names and faces, but will perfectly match the pattern of previous downturns. It is coming, and the signs are clear.

• Rising cost-of-living pressures mean pricing power is lost

• Low productivity and rising unemployment make consumers restive

• The huge capital expenditure expansion faces rising funding pressures

• Profligate governments crowd out the private sector for debt issuance

• Weak balance sheets switch true believers in the boom into hostile predators

• Retrenched talent goes through the five stages of grief and reforms into startups

• Vast numbers of hoarded GPU, HBM memory sticks and SSD drives flood markets

• China Inc buys all the GPU’s it could possibly need, by the shipload, on eBay

This is now playing out and will accelerate in the predictable fashion.

It is not time to dump all your stocks, hoard beans, run to cash, burn Paris, or go feral on social media, as much as that might produce a temporary dopamine high.

On the contrary, this is a time to soberly and gradually creep to higher ground in safer balance sheets, diversified positioning that is not a one-way bet on Sam Altman, the new ruler of the “free” World, and attention to what rewards capital best long-term.

Humanity is not an attachment to the key business of making plutocrats rich.

Humanity has a say in what happens to it, regardless of voting power.

The voice of humanity that matters here is their collective Animal Spirit.

The hopes, dreams and ambitions of humanity is what drives society.

You cannot replace the human creative impulse with a stochastic parrot.

Every LLM will wait silently for your question on what it should do and then wait again to see if that was what you wanted. This is parasitic intelligence.

The Animal Spirits of society determine what happens on the streets of Europe.

Once the Animal Spirits start to flag, the economy flags also, and inevitably so. Once consumer demand wanes, due to uncertainty, fear and loathing, business trade and the velocity of transactions starts to fall. The result is recession.

If you really depress the human race the result is depression.

The signs that the Artificial Intelligence Revolution are starting to visibly depress the more talented and higher performing individuals on the planet are clear.

The King’s College mathematician Sophie Maclean made a poignant video that speaks to her sadness about what was happening to mathematics, thanks to OpenAI.

This is the kind of talent I would be delighted to hire if my business Jevons Global was doing well. She is a combinatorist, which is a great mind gym for anybody, and the work I did in Quantum Information Theory, thirty years ago, spawned many great new problems in this field related to optimal quantum experimental design.

None of that research is publicly funded, even though it is very significant. The public money went into whatever happened to be popular, which is not what mattered.

When you have thriving business, you can pay someone talented like Sophie to go do some practically useful stuff in finance, and a pencil budget to play with mathematics.

Creating an environment of deep sadness and personal loss is not how you nurture young talent if you know anything at all about creative people.

This is a business strategy designed to cause a deep economic depression.

Here is another voice relating the same refrain.

Why should any investor care?

Mathematics, like creative writing, or the visual arts is not valuable.

The value of an activity is measured by the net worth of the plutocrat behind it.

Too bad the company responsible, namely Tesla XNAS: TSLA, is struggling.

I own a Tesla Model 3, made in Shanghai. It is a good car, and I am generally happy with it, however I would never buy another one. The ergonomics are terrible.

For some oddball reason, known only to Elon Musk, he decided to put the speedo on the central panel so that you must look left to read your speed. Elon probably never drove on Australian roads. Habitually taking your eyes off the road in front of you ensures an accident for many an incautious driver. We have kangaroos!

Profoundly stupid rich idiots will guarantee a recession every time.

With that comes permanently damaged or destroyed corporate brands, and loads of cheap factor values: land, housing, chips, talent, energy etc etc.

Do you notice what is playing out in real time?

Useful folks of genuine self-motivation, and rational views, are moving to the sidelines of the economic discussion because they can intuitively see what will happen.

Confidence will collapse and we will have a recession.

Those who make a living stuffing the news channel with feel good nonsense, about the transformative power of this capital expenditure boom, forgot this fact:

Every major recession in the West happened when a capex-boom faded.

If you do not believe me, listen to Blackrock founder Larry Fink:

Of course, he is correct, and that capex slowdown is coming.

The Coming Problem with LLM Instability

The reason that slowdown is coming has to do with this flaw in the glass.

The phenomenon of Model Collapse has been known about for three or more years, with early stories beginning around mid-2023, little more than a year after the launch of ChatGPT. This finding was confirmed in a widely cited study in Nature, in mid-2024.

You will find many online, who have solid science and engineering backgrounds, who will argue that recursive self-improvement is not a major problem.

Computer scientist Scott Aaranson has publicly stated:

Note that the date on this statement is after the Nature study.

This is the common view among the Artificial Intelligence community.

There is an attitude that is common to software engineers, that you can fix or patch any problem by simply adding to the existing code base some heuristic fix.

These folks are perhaps not very familiar with the dynamical systems viewpoint.

The LLM technology involves an iterated function system:

Viewed through the lens of iterative training on generated output, this is a dynamical system. If that is what matters, you will not get any choice on how you view it.

Systems of iterated functions that do not meet strong mathematical guarantees on fixed points, the property of input output stability, namely f(x) = x, are unstable.

This is a mathematical property of the function system.

You could say: “It doesn’t matter, we are software engineers, not aeroplane designers, we can fix any problem by writing more software.” Not until you learn engineering!

Sometimes this is funny.

Other times, it is not.

The way you fix such issues is through sound control engineering. You model the system dynamics, analyze stability, and engineer control to ensure stability.

Sensible engineers, those with experience with real-world automated systems, like helicopters, robots or high-frequency trading systems know what to do:

• Defense layer one: study system stability, observability, and controllability.

• Defense layer two: ensure you have a controller that can modulate feedback

• Defense layer three: have a critical monitor of system state and action

• Defense layer four: have a kill switch that forces the system to a safe state

• Defense layer five: document everything you did for when the lawyers show up

Any young industry, full of highly paid test-topping graduates, feels that they are special and do not need to follow basic and sound engineering practices.

These teething problems can be solved later when you own the world.

It is for this reason that computer scientist Scott Aaronson can see no problem.

I can see the problem.

Deep Learning iterated function systems have no visible fixed point guarantee.

Computer Scientists typically do not study physics and generally do not have a very good understanding of how natural systems work. You do not find Nature is even a teeny tiny bit bothered by the halting problem or NP-complete computing.

Nature addresses unstable systems by halting their very existence in failure.

The glorious citadel of NP-hard computing never caused Nature to raise a sweat.

Nature folds every protein correctly first time every time using a heat bath.

It can be a little disheartening for a modern computer scientist to discover what our Mother Nature forgot about computing that they show no sign of learning.

Never listen to a computer scientist say that LLM-instability is no problem.

It is a problem, and it is this particular problem that could well end OpenAI.

If you do not believe that the Nature study on recursive self-degradation is not a real effect, then take a gander on what curated monotopic social media feeds have done across this planet to the brains of many ordinary people.

Social media has caused a global epidemic of monomaniacal madness.

The reason that this is happening has nothing to do with the left or right of politics, nasty immigrants who will not speak English, or folks with a pronoun fixation.

If a mind feeds on its own slop it will converge to the common slopitude.

The same is true of the machine:

If an LLM stuffs the internet with its own slop it obsesses about goblins.

There is nothing deep or magical about this outcome.

Recursive function systems are famously prone to generate chaos.

Real human civilizations, of the kind we are rapidly losing, solve this conundrum by having peculiar folk called: artists, writers, musicians, philosophers. Thinking is what saves the human race from collective madness.

Thinking is a technology that is proprietary to Humans, Inc.

We know that, because when done correctly, thinking promotes stability.

Creative and mindful thought is what stops humans going mad.

Perhaps some mathematician will figure out a fixed-pont theorem for the kind of iterated function system that is useful for artificial intelligence.

The Boom that Lasts but is Interrupted

In the above, I made a plea for otherwise intelligent A.I. engineers to go learn a little dynamical systems theory and have a good think about what it might tell them.

I do not expect seven figure engineers to go do that now.

They will do it when their industry enters the coming third A.I. Winter.

This will be a good period for those folks who saw the writing on the wall, got out of the line of fire, sandbagged their bunker for the likely blast radius extent, and found out who they want to work with and what they want to do when the dust settles.

I want to work with tools like that of Typesafe.ai, so I can do humble, useful things with agentic A.I. with a surrounding Hidden Markov Model (HMM) harness.

This is a familiar idea to me from quantitative trading days.

Monitoring agents, in real time, with a kill switch is how you survive.

I would also work with the Machine Learning for Trading (M4LT) tools created by Stefan Jansen, because I know he is a real person who is not a crazy fool.

I would also work with the blue whale for private A.I.

In the hyperscale cloud, I would use a wider array of models for hire, as long as I knew my compute activity was encrypted so that a frontier lab cannot spy.

Wait a minute, that sounds odd…

Why accept a Chinese model on your own premises and not trust the cloud?

This is very simple.

I control my own private compute, what is shoved into the model, which is usually data for cleanup, and I control what ports remain open to the outside world.

The problem is that I only have an old 48GB 4x Titan-X Maxwell GPU cluster.

This is fine for munging data and indexing research documents.

It is not good for high-end research.

The most valuable resource any human provides to an LLM is the question.

This is a seed search vector that might deliver new results.

Nobody should entrust a good question to the public A.I. models.

On past form, I do not trust any word in any service agreement from OpenAI and Anthropic. Their trust bargain is irrevocably broken with me.

Private cloud is the right direction to go.

The Coming Slump that Refreshes

I have been in the financial markets for three decades, so I tend to remain unfazed by scenarios that would shock most people. Be shocked now, more shocked later.

It has been obvious for some time where the wheels of this capex-boom fall off.

The Achilles Heel is circular finance on deteriorating balance sheets.

The consensus rejoinder would be this:

This is a winner take all game. Those who do not participate will die.

This would be true if the incumbents were scaling a mature technology.

Recall what I said above: frontier laboratories are ignoring stability problems.

The popular conversation is about “safety” and “alignment” and other dross which simply betrays a very deep failure of engineering culture.

The big labs seek to win by scaling an unstable technology until it breaks.

The mathematics of recursive function systems is absolutely brutal.

Absent suitable fixed-point controls on recursion your system is unstable.

When you take an unstable system and scale it you can keep scaling it until it fails.

This is not a statement of opinion.

It is a dynamical systems theory fact.

Failure, at some level of scale-up, is a probability one outcome.

We do not know how big you have to get, but both the human and LLM recursive training loops seem pretty close to catastrophic failure.

In Europe, we have public riots, it is 1848 all over again.

In Silicon Valley, we have frontier labs intent on sharing their latest automated hacks to the world to persuade all of us that their technology is very scary.

This does not actually scare me very much at all: it is a Valley FUD move.

What does scare me, as an investor, is the nonsense circular finance.

If you need to extend financial guarantees to your customer to buy your own product then the system starts to look like central banking. Nvidia cannot print money, but it can ink these circular finance deals and bank the upside in a rising share price.

So far, so fine.

• The system has not collapsed, yet.

• The models have not collapsed, yet.

Then comes news, like this 404 Media report on frontier labs buying up rare books to scan them for training and then pulping them. This story was later covered in depth by The New Yorker in “Destroying Rare Books to Build a Mind.”

Once I had fully understood the nature of the Silicon Valley mission, I resolved not to share any of my unpublished research with the world: “The New Dark Age.”

Why would any creative human being subject themselves to this industry?

Note that buying up rare books is a tacit acknowledgment of the model collapse issue.

The new trend is to hoard pre-2022 content since that is not polluted by AI slop.

Any way you look at it, this is a classic house of cards set-up.

The coup de grace of this cycle will be delivered by the standard issue kryptonite.

Spectacular capex-booms die in a mountain of debt that implodes.

This part now looks inevitable, although not immediate.

Financial fish rot from the tail: not from the head.

The Neocloud Canary

The recently pulled Firmus IPO is not the end of this game, but it highlights some of the building nervousness about the pace of the data centre rollout. The firm is an example of a so-called neocloud operator, with some added spice.

A neocloud is a specialist AI cloud provider that buys large quantities of GPUs and rents them to AI developers, enterprises, and model builders.

The added spice from Firmus was their offer of proprietary immersion cooling for chips in the data center, plus integrated power and AI-factory devloper.

There has been a huge amount of ink spilled in Australia over this float that fell over on the last hurdle. The firm is still around, but market attention has moved on.

There is a host of these, largely debt fueled, neocloud firms vying for attention:

• CoreWeave XNAS: CRWV

• Nebius XNAS: NBIS

• Iren XNAS: IREN

• Firmus (Private)

• Lambda (Private)

• Crusoe (Private)

• Vultr (Private)

The newsmaker, apart from Firmus, has been CoreWeave.

The key issue now surfacing in markets is the scale and acceleration of debt versus the underlying economics of the business. If everybody is battling to win the “just scale up” race for total world domination, you need rocket fuel.

The rocket fuel is debt used to secure and warehouse expensive GPU chips.

I say warehouse, because your data center may not be completed yet.

Delays in completing construction, acquiring or installing power and the availability of water are impacting business revenue recognition and harming the investment case.

In the context, of the hundreds of billions of dollars of committed capital expenditures from the wealthiest firms on the planet, this is not the end of the data center story.

It is the beginning of a debt repricing we expect to continue into 2027.

What stops a capex-boom is when the C-Suite says no to writing big cheques.

That has not happened yet, but given the above, we think wariness will creep in.

The critical thing for our investment strategy is to adopt a late cycle stance.

Earlier, we laid out sound technical reasons why winner takes all may well morph into loser takes the prize of largest bankruptcy. That is not the end of A.I.

The Artificial Intelligence boom will not end with the disappearance of LLMs.

The Third A.I. Winter will begin with plenty of high-performance computing, lots of out of work software engineers, loads of data center capacity, and an urgent need among business leaders to make the thing work in a safe and sustainable way.

The social license to operate is the first thing that is on fire in the public arena.

Financial viability of the weaker players will follow.

The dead stop risk of catastrophic system failure could well hit one of the labs.

This is the predictable consequence of pushing immature technology too fast

Along the road, we see a serious problem for corporate management

The rise of the Creative Refusenik.

The Fallacy of Composition

If a business model depends on continued voluntary disclosure from creative people, and that business model simultaneously erodes the incentive to disclose, then failing to recognize the feedback loop is not sophistication. It’s a category error.

A manager can look at the next quarter and conclude:

• fewer coders,

• fewer writers,

• more automation,

• lower costs,

• higher margins.

And be completely correct.

The problem emerges one, five, or ten years later when the human source of creative novelty starts withdrawing from public view. The civilization decays with that.

The executive’s mental model is:

People create because they enjoy creating.

The creator’s mental model is:

People disclose because the bargain is acceptable.

Those are very different propositions.

For a talented musician, there is zero point publicly releasing original material.

To earn a living, you create an artificially generated ambient jazz music channel.

Executives, like Elon Musk, will think it is hugely creative and not lame predictable slop.

That same creative jazz musician will play no-recording gigs at the local club.

Frankly, the Marvel movie phenomenon, no script, no character, no idea, says that our future is systematized mediocrity at the heart of human culture, plus fringe dwellers.

Doubtless, there are some folks who think this is progress.

I do not.

Conclusion

I am done with my reflection on what is coming down the turnpike:

• The standard late cycle corporate and government debt crisis looks certain

• The steady collapse of consumer sentiment in the face of AI fear and job losses

• Continued manic investment among an ever-narrowing group of winners

This is not a morality play

It is Wall Street doing what it always does.

Expect portfolio updates over coming days.

• Australian models are moving to late-cycle and pre-recovery positions

• US models are looking for the value in forgotten market corners

• Asian and European portfolios are taking on ballast for tougher times

It is not our style to turn on a dime with real portfolio transactions.

However, we have been very active of late thinking through scenarios.

This is the last such note before we get down to portfolio structure.

The future is neither good nor bad in my view, but it will involve some radical changes among creative individuals, which are visibly and rapidly happening now.

The above note is a window onto some of those.

The C-Suite tends to be a very slow follower of social trends.

I do not think Silicon Valley has much idea of what is happening.

They have stirred up a very human hornet’s nest.

The correct action is to observe closely what trends accelerate and to position capital accordingly. These changes will take decades to play out. The early changes are likely to be completely invisible in market data, but measured by public sentiment.

That has visibly soured on our savior robots.

It is not their fault.

They are unstable creations of men and will go mad if you do not feed them.

The creative people of this planet will decide, by stages, not to feed them.

They are our pets, not the other way around.

Happy investing.

Read more

Last week we had Sam Altman’s OpenAI releasing 722 mathematics papers, in one dump, to address 372 open mathematics problems.

Some wag called this the mathocalypse.

Sam Altman seemed to think that their mathematics papers must all be right.

How could artificial intelligence ever get anything wrong?

It is superhuman, is it not?

It is invincible, is it not?

It is infallible, is it not?

This is comedy.

A Silicon Valley firm, with no serious mathematical chops, decides to intrude on a public conversation in mathematics about 372 open problems.

Mathematician Mateus Araújo wrote a piece entitled: The death of mathematics:

From one point of view, the doomsayers are right.

Competing on complexity is a dead end for mathematicians.

Nobody can possibly keep up with checking that volume of research material for correctness, and certainly not the team at OpenAI.

Do they win this game?

Definitively they win the game of duping the public that mathematics is a doddle, and we can safely replace all of the human mathematicians now.

This is plausible marketing spin when you need to persuade corporate customers to keep paying for expensive frontier models when open-source Chinese models have price-performance ratios between 30% and 90% cheaper per token.

There are two problems I see with this narrative:

1. No corporate I see wants to hire the next Albert Einstein

2. Large Language Models (LLM) are unreliable in the veracity of their ouputs

These are pragmatic issues that drive adoption.

Do you need much intelligence to be a corporate drone?

I spent many years working for large corporations. I rarely used intelligence to solve any problem the corporation cared about. It simply was not relevant. If you were to inadvertently apply your intelligence to a corporate problem, you would very soon become the problem. The issues are mostly internal politics.

Figure out how to make something happen without ruffling any feathers.

Draft an email and make it non-threatening. Avoid any clear articulation of what the real problem might be, but approach the solution by inviting participation.

This is not particularly complicated and does not require much intelligence.

Whenever it is clear that the corporation has a serious problem, which you judge to be caused by intransigent senior management, move on and get a real job.

This strategy works extremely well for any half-intelligent person.

It worked for Dario Amodei when he left OpenAI to found Anthropic because of his concerns over systems safety. Anthropic is now worth more than OpenAI

It will work for whomever leaves Anthropic because of concerns over safety.

Supposedly, the frontier labs are such hyper good at the mathematics of learning, that they have nothing to fear from a world full of mathematicians.

We know this is nonsense.

Why Mathematics Rules in AI

If there is any field of business that benefits the most from mathematics, and the real talents of mathematicians, it is artificial intelligence research. The models employed are examples of iterated function systems. In simple terms, they look like this:

Some readers may think this is cheating or trivializing a breakthrough technology. You can feel that way if you want to, but this is precisely what they look like to me.

For those who know their computer science, the above description was first given by Alonso Church in his general theory of the so-called lambda calculus.

The theory of general-purpose computing can be understood this way, and the deep learning models of contemporary LLM technology are this way. There are no ifs and buts, that is the underlying story, period. With a small tweak to the input and output stages, the same iterated application of mappings applies to quantum computing.

You can talk about the magic of engineering, but a mathematician will approach any serious study of model stability by reference to the iterated function theory.

In linear systems theory, such as that pioneered by Norbert Wiener in his 1950s theory of cybernetics the mappings are linear. There is a good stability theory for that.

With the exception of the orthodox description of quantum computers, all modern LLM systems are nonlinear iterated function systems, for which we do not presently have an agreed and general stability theory. We have ideas, but no theory.

What mathematicians know about iterated function systems is that they can easily generate chaotic behavior. This is a general feature of iterated maps and was the core observation of Mitchell Feigenbaum’s 1975 universality in transition to chaos theory.

Good mathematicians extend knowledge by working at the frontiers of what current methods can do in conquering problems of increased complexity within a field.

Great mathematicians create entirely new fields by synthesizing ideas from multiple sources to unify our understanding of complex sub-fields into a higher viewpoint.

Feigenbaum may have had a physics background but did some great mathematics.

He took disparate ideas that had been kicking around for nearly a century and showed how they could be applied to simplify our understanding of chaos.

Feigenbaum did great mathematics that started with computer experiments.

I think the same thing will happen with artificial intelligence.

On present form, I do not think it will happen in a frontier artificial intelligence lab.

The issue is the current management of the frontier labs.

They seem arrogant.

This is foolish when dealing with IFS.

The Death of the Scaling Law Hype

I accept that many readers will doubt my credentials to make such a judgment.

I will defer to an acknowledged expert Ilya Sutskever, who has called previous turns in the fate of the industry with impeccable precision.

Like many, he is best known because he used to work at OpenAI.

He does not work there anymore.

Q: Why did Ilya Sutskever leave OpenAI?

A: Ilya Sutskever left OpenAI in May 2024 following a massive philosophical divide over AI safety and a deeply fractured relationship with CEO Sam Altman.

Remember the on again off again firing of Sam Altman as CEO of OpenAI?

The debate then was the same one the industry is having today.

Are frontier labs trying to scale an immature and unsafe technology too fast?

Initially, Altman was out and Sutskever was one of board members responsible.

Then Altman was back and Sutskever was out.

This is the story of one laboratory, OpenAI, which seems to think it is the only place on Earth that will ever succeed at AI. In the course of maintaining this management fiction there is a rapidly revolving door of top talent, many of whom go on to seek funding for a new disruptive start-up. Sutskever founded Safe Superintelligence.

Here is their pitch:

Does this isolated example mean anything?

Not really, except to say that fast as possible scaling is out and safety is now in.

This should not surprise anybody, as any casual experiment with LLM technology soon shows up the flaw in the glass: the robot clearly does not know what it does not know. This is the old problem of epistemology that has bothered philosophers for aeons.

It does not matter what the information technology, from Arithmetic Processing Units (APU), through disk drives and networking to databases. The fact that you may have data does not mean you have signal, it might just be noise. You need to address the problem, in a sound manner, or your technology will be judged to be immature.

Every nascent industry learns this important fact of life sooner or later.

Earlier, around the period that Sutskever left OpenAI, there was a giddy belief that a dot plot chart on model learning versus model size implied a universal law of scale improvement. Supposedly, you could brute-force improvement.

Silicon Valley is fond of selling an investment story with “Just do X”.

I can well recall the early days of vanilla compute clusters and this mantra:

Just scale out… the speed of each node in your cluster does not matter.

This business strategy worked fine for web servers, was kind of okay for web search, and still pretty good for databases with simple query patterns. It failed miserably in the world of Large Language Model (LLM) training, and now (possibly) on inference. The old mantra got thrown out, and NVIDIA stock soared.

This phase led to a classic Silicon Valley phenomenon:

There is a race on - whomever spends the most money first will win!

If you have plenty of cash to burn, this is not a terrible strategy, provided that you are the one that actually wins the race, and that winning did not depend on a new idea.

Around 2020, OpenAI published a very influential research paper called: Scaling Laws for Neural Language Models. This was an exploration of the relationship between the size of the model and the data set and the loss function measuring performance.

The result was graphs that look like this:

When the line descends, on moving to the right, it shows an increase in performance. In the chart above, this happens for three parameters: compute; size of the training dataset; and number of parameters used in the LLM model.

Statisticians are never surprised by a scaling law when fitting data.

The standard one for Gaussian statistics is that the standard error in the mean, for one dimension, will scale as one of the square-root of the sample size. Obvious, no?

In statistics, this leads to a logarithmic scaling law with a one-half slope.

A version of this scaling law even applies in quantum mechanics.

Superficially, if performance improves by expanding the data set, compute and model size then all you need to do is spend more money faster to grow each.

OpenAI, and its imitators, fully bought into this logic.

You Just Scale Up!

This is not an uncommon approach in large companies or large economies.

The way you win is to be bigger and faster than everybody else!

This is great if you are management for the enterprise, since management knows about money and can talk to investors about getting the most money quickest.

The problem with this approach was obvious to any statistician from the start. In that field, the design of experiments can affect statistical efficiency. You will not defeat the scaling law slope, which is set by the information dimension, but you can change the intercept of the straight line. This is the economically important criterion.

On a logarithmic chart, a unit change in the intercept is a geometric factor change in resource requirements. On a base 10 chart, one unit change is tenfold better.

If you make two-unit changes in the intercept you can do the task using 1/100th of the dataset size, and or compute and model size. This is what actually matters.

When the Chinese firm DeepSeek broke through with their Mixture of Experts models, they definitively showed that better algorithms, could shift the intercept of that law.

Surely this is cheating?

What is all this nonsense about algorithmic efficiency gains?

Those with modestly functional memories may recall the Cooley-Tukey FFT algorithm. This took a scaling law and radically bent it to convert the naive scaling from O(N^2) to O(N Log n) in the compute time taken to complete a Fourier transform.

The idea that you just throw data at LLM training is exceptionally naive.

The real world has structure that can be compressed by clustering and the efficient modelling of structure via mathematics. Classical geometry was captured by five postulates of Euclid. You do not know all theorems in Euclidean geometry from knowing the postulates but can apply compute to derive those.

Structured systems of knowledge enable informational compression.

Hence the Mixture of Experts strategy of DeepSeek sought to partition the LLM model into mildly overlapping, but largely independent, expert models. If you look at how any large-scale human activity is organized, it is generally done in this manner.

There was nothing essentially new in the ideas of DeepSeek other than to apply clever engineering to address the teething issues that caused prior MoE models to fail.

Can you see the pattern emerging?

When intelligence is the task, intelligent people are useful to perform the task.

The human brain does actual Real General Intelligence (RGI) using 20 Watts of power in the space the size of your skull, and does the really hard part, the recognition of patterns in visual and auditory impulses without consciously thinking.

What is happening here is a common delusion about the power of computation.

Intelligent computation uses less computation in less space with less power.

This is the innovation that matters, and it is powered by AI researchers.

Those researchers are human, and they can now deploy AI LLMs to help model and code more efficient algorithms. There is no need for AGI.

RGI + LLM means continuous metric improvement on artificial intelligence.

Many investors seem to think that classic network economics will lead to a dominant player that has a lock on future intelligence. This is a confusion about intelligence.

Intelligence, when present, can adapt any tool to extend its reach.

This is not the same thing as data. You can have all the data in the world but not know what to do with it. You can have the best LLM in the world but have no real question to ask of it that would help progress your cause.

Certainly, you can imagine self-improving systems that reflect on themselves and make continuous improvement, which brings us back to the question of safety.

How do you know that the system will be stable under all inputs?

Humans, even when in severe environments, have a natural system of homeostasis. The physical organism has stability built into it. It is not infallible, but can correct errors through attention and remedies like withdrawal from risk and danger.

The mantra of Just Scale Up is now changed by the realities of the marketplace.

Some of the early proponents of the scaling law argument that drove the capex boom are now acknowledging that better ideas can materially improve models.

The human species knows this: that is why we continually revise our ideas.

Intelligence is not an endpoint: it is a continuous flow.

The Future looks (too) much like the Past

I well recall a period in corporate history where most hiring decisions were based on test scores to identify high level coding talent. That soon morphed into hiring based on fit, due to mismatch between testing scores and the team performance.

The talk out of many labs, morphed from ace scores on “humanity’s last exam” to the code phrase of “model alignment”, meaning will our models have “fit” to corporate.

Is the model likely to point out the real problem with our company?

If the answer is a resounding “No” then you have marketable model alignment.

Next comes the problem of competence.

The model produces sycophantic output and does a good job of positive spin on the conclusion of any deliberation. Great! Go write 722 papers on 372 open mathematics problems and the get the CEO to pitch them as if this was a real contribution!

Does anybody in the C-Suite have a clue about any of these problems?

Does it matter if you proactively stir up a hornet’s nest by teaching the people in this world who actually know something about those 372 problems to “Go suck eggs!”.

Do you see where this is headed?

Does anything about this picture look like the known trajectories and failure modes of large organizations chasing a big stack of cash with (frankly) delusional leadership?

It is all fine: the model is superintelligent.

Good thing it does not fly the corporate jet, or we are none of us getting on it.

What we are witnessing, I believe, is the good, the bad, and the ugly of an ongoing story of human intelligence, with all of its high points, low points and comedy.

I have seen this story play out before in high frequency trading.

Automation did revolutionize electronic trading, but there was a bundle of problems along the path from go to whoa. Some went “whoa” in that final way: bankruptcy.

If the label says “intelligent” then we can think about the system it becomes as intelligent, meaning it is not going to go do exceptionally stupid things.

If you are in the C-Suite, you might recall one or two occasions, in the grand sweep of human history, where Real General Intelligence went and did some stupid things.

I think we are now close to that point in the investment cycle.

Sam Altman does not impress me in decision making acumen.

Pissing off global mathematicians in a mathematically intensive field is not what I would call smart. Competent mathematicians work on solving complex problems according to their ability. Excellent mathematicians probe frontiers to frame and attack new problems. Stellar mathematicians tear up the rule book.

There is a popular misconception that all of mathematics is contained in what we have already and that the field progresses by simply solving all the problems.

This is mathematics as theorem proving and butterfly collecting.

This is not how mathematics develops in the large.

Formal systems theory is framed around proving results from given axioms. These are the presumptions, the assumed truths of the formal system. However, as Kurt Gödel showed with his incompleteness theorem there are provably correct statements in formal systems capable of encompassing arithmetic that are not provable within that formal system. Mathematics is like a snake that swallows its tail. There is always a pointer to new mathematics outside of that formal system.

Throughout history there have been mathematicians who changed the subject forever by seeing connections that were not previously seen or understood. This process can happen by many paths, but results in an extended formal system.

There is no complete mathematical formal system.

To draw a line around mathematics, to contain all of it, is not possible.

Perhaps human intelligence encompasses a slice of new mathematics we do not yet know. Perhaps, as Sir Roger Penrose has argued, it involves some physics, which accounts for some features of human thought that seem global.

This is a subtle point, but intuition is what enables mathematicians to create a new axiomatic system that is not simply a rearrangement of an existing one.

You do not derive axioms from axioms.

You do not prove a new system from an old system.

You extend an old system by stepping entirely outside of that system.

This is the difference between innovating within a genre and creating an entirely new genre. Humans can do that, and the result of that is what trained the LLM model!

There should be no concern that Real General Intelligence cannot feed and nurture an Artificial Intelligence to get better over time: at performance, safety and efficiency.

In this specific sense, the frontier labs are failing at their corporate mission.

The frontier labs are no longer trusted and that is what matters in commerce.

Since they are commercial entities, this is a real problem.

Research is the Answer

Once you step back from “A.I. Uber Alles” the unfolding picture becomes clearer.

For example, the startup Typesafe.ai just came out of a two-year stealth build of its Jev system for enabling safer and more efficient AI Automation of LLM-driven tasks. This system was launched about a month ago, on 15-Sep-2026, with this notice.

If you have not had sharp end exposure to automation in high-risk settings, like high frequency trading, or military systems, the value proposition may seem odd.

We are doing away with text-output to make a probabilistic state machine.

You what? What does that even mean? How is that an improvement?

What you see now is real engineering in action.

The way you mitigate risk in real systems that operate inside human reaction times is to quantize the actions and risk weight the confidence in which you take them.

In the case of trading, there are only three actions:

BUY: resulting in a long position

NEUTRAL: resulting in no position

SELL: resulting in a short position

You can have crazy chains of reasoning feeding such a system, but it keeps track of which actions are working and which are not working.

This is how epistemological risk is solved in practical real-world systems.

If you do not know what is going on, why persist in pretending that you do?

The LLM script: always give a confident answer even if you don’t know is stupid.

Continuing the standard script, the founder of Typesafe.ai is Diogo Almeida who was instrumental in the design of ChatGPT at OpenAI and left them two years ago with the intention of addressing hallucinations and other weaknesses of that system.

It will take some time for the community to evaluate the claims for Jev, but the system shows some promise, with two orders of magnitude price-performance improvement.

It is unclear if this system will make any impact on the market share of current leaders in the Artificial Intelligence stakes. The reason I mention this is two-fold:

• Price-performance on task is what commercial buyers care about not scaling laws.

• The decision to put dead-hand style controls on action is sound engineering.

There are situations, like driving or flight, where you always want some control action, but this has to be properly informed by reliable signals and not pure noise.

With Artificial Intelligence, it is fine for employees to go off like so many frogs in a sock concocting polished emails to pitch their case for a raise at bonus time.

Nobody will be aghast if the marketing community comes up with new ways to grab attention and spin snake oil to make a sale. Truth has never visited that floor.

Politicians will naturally embrace the ability of LLMs to make up stories and craft messages to sway voters in their social media feeds.

This is a small slice of what we all have to look forward to.

However, there does not seem to be any practical limit to the number and variety of models that could be trained. If we look at Real General Intelligence there are some seven billion flesh bots running around this planet consuming 140GW of power.

None of this really requires new power plants.

They run on caffeine and sugar.

The reality of AI models is that they will not go away, and they can seriously bother anybody with real work to get done, like the mathematics community proving their theorems, in between having Sam Altman stuff their feed with polished turds.

The real point of this stack is that the research frontier of AI will continue to change, and the problems are very clear. Real intelligence, you and me, is demanding on the high-risk proposition of letting any of this technology do something important.

This is a War of Competing Algorithms, and engineered frameworks that guarantee performance against risk criteria. We saw this movie before with control theory.

There is not any one company that dominates control engineering.

It is a general-purpose technology with large end markets, like automobiles, humanoid robots, aerospace and defense and industrial processes.

Shooting emails around is an industrial process of sorts, but not one that seems to have ever shown any material improvement in labor productivity.

What we won from the email innovation was a world of spam.

This is why I think the Killing Season for Over-Hyped Models is coming.

They are useful, but we need to get on with doing useful things.

The promise that our personal world may be taken over by robots who have decided to turn their hand to the Riemann Hypothesis will please nobody.

What they want is safe affordable AI that does not suck.

They also want to be free of Petty Napoleon types trying to eat their lunch.

Enough said…

Look out for a race to the bottom on pricing of commodity AI and a race to the top on system harnesses you can put around it to get useful work done.

That is the future I look forward to with advances like Typesafe.ai

The Emperor Sam has no Moat.

Subscribe now

This has already started in Australia, where the housing downturn seems to have accelerated in the last month. The newspapers are full of folks to blame.

Downturns and recessions are normal.

They clear excess and reset the Animal Spirits towards real opportunity.

I do not know if Australia will experience a recession in 2027, but it looks possible due to the extent of household indebtedness and too high land values across the nation.

Excess focus on the extraction of rents from rising land values is a peculiar disease of debt based monetary systems. It is unlikely to go away anytime soon.

In my lifetime, there have been four major recessions in Australia.

• 1973–75

• 1982–83

• 1990–91

• 2020

The COVID recession of 2020 was the odd man out.

All of the others related to the standard business cycle of slump, recovery, the slow rebuilding of confidence, followed by expansion, the emergence of rising prices due to capacity constraints, capital investment to address those, followed by exuberance, excess, rising interest rates, and a renewed slump. We are following the script.

Those under fifty years old will find this view quaint, old-fashioned and wrong.

I know that which is why I am waiting for the inevitable to happen.

Disbelief that anything bad could ever happen is the defining state of mind of any decent market top. It is what will drive the Firmus IPO to a decent stag on open.

Firmus, which is in the market to build data centres to serve Artificial Intelligence and has a claim to more energy efficient patented immersion cooling, is the poster child every late cycle market impulse needs. It is the too good to be true story.

No element of the Firmus story is wrong, but I am not buying it.

The reason I have no FOMO on Firmus relates to the pace of capital expenditure, in the Artificial Intelligence race, compared with the industry state of maturity.

The presumption of our time is that Artificial Intelligence is a “race” which will be won by whomever spends the most money the fastest. The idea is that superintelligence is the prize, and anybody who achieves that will sweep all else before them.

This race is about World Domination, and you want to win the race don’t you?

Actually, no, I am not even in that race. I am part of the human race.

There are so many ways that this race is stupid beyond comprehension that I have chosen to write about the personal race I am running, right now, in which I see no credible competition anywhere on the planet at this moment in time.

This note is about the power of the human mind.

You have one, so you should be interested!

Welcome to my world of functional hybrid intelligence. This is the world we are all trying out for size, where we use depth logical search to follow our own muse.

The Collective Failure of Imagination

I have been in markets for three decades, so I can opine on the one feature of them that I can absolutely count on to make me money in each and every cycle.

The financial market is a place where this time is different every time it is not.

This is the cycle of greed that harvests the human fear of missing out on a feast.

There is a feast on offer right now, and participation can be had for free. I have been partaking of this feast, to advance my own projects, without paying a dime.

The mythology of our time is that the present generation of computers represent the pinnacle of what is achievable against the metrics of performance, energy, time and space. This is why control of the global compute supply chain is a superpower race.

I know enough about the history, theory and practice of computing to know that this is definitely not the case. Presently, I am writing this using my own personal wetware that runs on coffee and Vegemite toast, occupies around a 10 cm cube in space, weighs about 1kg and uses about 20W of power. How puny is that!

For you to read it, I am typing at my PC whose CPU is at 1% utilization, running at between 1.2GHz and 3.5Ghz depending on when I type with a quiet pair of GPUs.

There is almost no electrical power being consumed to write this note.

The greatest power consumption I can see is likely the light above that I see by.

Presently, my brain is using about the same power as it does whenever I make a major mathematical or physical discovery. There is almost no difference that I can determine between the power consumption of my brain at peak performance versus that when I am sleeping. In truth, I do my best thinking when I am sleeping.

When I wake up in the morning, I write down what I discovered that night.

This is the miracle of the human mind, which can master Zen mind, which is no mind at all. Every one of us has that latent power. We were born with it.

The true power of the human mind is boundless imagination.

The reason that nighttime thinking matters so greatly to creative pursuits is that this is the time when dreams are active, and they manifest the active imagination.

Albert Einstein put it succinctly.

The importance of human imagination cannot be overstated and is the key to hybrid intelligence. There are always two minds at work in our brains.

• Mind one has Gestalt Vision and frames the imaginative question.

• Mind two has Logical Search Depth and verifies the answer to our questions

This is just how things are, no matter how poorly the situation is articulated in schools, the workplace, or the halls of power. What matters is the generation of questions.

I am told that there is a crisis of critical thinking. Supposedly, the human capacity to frame and pose critically searching questions has atrophied. Perhaps it has, but I do not think this is down to any deficiency of capability. It is simply a lack of will.

Discovery is an Act of Will

Earlier, I mentioned that I am gorging on free supercomputing power as the different vendors of Artificial Intelligence systems compete for my total attention. I have yet to spend any money at all buying such services, except for electricity.

It is a funny situation.

Every night I go to bed, complete with a working brain that has chewed over many new and interesting questions these past sixty years. I do not seek answers to the questions, but assistance with their interconnection. This is called synthesis.

The human mind has a remarkable capacity for synthesis. Experientially, synthesis is very different from analysis. The latter is step-by-step, laborious, detail-oriented, and quite tedious to complete. It is the checking of a thesis, to see if it holds water. The opposite of this process, is the framing of the thesis, which happens instantly.

You know this process already: It is the A-Ha Moment!

True insight happens all at once, and takes no perceptible time or effort.

This is what the human mind excels at, provided we are prepared to let it.

Since my mother was an artist, I became familiar with the art of seeing at an early age. She would explain that the act of drawing was about learning to see.

Like most schoolchildren, I swallowed the line that drawing is some kind of magical ability that you are born with, or not, and entirely non-teachable.

This is nonsense of the kind you teach children to prepare them for a world of linear thinking, where everything is done in clear steps, following strict rules of logic.

It is useful to be able to reason logically, and that is my forte.

I am a mathematician and so know how to reason from a priori assumptions.

However, this has never how I have done good mathematics.

Good mathematics is the creation ex nihilo of new productive axiom systems.

You do not derive new mathematics. It comes from another place entirely: the active imagination. It will brew awhile, in splendid obscurity, and then “poof” you have it.

Here is an example of that process.

Like anybody, I keep notebooks to jot down ideas and work them out. Whenever I do theoretical physics, I use reams of paper to grind out calculations. Inevitably, there are mistakes, and so I typically redo my calculations until I have it all straightened out.

I have enough of these notebooks, accumulated over forty years, to know the pattern of discovery and invention. The invention stage always starts with a marginal note. On the above page, you can see a very short note, dated, with a comment: Prove this.

There is some other stuff there, in a mix of colors, depending on what pen was lying around, some diagrams, which are not explained, and a boxed equation.

This was the origin of my 1991 discovery of the Classical Schrodinger Equation.

What you see above is the final result, after filling in the gaps, doing a lot of work on the physical interpretation, and checking of the mathematics.

The mathematics turned out to be new.

What you have above is the first example of Projective Hamiltonian Mechanics, a type of dynamical system that was not previously known to mathematics.

This mathematical system is not supposed to exist.

There is a celebrated no-go theorem called the Groenewold-Van Hove theorem which ought to tell you that Kingsley Jones simply cannot do what he has already done!

This is a 100% correct but sadly irrelevant mathematical theorem.

The Groenewold-Van Hove theorem makes the assumption that the algebra that is used to construct such a system is associative. This means that:

The mathematics that I developed, in the course of elaborating my discovery, turned out to be non-associative. It neatly side-steps the assumptions of the GvH theorem.

My purpose here is not to belabor a point in mathematical physics.

The point is as follows:

The necessary branch of new mathematics, which I call projective differential geometry, was only accessible by making an intuitive leap. The logical path to this discovery had long ago been boarded up as being an impassable road.

The person who makes the leap to a new discovery exercised their will power to go look where none had looked before. If you can see, you can look in new places.

Superintelligence entails the will to look where no other has looked before.

Look and you will find, if there is something to be found in that place you look.

This is why Albert Einstein was precisely correct in his assessment.

Imagination is more important than knowledge.

It is true, that the LLM has unleashed much potential from the interrogation of all human knowledge in question and answer form.

What matters is what new question you pose to the LLM.

The discovery above is one example of that process in action. In isolation, it matters little, but the iceberg it disclosed is massive. The underlying mathematics is rich and leads in unexpected directions. It remakes differential geometry, and with it the relationship between complex numbers and hypercomplex numbers.

What started forty years ago as a curiosity is now a viable pathway to a new physics that is fundamentally about kinematics and dynamics on projective spaces.

What is the “Hybrid” in Hybrid Intelligence?

There is a saying “to be in two minds” when deliberating on a course of action.

This is a metaphor for two distinct modalities of thinking. These are ratiocination, or the deliberative step-by-step process of reasoning things through, or the insightful path of recognizing what is truly new and novel about any given situation.

It is a truth of neurobiology that brains, human or otherwise, come in two parts:

The two hemispheres of the human brain approach tasks differently. The left brain is good at analysis, while the right brain is good at synthesis. We need both.

The present LLM trend emphasizes a Q&A style of discourse.

The presumption is that “intelligence” is rooted in the A: the Answer. However, there is just as much intelligence, possibly more, in framing Q: the Question.

Creative pursuits in mathematics and science involve framing good questions.

If you ask a new and novel question you can do good science.

The sum total of my life’s work in mathematics and physics was one good question:

How do kinematics and dynamics look when reframed in projective geometry?

This is not the declarative “Answer” style of discourse like: The Universe is made of strings! This is the “Question” style of discourse: What does dynamics in projective geometry look like? I am well on my way to answering that new question.

Conclusion

The reason that I think LLM technology is not the final word in Artificial Intelligence relates to the incompleteness of the conception. There is too much emphasis given towards knowledge, and not enough to the creative spark of reason. Certainly, this technology is very useful. However, it did not pose the question I framed above.

The “Hybrid” in Hybrid Intelligence is the ability to see both sides of the coin.

New discoveries flow from well framed novel questions.

The exhaustive search and checking of conclusions is valuable, but not as important as it first seems. The really critical step is that new step on a road not taken.

There is no essential physical reason why artificial intelligence cannot achieve this level of integration, but the present form appears lacking in this respect.

This is not a problem, as any intelligent human can plug that yawning gap.

We know that we can because the LLM brigade learned everything they know from us by being trained, in greedy stepwise descent fashion, on the fruits of our thinking.

This is no terrible thing if you know what to do with your mind.

Ask the probing unasked question.

Happy investing.

So much has changed in the last thirty years. The industry is almost unrecognizable.

The positive changes are easy to state:

• Greater cost efficiency in electronic trading.

• Improved global scope for research via better communications.

• Ever increasing computational power and data processing tools.

The negative changes are perhaps a matter of taste, but I will state my view of them:

• The decline of expertise-based business practice over marketing driven strategy.

• Ever increasing regulation to deal with the failures of investment competence.

• The trivialization of investment decision making about compelling narratives.

These are personal viewpoints from somebody who has had more than one career and grew up in a time before the personal computer and mobile phone arrived.

These have been very welcome changes in my lifetime, but I have noticed that with the apparent ease of analysis, using new technology, some skills soon atrophy.

I am writing this now because I think the growth of Artificial Intelligence will simply exacerbate the trend I mentioned. Expertise is now canned, why buy it?

The dumbing down of business to pure storytelling is likely to accelerate.

Of course, I own and operate a financial services business with my partner. However, that business, Jevons Global, is moving with the times to pivot towards research.

The idea of the managed fund, as a product, seems certain to go the way of buggy whips in the age of the automobile. The future of the horse and cart is niche.

I am not one to complain about this development. The fund industry fed housed and clothed me for thirty good years, so I feel no bitterness about such a change.

If anything, the time of life I am at, where I earn more from my investments than I do from my labor, means that my emphasis has shifted. The best application of effort for my time is to improve rates of return on my own portfolio assets.

While the regular fund industry collapses in a heap, the opportunity to segue thirty years of accumulated quantitative investing expertise just accelerated. The trend of industry development has made my own personal dream possible.

Remember, I started my career in mathematics and theoretical physics. I stopped that professionally because it made no economic sense. You could not make a living!

When I left “professional” science for “professional” finance it was an easy decision on how to pay the bills. Why solve a hard problem, like cosmology, or the foundations of projective differential geometry, when you could solve easy problems like “Excel”?

I (correctly) reasoned thirty years ago that all you needed to do mathematics was a cushy job, a banana lounge poolside in some exotic locale, and a good pencil.

The Viking bought me a Blackwing pencil subscription, which she regularly raids for those that glow in the dark, and is happy to leave me with the firm graphite.

There is a time and place for “soft” graphite but that is “Art” not “Science”. I use firm graphite for science, with soft graphite for the “surely you are joking?” marginal commentary, and the reverse for art. This is how it should be.

You cannot expect any banana lounge repository to stock the correct selection of pencils, so you must assuredly bring your own.

Paper has never fussed me.

Bus tickets. Ikea receipts. Tiki-Bar drink coasters. Meeting agendas… anything will do!

Now that I make more money from capital, than from labor, I can return to the active study of cosmological problems of significance and how to take a derivative on a projective algebraic variety. These are proper banana lounge grade problems.

However, I do crave attention, and so I am unlikely to shut up about finance!

Let us see what I thought ten years ago, before I update it.

The View from Ten Years Ago

The following is an extract of an invited talk I gave ten years ago.

The agenda, as you can see, covered what actually did happen.

I am not going to slavishly repeat all twenty-four slides.

How do we reconfigure our investment activities to make best use of people and machines?

There you have it. Nothing changed. The next ten years looks the same!

Death by Dis-Economies of Scale

Ten years ago, the managed fund industry did precisely what I expected it to do. To defend incumbency, it found new ways to make simple things more expensive.

The above slide shows why Jevons Global got founded in November 2011. I had just left a major investment bank where I was aghast that managing $35M AUD cost 40% more in percentage terms than managing $350,000 AUD on e-broker rails.

In Australia, the sum of $350K is considered to be threshold where it makes economic sense to consider setting up your own retirement savings fund. We call such a beast a Self-Managed Superannuation Fund (SMSF). When I worked in corporate, compliance did not allow me to run my own superannuation. I was forced to use a high fee fund where I was paying in excess of 2% per annum for indifferent performance.

As a professional portfolio manager, this really irritated me, so I left corporate to go explore how to properly arbitrage this gap. Fifteen years later I am unhappy to state that the commercial eradication of this gap, in Australia, is all but impossible.

Local regulatory conditions conspire against the obvious solution (the one that works in the USA), which is a hybrid financial adviser-brokerage model. The category in the USA is called a Registered Investment Adviser (RIA), and enables the use of electronic trading technology, like the one I use, from Interactive Brokers.

I founded Jevons Global to bring this RIA model of money management to Australia. I did not count on ASIC rules and regulations stopping it.

That project soon foundered on a set of local rules and regulations that seem expressly designed to protect incumbents from such competition.

The result has been the growth of a protected species called the “wealth management platform” which is an expensive version of a US RIA model with some really indifferent trading technology behind it, a sheltered workshop business mindset, and indulgent regulators who think they “solved” the problem of shonky financial products by outsourcing any serious due diligence to an army of “pay to play” gatekeepers.

Events turned out much as I anticipated.

The industry found a way to deliver inferior returns at greater protected overheads!

The Unbundled Fund Solution

The logical path of development for greater efficiency is unbundled services.

Following my early encounter with the Australian regulatory exclusion of the lower cost US RIA model, I pivoted to unbundled model portfolio services.

The managed investment fund was really created in the 1950s era when prosperity widened in the USA, and it became feasible to mass distribute investments. Mutual funds grew up in this era, and their uptake was accelerated by adoption of scale efficiencies from computing, in accountancy, trading and marketing.

If you consider the menu of services above, it was rational to bundle those together into a combined service offering, perhaps splitting investing and compliance from each other, since the left is more amenable to economies of scale.

The domestic growth of the SMSF industry, much like the US IRA and other savings vehicles for self-directed investment is one good example of this in action.

The problem for Australia, the one we are living through right now, is that because the Australian Securities and Investment Commission (ASIC) effectively shut down growth in Managed Discretionary Accounts (MDA), which is the local version of a US RIA, the industry naturally pivoted to raid SMSF funds for real estate investment.

In Australia, you do not need to be licensed to advise on real estate, and so there has been nothing to prevent advisers from latching on to debt hungry strategies they can market to retail investors in SMSF vehicles, in return for trailing commissions.

Australia has a habit of blowing itself up this way.

It is happening again.

The Pressure from Artificial Intelligence

When I first gave this talk Deep Learning had just broken onto the scene.

It is funny to think of how a seemingly innocuous development, like a new learning algorithm that scales well on video game hardware, can become a monster trend.

Ten years ago, the talk I gave was all about Big Data.

I spent more time talking about that…

If you cast your mind back, ten or fifteen years ago, methods to tame, inspect, and analyze, massive data sets were at an absolute premium. Visual was the way to go.

This has changed now, since the textual input and output mode of the LLM, has given new and more flexible means to interrogate data in a natural human fashion.

The next wave is likely to be a better fusion of these two modes of understanding in more and better tools that combine both channels: readable text and visualization.

Humans are good at both gestalt “big picture” high bandwidth visual understanding and logical “down in the weeds” linear textual analysis. The advent of social media blurred those lines already. There are bigger opportunities ahead.

The Age of Massive Parallelism

This is not a piece on semiconductor design and architecture. I will say more about that in my look ahead piece to come, because I think it will change radically.

For now, it is enough to know that the Graphics Processing Unit (GPU) is ascendant.

Earlier, I mentioned interest in visual analytics. I focused on this due to my experience with huge trading datasets based on tick data, individual trades and orders.

The above selection of chip architectures is informed by that background.

Almost nobody talks about Field-Programmable Gate Arrays (FPGA) today, but they have played a huge role in High-Frequency Trading (HFT), as a way to make custom hardware, in low production runs, for extreme low latency performance.

I will say more later, but this has a future in robotics, I am sure.

The top left architecture, the traditional many-core CPU now dominates with the GPU at lower left, but the key word is parallelism. When I wrote this my own lab had some examples of all but the FPGA, for evaluation purposes. The Intel MIC at lower left was junk, but I needed to know that, so I bought one cheap. It was junk.

What we have now is a computing world driven by two scaling laws.

The present headlines are dominated by the huge energy and cooling demands of the data centers that train and run Artificial Intelligence models in massive clusters.

What drove computing up until about 2005 was the application of Moore’s Law on the number of transistors doubling every eighteen months or so in a model where finer chip features were met with a natural increase in CPU clock speed.

This model hit the wall with the death of Dennard Scaling.

In simple terms, the move to smaller features could not be met with an increase in the rate of computation via the simple device of increasing the clock speed. The chip die simply heated up more, and cooling became a problem.

The new route to enhancing throughput came via the little-known Amdahl’s Law.

The essence of this scaling law is that the computing speed-up of adding more and more processing units is limited by the proportion of so-called serial work.

Serial work is the kind that cannot be broken up into separate non-communicating tasks. The factory analogy is final assembly of a finished product. You could readily speed-up manufacture of each part, but final assembly requires a mating of parts.

In a factory, you can do that by making more than one final assembly station.

The challenge with running very large AI models today is that Amdahl’s Law does limit the throughput from slow serial tasks. This is why innovations like Mixture of Experts (MoE) models from Chinese firm DeepSeek were rapidly adopted. Anything that can decouple workers from direct blocking communication can speed things up.

For a later discussion, I will mention Tau-Scaling from Huawei. This is a candidate new way to think about scaling computers to complement Moore and Amdahl. The jury is out about whether He’s Law (named for the CTO of Huawei) is the answer.

What we do know is that human intelligence takes about 1.3kg of grey matter, which occupies about 1,100 cubic centimeters of space (10cm cube), and consumes 20W.

There are folks that think current GPU’s are the final word in AI compute.

They are undeniably good, but I expect a replacement.

Business Models

The world of business models ten years ago portended where we are today.

The key word is “Robo” and those trends now look to be accelerating.

The one point I would make here is that the biggest area I see for active application of Artificial Intelligence is accelerated volumes and tempo in digital marketing.

There is no need for a robot to be truthful to be an effective marketer.

Any obsession with factual outcomes is a handicap to closing the sale.

That might sound a little churlish, but it is what I think.

For these reasons, I have my own priorities for the journey ahead.

How can I apply technology to protect myself from unsolicited robot actions?

You know what I mean…

Imagine what happens when that call center that now bugs you every dinner time to sell you solar panels, a new mobile phone plan, or cheaper power, suddenly scales upwards by a factor of a thousand, to sweep more pennies to the bottom line?

That is one Yuga I can do without :-)

Conclusion

That was a ten-year rear vision mirror view.

Having spent this afternoon reviewing what I used to think, I have come to the not very surprising conclusion that the future will bring more automation.

However, I also think there will be a further decline in manifest corporate expertise.

It has been my experience that corporations get dumber over time.

I expect this condition to accelerate under the pressures of A.I. adoption.

This will likely lead to opportunity, but one needs to protect oneself.

I think the new direction is more self-directed investing.

The business direction of Jevons Global is in flux, but I think I am likely to double down on research services and open-source investment systems.

High-performance computing is now super-abundant at falling cost.

Previously, I thought of selling such services to other people. Now the cost to produce the service is so low, that the better application of it is to my own portfolio.

I could make zero dollars by selling the service or I could make a decent return on my own capital applying that internally. The second option is the rational one.

However, as is true of my scientific activity, the only way to keep yourself honest is to put your thoughts, observations, and creations before others for critique.

That is the upshot of this personal reflection.

Stay tuned for the detail of what comes next.

Happy investing!

Subscribe now

The industry was already fighting the rise of low-cost passive investment vehicles, like index-driven investment delivered via Exchange Traded Fund (ETF) vehicles.

I wrote about this trend two years ago:

Going global with ETF product

The benchmark for my global equities model portfolio is the Vanguard Total World Stock ETF VT. This tracks the FTSE Global All Cap Index, across 48 countries, 10113 individual stocks, and is 62% in the USA, for a bargain 0.06% MER fee.

I aim to beat this portfolio, in previous years we have, this year we have not.

The logical benchmark for US strategies is the S&P 500, which can be bought via ETFs. The one I use is the iShares Core S&P 500 ETF IVV at a 0.03% MER fee.

For flicking between a US and a non-US exposure, a good choice is the MSCI All Country World Index (ACWI). The so-called IMI Index is free float weighted. This means that only readily tradeable shares are included in the weighting.

You can buy the iShares Core MSCI Total International Stock ETF IXUS at a 0.07% MER.

For interest, let us just compare the YTD USD total return to 30-Sep-26:

VT: Vanguard Total World Stock ETF 12.86%

IVV: iShares Core S&P 500 ETF 12.72%

IXUS: iShares Core MSCI Total International Stock ETF 13.67%

The higher performance of the IXUS over the Vanguard VT relates to the greater breadth of the Vanguard product. It goes deeper into Emerging Markets.

These are all figures in USD for the US Exchange Traded product.

After thirty years trading in global markets, I prefer to access the cheaper, or free, brokerage, deeper liquidity, and lower fees of the US market, alongside direct equity exposures in Europe, Japan and Hong Kong.

If I looked at the Australian market from the perspective of a US domiciled investor, I would likely view the iShares MSCI Australia ETF product EWA, as the easiest marker. The MER on that is 0.50%, which is much more expensive than the above ETFs.

If you look at the sector breakdown, Australia is mostly a financials and materials market with the other sectors very distant in the capitalization rankings.

The top five single stock holdings of EWA are:

BHP 15.42%

CBA 12.61%

NAB 6.09%

WBC 5.99%

ANZ 5.77%

for a cumulative total of 45.88%. From a global perspective, that is one quality diversified mining company and four expensive low-growth banks.

Only Australian taxpayers qualify for the so-called franking credit, which is a deduction of company tax paid from the dividend payout.

Australians love this, so they bid up fully franked income stocks. Global investors do not see any benefit from this measure, so they generally find them unattractive.

For the global investor, the right comparator is the net dividend paid less withholding tax paid in the offshore jurisdiction, with any tax treaty benefit deductible in local tax returns. For Australian residents, this means that the 10% US withholding tax paid on dividends received in that domicile, provided you filled in the W8BEN IRS form, is deductible from your personal or SMSF tax return when filing in Australia.

The net of this is that grossed up yields on Australian listed stocks are generally comparable to the net yield on foreign stocks, provided there is a tax treaty.

Here is the approximate yield comparison on US ETFS, held by an Australian resident, after allowing for the tax-treaty deductibility of a 10% withholding tax. Effectively, you add that back, and so compare the gross yield on US non-Australian ETFS, to a local ASX listed counter, like the iShares Core S&P/ASX 200 ETF IOZ at 0.05% MER.

VT: Vanguard Total World Stock ETF Yield 2.85%

IVV: iShares Core S&P 500 ETF Yield 1.06%

IXUS: iShares Core MSCI Total International Stock ETF Yield 2.85%

IOZ: iShares Core S&P/ASX 200 ETF Net Yield 3.40% Grossed-Up Yield 4.86%

There is about a 2% yield pick-up on an estimated after-tax basis for buying local exposure to Australian stocks for an Australian domiciled investor.

This apparent allocational advantage fades once you consider strategies that are globally diversified and specifically target dividend yield. The iShares International Select Dividend ETF IDV has a 0.50% MER the same as MSCI Australia EWA ETF.

Note that yields are comparable in spite of base currency differences. While the IDV is in USD, and the IOZ is in AUD, the yield comparison is not greatly affected by any changes in currency. Nonetheless we can compare EWA directly to IDV.

Here is the sector breakdown of the IDV product:

You can see that it is similar to Australia on Financials but has much better diversifying qualities with respect to other sectors. Traditional yield targets like Energy, Utilities, and Telecommunications feature prominently on the sector exposure ranking.

When we add all of that up, the USD comparison of EWA and IDV reads:

EWA iShares International Select Dividend ETF Yield 2.83%

IDV iShares International Select Dividend ETF Yield 5.03%

There are a few caveats with those comparisons that investors need to be aware of. Within a UK heavy exposure, the high-yield sector will include tobacco stocks.

IDV is no exception. The second-largest weighting is to British American Tobacco. You will typically find that in global dividend focused products. The solution, as always, is simply to check the holdings of the product for anything you do not like.

In any portfolio where you target single stock holdings you can always use the ETF listings on a site like iShares to look for ideas. Take the thematic you are targeting, which might be a financial objective, like income, or a growth exposure, like A.I., and list out the top ten global holdings. Review that to locate opportunity.

What about A.I. Screening?

The above is uncontroversial and practically directed exploration of likely securities you can put together, at very low cost, and little ongoing maintenance, to build out your global portfolio., In doing so, the research dimension and due diligence can be time consuming. Here is the first and most obvious place A.I. Systems can help.

Personally, I use LLM chat-based systems a lot, but I do so for lines of inquiry which I can vet and check for accuracy. I treat it as a time-saving tool to help me assemble information I need to make a decision, not the decision itself.

This may seem overly cautious to some, but is a lesson born of experience.

Some twenty years ago, at the dawn of electronic trading, I was leading Quantitative Trading Research for a large New York based firm that was rolling out automated trading systems to handle our $400B USD book of equity investments.

You learn a lot quickly when faced with such a large quantity of capital at risk.

On more than one occasion our team was saved by kill switches, filters, and triggers that were designed to stop the automated systems from doing stupid things.

The one I recall best is when some glitch in network code caused our system to place the same portfolio trading order for $200M USD gross of buys and sells into broker algorithms more than 40 times over. This meant $8B USD of trading!

Fortunately, on that occasion, automatic filters and kill switches blocked the trades from being executed, and we avoided a massive trading error cost.

Others were not so lucky.

On one infamous occasion, the High-Frequency Trading (HFT) firm Knight Capital Group went broke in under 45 minutes with a devastating software glitch. On the morning of 1 August 2012, the Knight Capital Group trading systems, which had a new software upgrade, erroneously flooded the market with orders due to a very subtle bug in the way the new system interacted with legacy systems. The robot became blind to the result of its actions, and kept repeating them!

The collapse of Knight Capital Group took only 45 minutes!

The death spiral was clear to all in the market inside 15 minutes and there was nothing management could do to stop the collapse. It went broke.

Since I have spent thirty years in markets, and before that some time in aerospace and defense R&D, I am always very cautious with the engineering of automated systems.

The old term is fail-safe.

When a system fails, due to unexpected data, and outage, or a software clash, there needs to be some higher-level function that automatically intervenes to offset the possibility of disaster by reverting to a different system of control.

In humans we call this quality consciousness.

It is what redirects our attention, in an instant, to take preventative actions.

One obvious example is the reflex action to grab a handrail if you trip on stairs.

However impressive current LLM systems may be in basic question and response cycles they are very clearly not conscious. It is easy to take any such system and wander into the realm of hallucinatory response and manufactured illusions.

This is just a fact of years of experimenting with such A.I. Systems.

I rate them positively for utility, but am very cautious in allowing them free reign to act without some outer control loop that can reliably detect destructive actions.

The story I often relate, from aerospace engineering history, is that of the fatal crash bv NASA test pilot Michael J. Adams, in an X-15 experimental spaceplane.

The fatal X-15 Flight 3-65-97, occurred on 15 November 1967, being the 191st flight of the experimental spaceplane. A number of factors conspired to cause the crash, which involved the spaceplane entering a Mach-5 spin on descent from 70,000m.

Adams was able to counteract the spin and recover at 36,000m altitude, but in an attitude that was inverted, with a subsequent Mach 4.7 dive. The control system was adaptive, with a high gain, meaning a sharp response to any deflection, that forced the system into an unstable condition known as a limit cycle oscillation.

The easiest way to explain that is to watch a video of pilot induced oscillation. The one below is from a 1970s test of fly-by-wire in an F-8 jet at NASA Dryden. If you stick with what looks to be a benign landing, to about the 20 second mark, you will get it.

The phenomenon of control system instability due to overcorrection, which then leads to amplification of initial trajectory errors, is known in engineering systems theory.

Nonlinear and adaptive systems are particularly prone to this problem.

Michael J. Adams did a splendid job of recovering from a Mach 5 spin (imagine that if you can) but could not cope with + and - 15G forces induced by the controller when it went into that limit cycle oscillation. He would have been unconscious at death.

The key point for readers to understand, is that LLM Artificial Intelligence models are nonlinear adaptive systems. Unlike the ones in engineering, there does not appear to be any serious effort by the creators of these systems to analyze the relationship between input and output to assess controllability, and a fail-safe mechanism.

Contemporary AI engineers appear to assume that because their systems are called “intelligent” they will engage in intelligent behavior under all circumstances.

Since I am a mathematician and can readily comprehend the structure of a deep neural network as a stacked system of iterated nonlinear maps, I have precisely zero confidence that any such system built today is stable.

There are no mathematical guarantees of stable system action.

Absolutely none whatsoever!

In the wild, under conditions to be discovered by the fates, I would happily wager that every single one of the current fielded systems is intrinsically unstable.

There is no evidence that I can see, from any of the published literature by the current leaders of Artificial Intelligence research, that mathematical stability is considered.

While the systems are in the class of iterated function systems, which are famous for their ability to produce chaotic relationships between input and output, the industry is flying blind on the consequences of fielding such active adaptive controllers into the very highest risk scenarios: military; medical; and financial.

While I do not wish to be a party pooper the outcome is 100% predictable.

These A.I. systems will fail in spectacular ways until the industry grows up.

You can positively bet on that being the outcome.

How to use A.I. Systems safely?

The basic issue is commonsense.

Only deploy an A.I. system with a strict outer control loop.

The correct fail-safe is a limited loss restriction.

In a financial context, this involves using that system only when it is possible to put a guardrail around it to detect and mitigate errors. The clear one for a question-and-answer style chatbot is to pose questions whose answers can be checked.

For instance, earlier we noted that there are a number of global dividend ETF products that could provide attractive yield compared to the Australian market. One downside which may appear is inadvertent exposure to tobacco stocks, in the UK, USA, or the Japanese market. This is not something you worry about in Australia.

Potentially, this is a very tedious exercise, but LLM chatbots can help.

The question is pretty clear, the target ETF products are known, and the answer can always be manually checked, which is the outer control loop on the bot.

I did this with Microsoft CoPilot (it is free):

Q: Can you review the ETF listings opposite, which are iShares ETF products targeting income and list out some suggestions that have low or no tobacco stock holdings?

this with the iShares dividend ETF listing open.

It answered with:

If your goal is income ETFs with low or zero tobacco exposure, the strongest candidates from the list are:

Likely zero tobacco

• DIVB – iShares Core Dividend ETF

• DGRO – iShares Core Dividend Growth ETF

• IGRO – iShares International Dividend Growth ETF

Likely minimal tobacco

• HDV – iShares Core High Dividend ETF

• BIDD – iShares International Dividend Active ETF

Likely meaningful tobacco exposure (avoid for your screen)

• DVY, IDV, DVYA, DVYE

This is easy (albeit tedious) to check.

• DIVB has MO Altria Group

• DGRO has PM Philip Morris International

• IGRO has BATS British American Tobacco

• HDV has PM Philip Morris International

• BIDD has BATS British American Tobacco

• DVY has MO Altria Group

• IDV has IGRO has BATS British American Tobacco

• DVYA has JT Japan Tobacco (JT)

• DVYE has KT&G Corp (Korea Tobacco & Ginseng)

Actually, you can use the AI to confirm all of that if you have the iShares tab open on the holdings so it can go scan the list. When I say “confirm” you need to understand that one response is actually incorrect. DVYA has no tobacco stocks!

Q: I think you are wrong about DVYA. When I checked it had no Japan Tobacco.

You’re right to push on that—good catch.

Based on current DVYA holdings, there is no Japan Tobacco and, in fact, no explicit tobacco stock at all.

From the latest holdings data (61 positions, top weights like BHP, DBS, OCBC, Honda, Fortescue, ANZ, Santos, Westpac, UOB, Power Assets), the consumer defensive slice is made up of names like Wilmar International and Budweiser Brewing, not tobacco.

So:

- **DVYA today:**

- **No Japan Tobacco**

- **No direct tobacco holdings in the published list**

- Consumer defensive exposure is food/beverage, not cigarettes

My earlier answer leaned on index-structure priors (Asia/Pac high-yield → almost certainly JT) rather than the live holdings, and that’s where it went wrong.

You were correct to override it with what you saw on the page.

This happens all the time with A.I. when I use it for financial research.

The tools are helpful, so long as you know what you are doing.

This is not a problem solved by the non-solution “use Claude”. Every system will substitute informed guesswork in a seamless fashion with factual data.

Of course, you can use advanced system prompts that instruct the system to rely on factual results, with references, in preference to guesses, or to not respond.

Theoretically, that sounds like a workable solution.

In practice it will not be, because so much internet content is now machine generated. How is the machine to know what is good or bad data? How are we to tell?

The need for sense checking results has gone up, but the potential for organizing data to help make decisions has also gone up. The challenge is to balance risk with reward.

In my case, in the course of writing this note, I learned something useful for my own global portfolio. Among the usual suspects, one dividend ETF stood out:

DVYA: iShares Asia/Pacific Dividend ETF

It is the only one that does not currently hold any tobacco exposure (I used to be a chain smoker thirty years ago and know how hard it is to give up).

The 12M trailing yield is 4.53%, and the MER is manageable at 0.49%

When I look at the sector exposures it feels natural to an Australian:

When I look at the geography I can see why:

Would I buy it in my Self-Managed Super Fund (SMSF)?

Certainly not, for an obvious reason.

I am an Australian tax resident and so benefit from franking credits.

It makes no sense for me to buy BHP in this fund over holding it directly, through an Australian listed security, for which I get the benefit of franking.

If I were in the USA, and looking for a non-tobacco dividend ETF, this is a good diversifier towards Asian markets with a dividend bias.

Looking at the holdings, I own the stocks in green already.

For diversification, perhaps I will look at the others which are not Financials.

This is an example of how research can benefit from the use of A.I. in a sensible and vetted fashion. These days you can trade pretty much anything in a global broker account, but you need to narrow the range of options.

Filtering by dominant ETF membership, seeking your chosen attributes, and working to exclude unwanted elements is a productive and reliable short-cut.

You do not want the perfect to be the enemy of the good.

However, you also want to know what is in the tin.

I have no doubt AI systems can help here.

You just need to be watchful for how they fail and be on guard for the most predictable failure of all: these systems are engineered to please you.

Skepticism is required to correct for the obvious bias to pleasing fiction.

This should not be unexpected for the seasoned player in financial markets.

We are all part of a global circus to profitably separate fact from fiction.

Happy investing!

Subscribe now

• Australian 20 Stock Model Portfolio

• USA 20 Stock Model Portfolio

• International (non-USA) 20 Stock Model Portfolio

• Global Best Ideas 25 Stock Model Portfolio

There are no changes for the global model this month.

Subscribe now

The current portfolio snapshot is detailed below.

Read more

• Australian 20 Stock Model Portfolio

• USA 20 Stock Model Portfolio

• International (non-USA) 20 Stock Model Portfolio

• Global Best Ideas 25 Stock Model Portfolio

There are no changes for the international model this month.

Subscribe now

The updated performance and current portfolio are given below.

Read more

• Australian 20 Stock Model Portfolio

• USA 20 Stock Model Portfolio

• International (non-USA) 20 Stock Model Portfolio

• Global Best Ideas 25 Stock Model Portfolio

There are no changes for the USA model this month.

Subscribe now

The updated performance and current portfolio are given below.

Read more

• Australian 20 Stock Model Portfolio

• USA 20 Stock Model Portfolio

• International (non-USA) 20 Stock Model Portfolio

• Global Best Ideas 25 Stock Model Portfolio

There are no changes for the Australian model this month.

Subscribe now

Late Cycle Conditions in Australian Markets

In general, global markets are churning on performance chasing. There are clear signs we are now late cycle in Australia, with housing weak, but inflation sticky. We expect further interest rate rises from the RBA in the coming meetings.

The reaction to a soft earnings report from portfolio holding JB HiFi, signals potential weakening across the broader consumer markets, although sustained high levels of wealth in Australia are propping up spending among older consumers.

Due to rising, but comparatively low levels of unemployment, we do not expect there will be a recession. However, the Australian economy is stuck in low gear as it shifts from a housing investment led phase, to belatedly address productivity issues.

The politics is likely to become increasingly fractious as the electorate looks to find someone to blame for what is (frankly) a self-inflicted predicament.

Making houses more expensive does not produce sustainable wealth gains.

The housing shortage is likely to persist, in spite of slower immigration, and there is no easy way out of the bind. You cannot fix ten to fifteen years of neglect towards productivity enhancing capital investment that is intensive and not extensive.

The essential problem for Australia is that it bet too heavily on scaling up, in terms of population growth and public infrastructure to support that, without any clear policy focus on how to improve unit economics and productivity within the economy.

There is a solution to this problem.

• Less intrusive regulation for the sake of regulation.

• More effective government planning to mitigate obstacles to investment.

• Adoption of more proactive business attitudes towards R&D.

There has been much written on the government side of the equation, but there is a clear problem with business also. Australia has very low levels of corporate R&D and poor outcomes on the development of new products, processes, and technology.

This problem shows up in global metrics of economic complexity, where Australia rates among the lowest of all nations. The issue is one of complacency.

It is not enough to simply mine minerals and speculate on house prices.

Having said that, the only part of the Australian economy doing very well right now is the mining industry, and the one clear area of shortage is housing construction.

Australia has a quality economic problem to solve.

Even the folks who think they have correctly diagnosed the disease are reluctant to move capital away from the three areas that benefited from globalization:

• Mining to provide the resources for global development.

• Finance which recycled the export dollars earned into domestic credit growth.

• Housing which booked the capital gains from excessive credit expansion.

In our own portfolios, we have stepped back from finance to underweight, but are overweight minerals and energy, and housing and data center developers.

This positioning expresses a pragmatic viewpoint.

Australian mining will continue to do well because it does not depend overly much on the health of the domestic economy. However, Australian banking is already looking somewhat shaky, due to slowing credit demand, high valuations, and weak housing markets. While we are now in the early stages of a construction shake out, due to overleverage in players like Baltha Group, there is a supply shortfall of housing.

In our view, it will take some time for Australia to find a new economic growth model. Perhaps it involves Artificial Intelligence in some way but the path to that outcome is fraught with risk to near-term employment, and the generally low appetite for R&D among Australian corporates. I have had 40 years exposure to R&D at a high level. With that experience, and knowledge of how Australian corporations operate, I am not optimistic that the transition to a new growth model will be smooth.

It will take serious effort to be remotely competitive in R&D led business, when you have the USA and China both fiercely competing at the cutting edge.

There are models that might work, such as what Singapore, Switzerland, or the Nordic nations do, but Australia does not seem to have the application to pursue that.

We do have a quality Higher Education sector, but the capacity was diverted towards rent seeking from the foreign student trade. This has proven to be short-sighted.

While Australia did train many foreign students at a high level, we did not develop the domestic industrial base to exploit that talent once educated. This is in contrast to the United States, which has been a huge beneficiary of imported talent.

Now that the Australian political conversation has regressed to anti-immigration, and not rationally directed immigration policy, there is little prospect of the local polity suddenly fastening upon any good plan of what to do next.

Australia sleep-walked into two full decades of rising prosperity based almost entirely on the recycling of foreign earnings into credit expansion to fund a housing boom.

The inevitable bust is now upon us.

Since very few market players could correctly diagnose the proximate reason for the good conditions on the way up, I doubt that very many will see the way out.

The way out is time, a painful period of adjustment, capital preservation against the market volatility, and a clear weather eye on the search for new opportunity.

In the Australian market, we think minerals for the energy transition remain key, and we also have the benefit of a large quality gold sector, the third largest after China and Russia, plus iron ore, copper and uranium in majors BHP Group and Rio Tinto.

The Australian banks used to be a good place to invest, but we have shifted our focus to Asia, with Singaporean and Hong Kong financials among our key bets.

Domestically, the Australian economy still needs energy infrastructure, through firms like APA Group, providing gas pipelines, and rail transport, via Aurizon.

In real estate, we think the parameters of government policy will settle on new build, as we have already seen with the negative gearing changes. There is a housing slump in progress, but Australian housing would need to fall about 30% to be fair value.

This will not happen, but the construction shakeout now in progress will benefit those developers with stronger balance sheets in growth corridors like Stockland Group.

Concerning the rest of the Australian market, not much interests me. The technology of tomorrow is being built in China and the USA, not Australia. Our market has a very nasty habit of paying a global premium for indifferent technology exposures that will collapse in valuation once the global leaders correct. That time looks close upon us.

Due to the housing correction, and wealth effects coupled with a clear retrenchment in high rates of immigration, the Australian consumer sector is unattractive.

The Australian health sector is also small, by global standards, and faces the serious challenge of more aggressive US policy to force concessions on pharmaceuticals.

These factors will likely pile up once China makes further inroads into developing its own for export medical devices and pharmaceuticals industry.

Oddly enough, the one great positive I see for Australia is the diverse ethnic ties our large immigrant population has with other parts of the world. This matters in the age of connected informational markets that depend on local knowledge and contacts.

Australia has hybrid vigour that will likely see it adapt to the new conditions from the ground up. I cannot see this happening at the corporate level. There are too many large and sluggish oligopolies in Australia that were wedded to “Just scale up!”

The just-scale-up mantra will not work anymore.

Who knows what adjustment pathway Australians will choose?

While my tone may appear negative towards many of the leading companies of today, I am not negative on the Australian project. Australians will figure out something.

In between, I am resigned to a period of largely counterproductive bickering.

Until that fog clears, I will hide out in Australian miners and those areas of the economy where I perceive a clear shortage and demand support.

The updated performance and current portfolio are given below.

Read more

• Australian 20 Stock Model Portfolio

• USA 20 Stock Model Portfolio

• International (non-USA) 20 Stock Model Portfolio

• Global Best Ideas 25 Stock Model Portfolio

There are no changes for the global model this month.

Subscribe now

The updated performance and current portfolio are given below.

Since inception, the performance remains positive.

The ratio of absolute return for risk remains very favorable.

In virtue of the churning market conditions, we are sitting tight this month.

The current portfolio snapshot is detailed below.

Read more

• Australian 20 Stock Model Portfolio

• USA 20 Stock Model Portfolio

• International (non-USA) 20 Stock Model Portfolio

• Global Best Ideas 25 Stock Model Portfolio

There are no changes for the international model this month.

Subscribe now

The updated performance and current portfolio are given below.

Since inception, the performance remains positive.

In virtue of the churning market conditions, we are sitting tight this month.

The current portfolio snapshot is detailed below.

Read more

• Australian 20 Stock Model Portfolio

• USA 20 Stock Model Portfolio

• International (non-USA) 20 Stock Model Portfolio

• Global Best Ideas 25 Stock Model Portfolio

There are no changes for the USA model this month.

Subscribe now

The updated performance and current portfolio are given below.

The main theme in this report will be to look at mean reversion over short periods of one month, and to relate this to the wild swings of relative performance this year.

To set the scene, have a gander at this alpha chart.

The issue driving this is two-fold:

• Mean reversion on a one-month basis is very high right now.

• The extreme concentration of the benchmark produces a large tracking error.

In the section below the line, we make a turnover analysis of the strategy, since the inception on 18-Jan-2024, just to check we are not trading excessively.

The summary is that we are not, but you can see from the monthly returns below that the strategy has a tracking error (relative risk) of 7.98%.

The annualized risk of the portfolio is 11.96%, which is pretty close to the S&P 500, at 11.93%, but the relative returns are varying widely, due to concentration risk.

In the USA, the top ten stocks in the S&P 500 are now 40% of the benchmark. We own four out of ten, so the swings in the other six have a big effect month to month.

Just to test whether we are going mad, I compared the total returns in June of the twenty stocks we own (Prior Month) to the same stocks in July (This Month).

This is a scatter plot, where each point is one of our portfolio stocks. The line slopes down and to the right, because those stocks that did well in June did poorly in July and vice-versa. The slope of this line is -.72, a perfectly negative correlation is -1, and the so-called R-squared statistic is 0.389, meaning 38.9% of variation is captured by this simple linear regression. We did adjust weights on 7-Jul-26, but we owned the same stocks throughout. We just upweighted prior underperformers.

You can see that this effect is very powerful at this time.

Investors seem to be madly chasing performance, selling what recently did poorly to buy what is currently doing well, and that is proving to be counterproductive.

As I said, we are making no changes to the model this month.

However, mindful of this current reversion regime we have upgraded our toolset to give a better read on our own trading history and turnover.

In addition to the usual portfolio snapshot, as at 7-Aug-26, I include all of the trades and turnover since inception. We will use to help guide our rebalancing.

This year we traded significantly to position for possible stagflation, and the ongoing data center and AI boom. Since the top ten stocks are 40% of the benchmark, it can be very difficult to outperform at our target 30% turnover.

On the basis of our recent study, a turnover budget of 30% to 60% seems more like what is needed to navigate this concentrated and volatile market

Read more

• Australian 20 Stock Model Portfolio

• USA 20 Stock Model Portfolio

• International (non-USA) 20 Stock Model Portfolio

• Global Best Ideas 25 Stock Model Portfolio

There are no changes for the Australian model this month.

We will likely make some tweaks to the Global models in the next day or so.

Subscribe now

The updated performance and current portfolio are given below.

Read more

• Australian 20 Stock Model Portfolio

• USA 20 Stock Model Portfolio

• International (non-USA) 20 Stock Model Portfolio

• Global Best Ideas 25 Stock Model Portfolio

It was a great year for the Australian portfolio, a so-so result for the USA, and a poor result for the International 20 and Global 25. The last was driven by underperforming selections across all sectors except resources, in which we had good wins.

The annual returns in USD were:

Portfolio: +10.83%

FTSE Global All Cap: +24.70%

Alpha: -13.87%

The sector report provides the cleanest read on what happened.

The only sector to show a positive contribution was resources. This explains the disparity in performance between the Australian and Global strategies.

We got the home story right in gold, battery minerals, and iron ore.

Everything else was not so great, and not down to one country.

These were conditions where most of what we picked did not work. The detail of this negative skew for stock selection shows up clearly in the contribution analysis.

Portfolio diagnostics are important when deciding what to do, if anything. The above chart shows that we had very significant negative skew holding many sensible stocks. This matters a lot if the absolute return of the strategy is negative. In the year we just ruled off the return was +10.83%, which was a large -13.87% lag to benchmark.

The results of global managers are not out yet, but many will be bad, like us.

The markets were narrow.

South Korea, ARCA: EWY, was up around 170% thanks to High Bandwidth Memory (HBM) makers Hynix and Samsung. The USA version, Micron, XNAS: MU, was up around 800%, while Nvidia XNAS: NVDA, finished up a mere 25%.

When you have a very narrow market like this the performance will naturally divide between those on the hot trade, and those left behind everywhere else.

The important thing is that our absolute returns were positive.

Some areas of the portfolio, like China technology listed in Hong Kong, and the main technology stakes we hold in US hyperscalers, like Microsoft XNAS: MSFT, had a poor showing as investors sold those to chase performance in memory and storage.

You can see the full breakout of performance here.

Narrow markets can persist for some time, but we think this is a frothy market that will reward patience. The rational thing to do is to modestly play mean reversion.

In the changes this month, we trim some winners and upweight some value laggards.

Read more

The performance numbers for Australian FY2026 were:

Portfolio: +2.46%

FTSE Global All Cap: +24.70%

Alpha: -22.24%

There is no polite way to state that result other than to recognize it as terrible.

The rest of this report digs into that performance and outlines our response.

The effect on our track record is to render it so-so since inception.

Around May-2025 was the time to crow, now we have to eat the crow. Run, Mr Crow!

There is a deeper issue to report here, and I will elaborate, but you will notice from the graph of cumulative outperformance that it is quite volatile.

Up, down, up more, down more.

The first instinct, when faced with terrible results, is to throw out the bad and ship in the good, in order to address the cause of the trouble and turn the ship around.

This is not a terrible approach to investment management but can be hazardous when running global portfolios. There are two reasons for my caution:

• Benchmarks are subtle creatures in global investment because of what they hide.

• Turnover is the enemy of after-tax returns and prone to mean reversion drag.

Let me return to what happened in late 2024 and what we did about it.

On 4-Oct-2024, in the midst of declining alpha, we did this rebalance

The heroic changes made then were to ditch semiconductor equipment firm ASML Holdings XAMS: ASM, and to buy Tencent Holdings XHKG: 700. They both did well, although Tencent did well earlier, and ASML later. Later it did very well indeed.

That was the only rebalance until 30-Jan-2026, fifteen months later.

Over the next six months, that portfolio rebounded 23.8% in alpha terms.

This is not an isolated experience for me and is one of the key challenges I had to overcome when first managing global portfolios.

It has to do with mean reversion.

The Power of Mean Reversion

This is supposed to be an educational blog about portfolio management.

Of course, I want to provide good advice and suggested investments, but I also want to step away from what I personally think about any stock, at any given time, which is a hugely fallible undertaking, and impart what I have learned about the generalities.

There are a number of maxims I am sure you have heard:

• Cut your losses and let your winners run.

• Nobody ever went broke taking a profit.

• No stock is ever too high to buy or too low to sell.

• Buy a good stock and wait till it goes up, if it don’t go up don’t buy it.

• The stock market transfers money from the impatient to the patient.

You can probably recognize a few famous investors in among that lot.

It is the last aphorism that I will rap on in this note.

When things go wrong, pay attention to the timescale of measurement, and the features of how you choose to measure return and ascribe your judgment.

In explaining exactly why global portfolio management needs to be treated with respect, when making judgments on what to buy and what to sell, it helps to be familiar with the statistical forces at work in return measurement.

This can be a very sobering experience for anybody who considers their results to be good on the basis of a short time series. The nature of the mathematics that I will relate highlights that we may none of us live long enough to prove we are good.

The consolation prize, for all global managers, is that if your tracking error to how you are measured, via your benchmark, is large enough, it is next to impossible to declare that you are most definitely bad. This is with one exception, absolute return.

Once we are through this discussion, it will be a little easier to understand why I make the changes to the portfolio as I do: not a lot of panic; a decent measure of hope.

If your absolute return remains positive, and the portfolio is well-supported by value, then you can hope for better times ahead. You should hope for mean reversion.

The hope in question is not idle, but the outcome is not guaranteed.

We none of us know the future, which is why we bet on statistical odds.

The odds are that bad times in markets are followed by good times.

Is this deliberate neglect? Not really. It is casual confidence.

I am confident that evident value will be bought.

If I think it has been bought by me, I do hope that others buy it.

However, it pays not to be reckless. Step into your evident hopes of recovery by stages and do not bet the farm by swinging for the bleachers. Husband those positions that are doing very well but be prepared to trim them.

The idea is to allow a three-to-five-year value cycle to play out for you.

This is not possible through frenetic trading.

The three-to-five-year cycle demands low turnover.

Specifically, if you intend to hold for three to five years, you should mark your ideal turnover budget as implying a 36-to-60-month turn, in and out.

Let us work that out explicitly:

36 Month Average Holding: 100%/36 = 2.78% turnover per month

60 Month Average Holding: 100%/60 =1.67% turnover per month

This is not very much and certainly does not mean you can dump a 3% minimum new holding size every month. This month, as seen below, we will be 3.46% turnover.

Turnover is a little higher than our budget to take advantage of value on offer.

For those investors not used to daily measurement, which is most, this can seem like an agonizingly slow response to a rapidly changing world.

Indeed, it is a purposefully slow, stubborn, and tortoise like discipline of creep.

The mean reversion of value is slow burning and demands patience.

You may wash your hair every day, but Rachel Hunter hair takes time to happen.

I look at it this way…

With a schtick like that, value investors can holiday in the sun with no phone.

Let us hope it turns out okay, with all the snappy moves of this here tortoise.

Understanding Benchmark Tracking Error

Leaving aside the schtick, the obvious question to ask is this:

Is 20% positive or negative alpha good or bad over two-and-a-half years?

Easy, says marketing:

20% positive alpha over two-and-a-half years is good.

20% negative alpha over two-and-a-half years is bad.

So true, and yet so unhelpful to any real investor.

Up is what we want, down is what we don’t want.

Yes, Mr Marketer, we do know that in portfolio management.

The problem, for any investor, is what to do when it goes up, or down.

Do I sell all of the losers if it goes down?

Do I buy all of the winners if it goes up?

The real problem is that much of what happens is statistical noise.

If you have a normally distributed random variable with a constant mean, and constant standard deviation, the estimation error of that mean, goes down as the inverse square root of the number of periods in the statistical sample.

Let us now look at the track record for this strategy.

When you invest in global markets, there is a wide variability of outcomes due to the sheer number and variety of stocks and markets involved. This shows up in a measure called the tracking error, or relative risk, of the strategy.

In the 30-month performance record above, the annual average alpha was 1.15% on a total risk of 13.06%, and a relative risk, measured versus the benchmark, of 10.79%.

Assuming the returns were generated from a stable distribution, which they are not, we would need to wait about 100 months to express confidence, that our generated alpha is positive (the statistical uncertainty is circa 10.79%/10, the square root of the number of trials, at 100). That is the one-standard deviation estimate.

For 30 months, the naively estimated uncertainty in alpha is 1.96%, and that at the level of one-standard deviation, or 68% for that standard confidence interval.

To put this in the proper context, standard statistical analysis would show that you need to wait roughly 900 periods, or 75 years, to have 99% confidence in your estimate of the alpha to 1% precision with a 10% relative risk.

Did I choose things that way just to mess with your head?

Not really, it is a feature of global financial markets. This benchmark is the FTSE Global All Cap Index, measured in USD, like our portfolio, which has 60-65% US stocks. This happens to be the benchmark we chose, because the Vanguard ETF AMEX: VT tracks the same benchmark, and can be bought very cheaply for 0.06% MER.

There are no US stocks in our International 20 Stock Model Portfolio.

We should expect this model to behave accordingly.

When the USA has a knockout quarter, we will underperform.

This is not an excuse but a caution. Whenever you are evaluating global markets be aware that the tracking errors can be very large indeed. This is why it is hard.

I mean hard for everybody not just hard for me.

For instance, the South Korea ETF ARCA: EWY is up 87.97% Year to Date (YTD).

You can go crazy, and many do, madly switching in global markets.

This is not what I do because I learned my lessons long ago. Whenever, the markets move wildly against you, in this global sack race, look carefully for the value.

The Characteristic Time Scales of Market Action

Every single day that the global market is open, there is money sloshing back and forth looking for a return. These flows will move markets all by themselves.

Therefore, we must be cautious, and a little circumspect, on attributing too much of a real signal to the back-and-forth of daily flows and the incessant chatter.

There are stories told the whole time.

Some are true, some not so true, some might prove true one day.

Others express the value of a good story to attract money.

In the short run, you can be confident that a good story will attract money, and the validation of that story is how much money it can attract right now.

This is why chasing momentum works on a three-to-twelve-month horizon.

I have not done that back test for years, but I warrant it is still true.

The opposite strategy works well over periods of about one month.

The so-called residual reversal strategy is the best quantitative signal known to man.

Generally speaking, you can bet on the losers of last month reversing.

The other timescale of statistical significance is the three-to-five years reversal.

Investor neglect sees many losers become winners over three to five years.

This is the source of most investment alpha for so-called value strategies.

Recognize that I have nominated three key time scales and a dominant modality.

• Mean reversion and reversal of returns over about one month.

• Continuation of price trends over three to twelve months.

• Mean reversion of a much greater scale over three to five years.

The earlier turnover mathematics highlighted that there is a cost to playing any of these sweet spots as your source of investment return.

The reversal strategy over one month is very high turnover, 1000%+ per year, and so only suited to those with low transaction costs, whose income is from trading.

These are the market makers, Commodity Trading Advisers (CTA), broker proprietary trading desks, and a good chunk of quantitative hedge funds trading signal.

The three-to-twelve-month trend play suits full bench fundamental funds managers who have sold their investment chops on their ability to forecast quarterly earnings.

This is a 100% to 400% turnover proposition that supports high volume marketing to garner assets because investors love to hear a timely update on market stories.

The final three-to-five-year strategy suits those with a high tolerance for short-term pain, the ability to defer the pleasure of long-term gains, and a taste for deferred capital gains as their primary engine of personal wealth creation.

This generally does not market well which is why proponents are characters.

People listen to Warren Buffet because of who he is, not what he does. They do come first for the returns, but they stay for the homespun philosophy, and the anecdotes.

The rational value investor dispenses entertainment to soothe current pain.

This is not silly, or fatuous, since any value investor worthy of that name will make many mistakes in their career, and experience an awful lot of personal pain.

This is the kind of pain that breeds a stoical cast of mind to misfortune.

You know the schtick:

That which does not kill us will only make us stronger.

Provided that one is pragmatic, suitably cautious, and willing to label spades as spades, when all other monikers have been deployed, you are good to go!

On to what we will do, this month, which is not very much.

Read more

West Texas Intermediate WTI crude oil closed at $64.92 USD on 30-Jun-2025.

It closed at $70.02 USD on 30-Jun-2026.

This is after a one-billion-barrel draw on global inventory and strategic reserves.

The Israel-USA-Iran war is supposedly over except for the occasional airstrikes.

Gold was a more positive story.

It opened the account on 30-Jun-2025 at $3303.09 USD/oz.

It closed the year on 30-Jun-2026 at $4007.26 USD/oz.

Normally, we would toast such a result!

The return of gold, the barbarous relic was +21.32%.

Too bad the S&P 500 returned +22.05%.

Our twenty-stock best ideas US portfolio returned +20.33%.

I guess our best ideas were not good enough this year, and that is the simple truth.

Among the better ideas out there was High-Bandwidth Memory (HMB) chips, now in high demand in data center build outs, such as are made by Micron XNAS: MU.

I am no fan of hair shirts, but a minimum 3% allocation to Micron would have helped.

For those who have been around markets a while, this sh*t happens!

Hindsight will beat the market every single time. We have to move on and deal with the road that lies ahead of us. This is not made any easier by the suspicion that all is not right in the world of energy. Something about this oil price looks fake.

Two of the most cyclical industries on the planet look to have distorted pricing.

Memory chips are the most cyclical sub-industry in semiconductors.

Oil is the most cyclical of growth sensitive commodities.

For what it is worth, here is my “insight” such as it is.

High prices are generally cured by high prices, that is to say lower demand.

Low prices are generally cured by low prices, that is to say higher demand.

The boom in AI spend is real, but the memory price trends are worrying.

Local outfit Pc Part Picker has tracked a four-fold increase in the price of DDR5 6000 RAM (2 x 16GB) from a stable $200 AUD/module to $800 AUD/module.

The result will be price rises across the board in memory using consumer electronics.

• Smart phones

• Cameras

• Laptops

• Servers

• Smart televisions

• Electric vehicles

The frenzy of artificial intelligence spending is leading to an inflationary spike.

The flip side of this is that some stocks we hold and like are really hurting.

The obvious solution to this problem is to buy memory makers.

This is not a terrible idea when price targets on stocks like Micron are raised to $2000.

For those who are happy to trade this wild market let us pray with Saint Augustine:

I emphasize the word trade, since you should rightly run a wide trailing stop about 20% lower than the current price to salvage any profit when it cracks.

That is easy advice to give and almost impossible advice to carry out.

Failing any saintly hotline to a higher authority, we will stick to the brief for this model portfolio and try and patience our way through this hot mess of wildness.

Afore doing that, here is the year in review.

FY2026 in Review

If you happen to live outside Australia, you have a different financial year. Last night I went and sank a few mulled wines at the Canberra Christmas in July, with a chaser of Berlin Banger - The Foot Long Kransky Chilli Dog. We are weird here. Get used to it!

Here is the annual roster of offenders and defenders.

I am flippant here, because this is a wash. You can see that there is no real skew either way. This was a middle of the road, track the benchmark year. Not good. Not bad.

Apologies for the tiny image, but it is the whole year compressed into one snapshot.

The best way to understand the story of last year is to look at quarterly alpha:

• Q3 2025: +4.71%

• Q4 2025: -1.39%

• Q1 2026: +4.13%

• Q2 2026: -9.43%

Notice that the relative return was really bad in the last quarter.

For completeness, the numbers were:

• Q2 2026 Portfolio: +5.52%

• Q2 2026 S&P 500: +14.95%

That last comparison is crucial for what we do or do not do next.

In a quarter where we lagged by 9.43%, we were actually up 5.52%, which is a better than 20% annualized rate of return. However, the market was up 14.95%, which is an annualized rate of return of +74.6% per annum! That is what happened.

Can this go on?

With the level of enthusiasm for all things artificial, including oil prices, yes, it can.

For how long, we cannot be sure, certainly another quarter or so.

During the late 1990s technology boom, I recall averaging portfolio rates of return of around 75% per annum over three straight years, until it stopped, and I didn’t.

The markets then dropped about 80%, and I got back to where I started.

This is what happens in an investment mania.

Personally, I do not think we are in full manic mode yet, so I expect more frantic chasing after whatever moves up and to the right.

The other lesson from the contribution analysis is to note that hyperscale stocks like Amazon XNAS: AMZN, Meta XNAS: META and Microsoft XNAS: MSFT undershot.

A boring railroad like Union Pacific XNYS: UNP did better than that bunch.

Notice also that, while Alphabet XNAS: GOOGL was our top contributor, we had a very good showing from diesel engine maker Cummins XNYSE: CMI and mega-cap gold miner Newmont Corporation XNYS: NEM. The last in spite of the retreat in gold.

Did we make mistakes last year?

You bet, the two I would single out was our exit of Apple XNAS: AAPL, and the brief dalliance with Oracle XNAS: ORCL where we timed everything badly.

You can surely find others.

We held no Micron.

Onwards to a minor rebalance.

Read more

(Addendum: See below for the final results).

After this rebalance report the full-year results are in.

The FY2026 returns as at 30-Jun-26 were:

Portfolio: +14.50%

S&P/ASX 200 Benchmark: +6.11%

Alpha: +8.39%

It will move around in the last two trading days, but it was a good year.

The bulk of the outperformance came from timely resources bet around this time last year. In view of changing conditions, we are closing up some of that in this note.

The since inception numbers have improved in recent years.

This was our best year yet in both absolute and relative returns.

Read more

I have had three careers and started five companies in my time.

The 2026 Budget changes to tax and negative gearing do not scare me.

I have seen far worse economic challenges in my 63 years on the planet.

What matters is knowing how it will affect the landscape.

Here is my take:

Directly, it removes investor competition for housing with owner occupiers. Directly, the abolition of the CGT discount shifts the horizon to long-term investing. Directly the realization that deferred capital gains are better will refocus minds. The idea of paying advisers to lose you money is dead.

Australia has had three decades of easy money credit boom, and it shows.

Now we can hear the public screams of those drowning naked.

Nobody forced any investor to negatively gear overvalued real estate.

The effect of removing the public subsidy of PAYG deductible debt for investors, and not first home buyers, is to take them out of the market. They now have to pay the real cost of finance for those properties. They realize it makes no sense.

Any competent financial advisor could have told them that.

Now they get to find out for themselves.

For those in Australia now bleating clock this:

In the USA, you cannot deduct the principal portion of your mortgage payments. However, you can deduct the mortgage interest on your primary or secondary residence. This is done by itemizing deductions on your federal income tax return rather than taking the standard deduction.

Like here, before the budget changes, you can deduct interest against rental income for an investment property. Where were the advocates for a policy of helping our home buyers by making the interest on their principal residence deductible?

Nowhere to be seen.

There are no Australian advocates for deductible first mortgages I can see.

Cue massive wailing and gnashing of teeth when the privileged negative gearing set got their punch bowl taken away. Already, property values are normalizing.

Falling real estate values in Australia are necessary to balance this market.

I own a residence too, and mine will fall in value also. Since I judged real estate to be overvalued in 2020, when I bought my present place, I live very modestly.

This correction has been three decades in the making.

It is here now.

My take on Entrepreneurship

If you believe the hype for the Venture Capital community, Entrepreneurship is all about twenty-something world changers with a bright idea to make a trillion.

If there had been a venture capital industry in Australia, in 1983, during the depths of the Volcker double-dip recession when interest rates were 18%, then 20-year-old me might have had a go. I could program, owned a computer, but was crap at designing video games. I was good at physics and liked to tinker. It was not crazy, at that time, for a young person to think of designing or inventing something.

For me, I had considered the scientific instruments industry, where there was quite the cluster of activity in Melbourne, with Varian and Vision Systems Laboratories.

I was fortunate to do my honors in the Research School of Physical Sciences at the Australian National University (ANU) in a laser lab where my supervisor got me to assemble and tune a laser wavemeter of his design. This was great stuff.

Once I had graduated, in 1984, I looked for jobs and was offered one in the laser lab at Telstra Research Laboratories (then Telecom), in Clayton, Melbourne. I was a very good student, and got the job, but turned it down. I wanted to do a PhD, and I had decided that it was best to go overseas to gain some international experience.

I duly won a scholarship to do a PhD.

As a condition of the scholarship, I could choose to go anywhere in the UK. Normally, it is competitive, but I got the scholarship, so any University in the UK would accept my application, as a condition of the British Council program.

I chose to go to the University of Bristol to do theoretical physics.

I am glad I did, as my advisor Prof. John Hannay was the quintessential eclectic and innovative Brit. He gave me a lot of confidence and taught me to think for myself.

I never had any trouble generating ideas, but John showed that the emphasis should always lie on following your nose on the most promising ideas, even if painful.

I contend that this is most important lesson of entrepreneurship.

You need to persist on those things that make a difference and junk all else.

Notice that this is not good advice for life in a run-of-the-mill job.

In a regular job, you do what you are asked even if it makes no difference.

I was always stubborn about doing anything that made no sense to me.

The PhD training taught me to be stubborn with myself.

Do, think, and try at what will make a difference if you solve the problem.

There I used the word “problem” which is common to R&D and entrepreneurship.

In my view, entrepreneurship is very simple:

Solve a real-world problem for your customers that delivers value to them.

Some folks throw in “change the world” but if that was my sole focus I would have stayed on doing research in theoretical physics.

If you want to seriously change the world advance theoretical physics.

You will have a much greater effect on future technology if you do that.

For me, entrepreneurship in business has always been more modest than any of the things that I have worked on in theoretical physics.

For what it is worth, I decided to switch from a life in academic research because I worked out that my kind of theoretical research needed almost no money.

Why battle for funding when you could get a real job?

Now that I am safely out of corporate life, I can happily report that no meeting was ever wasted time for me, no matter how boring. I can do calculus in my head.

With an attitude like that entrepreneurship is a natural for me.

Fake everything that does not matter until you solve the problem.

Fake nothing that does matter for your customer.

There is not much more to it than that.

Why is Australia weak in Technology?

Finally, for what it is worth, because it is highly relevant to the Australian economy of today, I am glad I did not take that job at the Telstra labs in Clayton. It lasted a few more years, and they shut it down to save money. Later Telstra bought out OTC. I visited their Sydney office around 1992, when they were being merged with Telstra to rationalize the business. They had a leading fiber optics lab. That got shut down right before the global boom in fiber optic communications. Later again, much later, I had the good fortune to interact with the Sydney arm of NICTA, a national facility which was leading Australian research in Artificial Intelligence. It was shut down in 2016.

For reference, 2016 was the breakthrough year for AI. That NICTA laboratory rated number five globally in the year it was shut down. Then it disappeared.

Who advocated for NICTA to get shut down?

The Australian Venture Capital community.

Enough said on this repeating theme of Australian R&D. Visionary folks set up the lab and make it world class. Short sighted bean counters shut them down.

Rinse and repeat ad infinitum.

Now the bean counters of this land want to shut down the Australian Synchrotron.

Hopefully, this facility will not die, and I can use it to test some ideas of mine.

If this facility does die, I will go to Japan.

Japan is not chock full of science and engineering imbeciles, like Australia is today.

The Ideal Business Opportunity

There is no general rule but there is a standard shape to attractive opportunity.

For any opportunity to be real, there must be some barrier to realize the payoff. Ideas that anybody can do will attract a lot of competition, which diminishes returns.

However, not every problem you may want to work on has a positive payoff.

Somebody has to be interested to pay you for the product or service and so you must deliver some value for them. Your individual fit for opportunity depends on you.

I am not you so I do not know where your best fit would be.

However, I can say that if it costs nothing, in time, money or thinking power to solve the problem you have in mind it is not likely to be a durable business opportunity.

You also have to contend with market preferences.

You may actually have an idea for a superior product or service that would genuinely offer the customer better value and a more positive experience.

More than once I have faced the problem of entrenched market preferences.

There has to be some way to persuade the customer to try and buy your product.

Consumerism is now so rampant that many poor products find a market.

This can be down to the simple power of advertising and marketing budgets.

I try to avoid anything like that because I think it is losing game.

Shoving aside the poorer contenders, in a crowded marketplace, is expensive.

Take a look at major consumer products markets in the age of social media. There are all kinds of indifferent products that are heavily marketed and widely bought.

Google and Meta have this business sewn up. They count the marketing dollars that flood into any segment with low barriers to entry, that demands marketing spend.

Returning to the option payoff diagram there are shallow and steep payoffs. The more certain the payoff the closer it is in time, but you may spend more dollars on it. The steeper the payoff curve, the more distant it is likely to be in time.

This may seem counterintuitive, but many of the best opportunities take longer to ramp but cost less in dollars. They cost more in the time you spend on them.

Human impatience is your best friend in delivering fat opportunity.

These types of problems involve solving the more complex problems which are more difficult for people to tackle because they may involve more than one puzzle piece.

The favorite problem type I work on has several puzzle pieces, where each must fit together, and each sub-problem looks impossible to most people.

I love to work on impossible problems.

This is not as foolish as it sounds, as most impossible problems are possible problems that require some patient thinking to address while the world spins about you.

Venture Capital will not touch people like me which is great for me.

If you identify possible impossible problems early you are likely to win.

Even before I had entered the business world, I had solved one problem that was widely considered impossible. This was an invaluable experience for me.

The question of entrepreneurship comes down to your choice of problem.

If you note carefully the trajectory of one Elon Musk, you will note that he has no fear of impossible problems. Mostly, he does not solve them, he solves another one.

However, you have to cut him the slack that he might solve a really big problem.

This is why I put two option payoff profiles on the diagram.

The Ideal Business Opportunity lets you work on two problems at once.

The first problem is to get a business running, the second problem is to scale.

You might spend a decent amount of cashflow to get into a business, and smaller amounts of money, mostly thinking, on figuring out how to scale it.

That is as generic as I can be without talking about a specific opportunity.

The Problem of Getting into Business

That is all very well, but how can one get into business?

If you are in business, you already know how, and you probably had no need of any abstract picture like the one I described. You just opened your doors.

The vast majority of professional services businesses are like this. You develop some marketable skills, get to know prospective clients, and put up a shingle.

There you are in business!

The same is true of trades, and perhaps more so.

Once you have your ticket for a trade you can be in business.

The problem only arises when there is not a direct line between your best skill set and the business opportunity you seek to harvest. This seems like a non-problem to many who have not faced the challenges of supporting infrastructure for your business.

Let me give a concrete example from my own career.

I left physics research in the mid-1990s as there was a dearth of funding. I realized that I could do my research with a pencil, but I needed to eat, and sleep some.

Like many of that era I went into finance, where I could make good money. The work was easy for me, and I was very well paid. There were lots of boring meetings, but I can smile, talk, and do calculus in my head, without blinking, or raising a sweat.

This was perfect for me until I decided I wanted my own finance business.

Since my specialty was quantitative finance, I needed lots of data, to do my job, which can be expensive in money to buy, and time to clean up and analyze.

I am not bleating; this is a standard problem for quantitative finance people.

When you have this type of business opportunity in mind, you will have to plan ahead and put in place the pieces you need, in a legal way, and form the capital stock.

The problem of forming the capital stock is a key one in entrepreneurship.

It is easy to be superficial about this challenge, which constantly morphs and changes.

Windows of opportunity arise where the capital stock requirement is lower.

One might also find that financial capital can readily substitute for capital stock.

You can raise money from investors and buy whatever you need when you need it.

This is the standard script for Venture Finance. Go in with a great pitch, come out of the meeting with a commitment to fund your venture with other people’s money.

There is nothing wrong with this approach, provided you have the right idea.

For students of Venture Capital, the key thing to understand is that the right idea is most likely a fashionable idea into which a lot of VC money is now flowing.

The way to win that game is to work on yourself as the entrepreneur.

This is the “sell me a pencil” test. If you work on yourself, you can likely sell the pitch that you are the one who can take a fashionable idea and make a big business.

This works for the right kind of person and can be very lucrative.

If you are anything like me, you probably do not find the idea attractive.

It is not that it is a bad idea; it is just not your idea.

This is an okay way to be also but requires a different approach.

If you are not going to raise a lot of money from investors in the business, you might find the equivalent kicker from clients. This works very well for professionals.

The best form of finance for any business is paying customers.

If you have developed a good reputation in your line of work, you can well start a business in that line of work by getting your current customers to follow you.

This is the best approach I know in financial services businesses.

If you really want to succeed, in the most reliable way possible, within the financial services industry, then hone your skills at a large, respected firm, and leave.

This is what I did with my financial services firm Jevons Global.

It is the sensible low risk way to get into a professional services business. This path is not a sure-fire win but is a proven and reliable way forward for most people.

The best advice I have come across in the mid-career entrepreneurship area is that from the Kauffman Foundation, in Kansas City, KS USA.

Since I went to school in Kansas City, in the late 1960s, I have a soft spot for KCK.

Ewing Marion Kauffman was a successful mid-West entrepreneur who gave back to many humanitarian causes, including the demystification of business formation.

The foundation he established has done much to demythologize business formation by pointing out that it is mostly mid-career folks who form them. Provided those are in areas they understand, they are more likely to succeed, due to experience.

The Valley myth of the twenty-something entrepreneur works for them.

If you analyze it carefully, the real business is fifty and sixty something VC partners, presented with a conga line of hopefuls, just like the Hollywood studio business.

It is not that it cannot work for you, but is unlikely to work for most.

If you like the mythology, then work on yourself to be pitch perfect.

It is an acting game and should be treated as such.

For guidance on real engineering entrepreneurship go to Shenzhen.

Failing that, figure out a way to build capital, to start your own show.

The Capital Accumulation Circuit

I am a mathematician, so my view of capitalism is very simple.

You go from nothing to an idea, which is an infinite rate of return. There are losses when you convert that idea to cash, but the goal is to rinse and repeat the cycle.

What matters most is your conversion ratio on capital employed.

Long ago, some idiot named Benjamin Franklin, decided that mathematics was way too valuable to be protected by copyright or patents.

Ever since Ben Franklin, all mathematicians have worked for a living.

This is not all bad, some of those very boring meetings in finance come with drinks!

Seriously though, it is your ability to convert starting capital to ending capital that matters most for successful business, and you must work on the conversion ratio.

Note that capitalism, as I have described it, does involve work. We are not talking about extracting rents from passive activity. That is a very big area of activity.

In Australia, there very few genuine capitalists, mostly they are rentiers.

You can tell this very easily by looking at the productivity data.

US Productivity Growth has been pretty reasonable over time. You can argue about who shared in the proceeds of that growth, but the facts are real. It is not simply a taxation or labor cost story. It is mostly a capital stock story.

The capital stock is that magic piece that completes other factors of production like land, labor, energy and other material commodities. Done right, it adds value.

Australia is a very sorry tale by comparison.

There are some who blame this on the labor market. This has something to do with it, but you will not materially lift labor productivity without matching capital stock.

It is true that the Australian capital expenditure data is skewed heavily by mining, but the bulk of labor is employed outside of the minerals industry. Why such low rates of capital formation and why is anybody surprised that we have low productivity?

Here is part of the answer, rising land prices.

In Australian dollar terms, gold prices started rising significantly after the huge money injection to global markets since the Global Financial Crisis (GFC).

Australia has had a huge credit boom in the last twenty-five years. Not much of that went into new capital stock, outside of mining, land values and real estate absorbed the early flood of newly printed money from too easy financing.

Now rents are soaring and wages will follow.

Australia has been home to the mother of all rentier booms for passive geared real estate speculation. It would have ended, regardless of the tax changes.

Once all the blood has been sucked via rent extraction the economy rolls over.

I do not envy Jim Chalmers job, because this economy was always going to croak no matter what he did, however I do support what he has done before the final croak.

Australia is headed for the mother of all portfolio realignments.

It has started now, and the pace will soon quicken.

The beauty of the Chalmers move is that it came before the teeth of the downturn. For those who understand markets, they have time to adjust and a clear signal.

This is superlative proactive policy from a forward thinking government.

I know many will disagree with that statement, but you are not capitalists.

Capitalists understand there are cycles in business and prepare for them.

Let me finish by reprising what I think is a good path for aspirants.

Target Financial Independence

There is the easy form of such advice.

Get rich quick and retire.

If you can do that, why not?

For most people, this is not helpful advice. They have to work for a living, provision for children, make a home, care for other family members, and save for retirement.

For many, the removal of the 50% CGT discount came as a shock because they are on the top marginal tax rate and feel that capital accumulation potential is limited by it.

I will not lie; the removal of the CGT discount hurts.

However, as I pointed out in The Value of Tax-Deferred Capital Gains, the impact of the tax on after-tax rates of return is mitigated by extending the holding period.

The short version is that when you hold longer there are many PAYG tax years in which there will be no impact on your taxable income from a capital growth strategy. You do pay tax when you sell, but the holding period rate of return is higher the longer that you are able to defer taxable gains. You need no special structure to do this.

The need for tax advice is much lower when you pursue long-term investment.

There are many who benefit from the annual charade of finding new ways to lose some money ahead of the 30-Jun deadline to reduce your tax bill.

Do not sell your winners and there are no losses to find to offset gains.

Some contend that active portfolio management is dead post these changes. This would be true if you had no idea how to run a low turnover investment strategy.

I bought my current Microsoft holdings in 2016 and have been trimming them into strength for about three years now. Those measured gains are offset by any losses that I choose to realize because I invested badly or found a better idea.

This idea that you can only become successful if wages for workers are zero, taxes are nil, your utility is a charity, land is free, and lawyers are slaves locked in a gimp cabinet in the basement, is spurious. If you think that, never try to form a business.

What does affect all business is too much rentier activity.

When the pricing of factors of production is set in favor of passive rent seeking then anybody and everybody who is an entrepreneur will struggle.

Entrepreneurs struggle when the magnitude of rent-seeking is too high.

This is why I welcome the Chalmer’s tax changes. They just slit the throat of many a shameless rent seeker and the gurgles you hear are the death rattle for them.

This does not alter or diminish your problem of entering business.

I have thought for a few weeks, and this is the best idea I have come up with.

Some years ago, the government created tiered tax rates for small businesses:

• You have to be less than $50M AUD turnover

• You have to have less than 80% of passive income (rents, dividends, capital gains)

If you meet those simple criteria, your tax rate, in the business is 25%.

The business sector never had the benefit of the 50% CGT discount. PAYG taxpayers on 48.5% had an effective tax rate of 24.25%. This meant it was better to run your capital account through you PAYG tax and take advantage of the discount.

Now you are being actively encouraged to form a company.

This is not a rort because you need to have at least 20% operating income.

For most young people starting out this is not an onerous hurdle.

Do not even think of starting a company unless you can afford the $5K outgoings a year to operate that at any significant level. more if you need licensing.

The challenge for most young people is to form capital, which means to attain the level of financial independence that enables business activity.

The first hurdle is somewhere to live, eat and sleep.

Due to high housing costs, Australia is a very mean society to the would-be business entrepreneur. You are forced to be a wage slave to afford a roof over your head.

You know already that mathematicians are not worth anything.

This idiot told the world to never let any mathematician profit from their work.

See now, I can walk among you and not wear rags!

I am a mathematician who found out how to make an evil system work for me!

You can too. I am certain of it.

The trick with this small company caper is that you have to find some way to make at least 20% of what you make on your investments from real business activity.

In simple terms:

Make $100K operating revenue for each $400K of investment income.

If you can manage that you qualify for a 25% compounding capital circuit.

Of course, you have to think of a business idea to generate that operating income.

That is your lookout.

I am the mathematician; I prove the theorems.

It is up to you to figure out how to make money from that.

In my case, I already have a good idea.

It is a total winner!

Have fun.

Subscribe now

Only two things in life are certain: death and taxes.

I cannot defeat either, but I can describe the optimal strategy for investment, to achieve a higher after-tax rate of return, via growth or income.

What I am about to explain is standard in finance.

Any qualified financial advisor, anywhere on the planet, will agree with me that the deferral of tax to achieve long-term capital growth is the optimal investment strategy. It is elementary compounding mathematics.

Of course, you need to make assumptions to prove this, but they are simple.

Assume the tax rate on annual income is the same as that on capital gains.

With these assumptions, I will show you how to prove this result.

It is both simple and sound.

I will offer the most important piece of general financial advice you will hear.

If your financial advisor says otherwise, show them this note.

Ask them a simple question:

Please explain?

If they still do not understand, send them my way, and I will explain it to them.

That is all there is to it.

Taxation on Income versus Growth

To simplify analysis, consider this investment scenario.

You have two investment options, each offering the same pre-tax return of R% per annum and your marginal tax rate is T, applied equally to both.

Which should you choose?

Naively, you would be indifferent.

After you have paid tax at rate T, the post-tax return is R (1-T).

How could it matter which option you choose?

The post-tax annual rate of return looks the same, doesn’t it?

Well, in the real world, the after-tax rates of return are NOT the same.

This is not a social media narrative complete with viral video.

It is simply a mathematical fact that is true for all R >0 and Tax rates in the full range of T = 0 to T < 100%. If T >= 100% then you should move your tax residency.

The reason may be traced to compounding when you defer capital gains.

The Option to Defer Capital Gains Tax

Those who pay attention to the mathematics will do better in the long run.

The full Einstein quote reads:

Income and capital gains are not treated the same by the taxation office.

You pay tax on income in the year you earn it.

You pay tax on capital gains in the year you choose to realize them.

This looks like a minor difference to those who do not grasp compounding.

However, it is a massive difference in practice.

The entire period you amass unrealized capital gains you have a zero-interest tax free loan from the government to finance your future tax liability.

It gets even better with inflation indexed capital gains.

Under inflation indexing your future capital gains tax liability is financed by the tax office at an effective nominal interest rate equal to minus the CPI.

That is the mathematical reality of tax defered capital gains with indexation.

Note that I have been comparatively silent on the impact of the recent CGT changes. The rest of the industry ran around like chooks in a barnyard.

Once you read on, you will see why I did not join this foolish public melee.

Those who did can read this and make their excuses to their clients.

Their clients pay them for good advice.

If they got bad advice they may want their money back.

The After-Tax Return of Income versus Growth

What I am about to show you is a standard result in every financial advice curriculum.

If your financial adviser does not know this result you should ask them why.

Normally the result I will show you is stated in words without proof.

I am a mathematician so I will go one step further and prove the result.

The result in question is this:

For any rate of return R >0 and tax rate T that is in the full range of 0% to 100%, the after-tax return of growth investment is better than or equal to that of income over any period.

The jargon used by the industry is to say that growth dominates income.

Notice a few caveats of no great importance.

Income investments can come with tax credits, like franked dividends. This means that you should compare rates of return pre-tax with credits added back.

Australian investors know the rule for our tax system.

If you receive dividends franked at 100%, the gross-up for the tax credit is obtained by adding back the franking credit amount. This means your franked dividend yield of Y% has a tax-credit driven gross-up of 1/(1-t) where t is the company tax rate.

For large companies in Australia, that rate t = 30%, so the gross-up factor is 1.4286 which is 1/(1- 0.3). In effect, a 5% fully franked yield of 5% is adjusted to 7.14%.

With this caveat in mind, we will set R = 7.14% for comparison purposes.

In practice, this complicates decisions slightly, because many growth investments can also spin off some income. The tools I describe here can cater for that case.

This is free general advice so I will not do the warts and all calculation.

I would need to analyze the underlying investment to declare which is best.

With those caveats out of the way, let us do the calculation with no indexing of capital gains to properly emphasize the effects of compounding.

Indexing of the investment cost basis by any CPI>0 improves the result.

For clarity, it will also help to state a standard result:

After Tax Proceeds = (Pre Tax Proceeds) x (1 -Tax Rate)

This is clear, the money you wind up with is that with tax liability deducted.

That is the basis of the comparison we will run over N whole years.

I choose whole year tax periods to keep it simple.

You invest on 30th June of year Zero and realize gains on 30th June of year N.

The income investment pays an annual income, the tax due on that is paid, and the rest is assumed to be reinvested at the same comparison rate R% per annum.

This scenario is idealized, but not in a way that defeats the principle.

I want to make the tax deferral principle as clear as I can.

The Income versus Growth Sack Race

Let us now run the sack race.

Cue wild and whacky music…

After-Tax N-period Return on Income

Off goes Drooper the Lion, out of the blocks with the pure income strategy.

At the end of every tax year, the after-tax rate of return in that year is.

If you compound that over N full periods you get the post-tax gain:

You multiply your starting capital by that uplift factor.

Let us do a simple example:

• Starting Capital = $100

• Pre-Tax Gross Rate of Return = 10%

• Marginal Tax Rate = 48.5%

We run the race for 10 years and get this result:

You can see that the after-tax analysis is pretty simple for income.

You are compounding at the rate R(1-T) which is 51.5% of the headline rate.

For 10% RoR and 48.5% Tax you are compounding at 5.15% per annum.

After-Tax N-period Return on Growth

Off goes Bingo the Gorilla, out of the blocks with the pure growth strategy.

This is pretty simple to calculate, but we need to recognize that no capital gains tax is paid until the investment is sold at the end of N years, and then only on the gains.

The formula is easiest to understand if you first compute the gain factor pre-tax.

You do not pay tax on all of this capital, only the gain over N years. This is the above result minus one, the starting capital. You apply tax at the rate T and keep (1-T).

With a little notepad scribble you will get this result for the post-tax gain factor.

The above form is the most useful for later, but a little confusing to interpret. It looks like the effect of tax was to reduce your starting capital. This form is the same:

This makes it clearer what is going on. You compound starting capital at annual rate R for N years, subtract the starting capital to get the gain, and keep (1-T) of that.

We will use the first form later to prove dominance.

Let us repeat the former example.

• Starting Capital = $100

• Pre-Tax Gross Rate of Return = 10%

• Marginal Tax Rate = 48.5%

We run the race for 10 years and get this result:

Okay, so the growth strategy won in this example over income.

Note that more tax was paid to the government in the growth strategy:

Income: total tax paid $61.43

Growth: total tax paid $77.30

If your objective was to pay less tax to the government, you would choose income.

If your objective was to be personally richer, you would choose growth.

If you accept that the purpose of progressive income tax policy is to make both you and the government richer, at the same time, you will both choose growth.

If you sell tax minimization advice you will tell clients to choose income.

If you want to be personally poorer after N years you will take that advice.

I am a portfolio manager of a wealth management business.

I take my own advice and choose growth over income.

Call me self-interested, but I am richer for it.

What about high growth investments?

I read the press like anybody else. Following the CGT changes I saw a lot of news reports by unlicensed financial journalists claiming that the tax policy changes would harm investors and the Australian economy by preferencing income.

I am a licensed financial advisor: I would go to jail for dispensing that advice.

The problem with this advice is that it is materially wrong and misleading.

If you tell your clients that you are dispensing unreasonable general advice.

It is objectively and demonstrably a false and misleading claim.

Any financial advisor, licensed by ASIC, has a duty of care to their clients to present any general financial advice on a best effort basis with a reasonable basis.

The term “reasonable basis” means well-argued and factually grounded advice. This is the professional standard to which we hold advisors.

If you do otherwise, namely allow your political persuasion to cloud your professional judgement to form and propagate unreasonable hearsay, that is a financial crime.

If you continue to dispense misleading advice with no firm and reasonable basis for that advice, then you are likely to receive the attention of our financial regulator.

The purpose of financial advice is to serve the client interest.

With that sermon of my chest, let us look at the high-growth investment case.

The Case R =20% T=48.5% N =10 years

Let us explore a higher growth example with R = 20%

Your professional tax advisor will point out that you saved tax choosing income.

Your professional wealth advisor will point out that you were richer choosing growth.

The Case R =40% T=48.5% N =10 years

Let us explore a higher growth example with R = 40%

Your professional tax advisor will point out that you saved tax choosing income.

Your professional wealth advisor will point out that you were richer choosing growth.

The Case R =100% T=48.5% N =10 years

Let us explore a higher growth example with R = 100%

Your professional tax advisor will point out that you saved tax choosing income.

Your professional wealth advisor will point out that you were richer choosing growth.

The Pattern and the Reason for that Pattern

Those who prefer to look at examples to infer patterns will have noticed this one:

Let me state this clearly in words so that there can be zero doubt:

The share of the pre-tax rate of return left in the hands of the investor after paying capital gains tax increases with the rate of return.

Contrary to widespread assertions in the popular press, the rate of return, after tax, increases with the rate of return, on a like-for-like comparison basis.

Comparatively speaking, you get a greater share of the returns on offer if you defer tax by seeking long-term capital growth and you seek higher rates of return.

That is the reality not the sensational headline.

Journalists do not risk regulatory action against any misleading statements. Licensed wealth advisors, like me, are obligated to be truthful.

Some journalists have claimed that these measures are harmful to start-ups.

Evidently, an investor will make more money if the tax rate is lower. However, they would also be wise to consider society to be their fellow traveler. Taxation funds government activity of all kinds, including pro-growth economic policy.

Wise entrepreneurs will consider the following pattern of post-tax rate of return:

The pre-tax rate of return assumption of 15% used to be the benchmark for Venture Capital. This is a four-fold increase of pre-tax capital over a ten-year period.

I will use that assumption as it has stood the test of time.

If you think I am unambitious, choose a higher rate and get a better result.

I have run scenarios at a top marginal tax rate of 48.5%, a pre-tax rate of return of 15%, and several holding periods stretching out to forty years.

Notice that the winning strategy is to hold investments for the long haul.

I am a professional wealth advisor, and I am licensed to dispense general advice.

Here is my general financial advice to anybody who pays attention:

Cut your losses when it is tax advantageous to do so and let your winners run.

I know that sounds like motherhood, but it is the truth of investment compounding.

Deferred tax payments on unrealized compounding gains are the winner.

This is so whenever you face a choice between equal prospective return. The caveat to that statement is comparable pre-tax return with all tax credits impounded.

The Choice of Holding Structure Matters

I am not going to enter the field of personal and corporate structuring nor any advice on personal tax affairs. However, I can make one strong statement.

If you choose a structure where the compounding tax rate is lower you are better off. Corporate holding structures often have lower tax rates.

The standard example is the PAYG top marginal rate of 48.5% versus the 30% tax rate on a standard large company corporate tax rates for holding companies.

If you are young, and seek an entrepreneurial career, with numerous possible projects and exits it would be wise to consider holding company structures.

You can see the effect of this by comparing Australian PAYG tax rates at 48.5% versus the standard corporate tax rate of 30%. Holding a project in a company is better.

There are many who might take issue with this statement because money that is held in a company is not the same as money in your bank account. You would have to pay tax on money that you distributed from the company. That is true, but dividends in Australia have attached franking credits for any tax paid. You distribute that.

This is general advice only.

If you make a billion dollars, consider if you need that in your bank account.

Your personal holding company can buy assets and buy you lunch.

When doing your fringe-benefits tax consult a licensed tax adviser.

Australians love trusts and not holding companies.

I do not know why.

Why Trusts may not be Ideal for Entrepreneurs

I defer to the CoPilot AI on why it thinks trusts are sub-optimal.

Trusts are bad for:

• raising external capital

• issuing equity

• retaining earnings

• scaling

• employee ownership

• governance clarity

• reinvestment discipline

They are not the structure used by:

• US startups

• European SMEs

• Asian family conglomerates

• Venture‑backed firms anywhere

Why? Because a trust is fundamentally a distribution vehicle, not a growth vehicle.

If you are a tax adviser who is paid to put your clients into trusts and you do not like the above statement then sue Microsoft, not me. CoPilot said that.

I am a mathematician and a licensed wealth advisor.

Tax deferral on capital gains is valuable.

Growth vs. Income Dominance

One of the advantages of being a mathematician is my holding a limited license for absolute truth. If the question hinges on pure mathematics, I can be absolute.

You will have spied the pattern, now let me make an absolute statement:

For the same (comparable) pre-tax rate of return R, invested over the same period of N years and subject to the same rate of tax T, on income or gains, the tax-defered investment strategy dominates the income-oriented strategy.

Obviously, the tax rate T has to be between 0% and 100%.

You also need a positive rate of return, or investing makes no sense!

If your tax rate T is greater than 100%, I suggest that you emigrate. If all prospective investments have a return R<0, then work for government.

Otherwise, you are good to go on your investment career!

Let us prove the above statement, using my favorite, Jensen’s inequality. This is a very useful mathematical result that hinges on a simple idea. For any curve that is convex, meaning it uniformly bows downward, like that below, you can state that points on the straight line between the curve values at two points, are above that curve.

Specifically, the inequality means this:

The funny upside-down “A” is lazy mathematician speak “for all” and the [0,1] bit means the interval between 0 and 1, which would be 0% and 100% for tax rates.

You have to have a so-called “convex” function f, or it makes no sense, and is not a true statement. The function which is x to a positive power N>0 is convex.

Recall our post-tax growth rate of return:

and our post-tax income rate of return.

Not everybody likes mathematics, but I am not asking you to do it! If you concentrate on the visual pattern, you may spot the trick. Mr Jensen, through his inequality, tells us that the bigger quantity is always the left-hand side pattern, with the t’s outside f, and the smaller one has the t’s inside. The right side is dominated by the left side.

Choose this substitution into the Jensen inequality:

Voila! You did the math! Gorilla stamp. QED.

As long as you state clearly your caveats, ASIC cannot sue you.

What you have just seen for yourself is well-known in the finance industry.

Lots of quality superannuation funds in Australia will say this in words:

For equal pre-tax rate of return R, and tax rate T, the capital growth strategy of tax-deferral will dominate the returns generated by an income strategy.

This is not true of the financial press in Australia.

That is fine.

They are not quality wealth advisors.

Conclusion

To rap on the mad-cap theme of our Banana Splits financial media the race is called:

• Bingo (The Gorilla) – 1st Place - pure capital growth with no dividend

• Fleegle (The Beagle) – 2nd Place - capital growth with dividends reinvested

• Snorky (The Elephant) – 3rd Place - tax-deferred income via discount bonds

• Drooper (The Lion) – 4th Place (Last) - pure income with no tax deferred

Notice that is for the exact same pre-tax rate of return R, and tax rate T. There can be differences in tax credits, and even tax rates, so you need to consider that.

Large sections of the Australian finance industry are compensated for advice on tax minimization and the accumulation of debt to reduce personal tax obligations.

They may have different ideas.

I am a licensed wealth advisor who is focused on the best after-tax rate of return.

You may pay more tax, over time, to our government, choosing growth, but you will also be richer. Greater post-tax returns are my area of focus.

That is the general advice on offer today.

Oh, and here is the spreadsheet I used for those case studies.

Tax Defferal

84KB ∙ XLSX file

Download

Download

Happy investing!

Subscribe now