Everything
Mean Reverts.
Everything.
Stocks, property, bonds — they all suck right now. That's not the end of the story. It's the beginning of the opportunity.
for mean reversion
from a 20% loss
to make $1
portfolio archetype
Mean reversion is the iron law of markets
No market rises in a straight line forever. Everything — US valuations, emerging markets relative to the US, value versus growth, commodities versus stocks — stretches until it snaps back. Your job as an investor isn't to predict when. It's to understand the mechanism deeply enough that you don't panic when the waiting gets uncomfortable.
This week: what mean reversion actually is, why the volatility gremlin makes losses more dangerous than gains are helpful, a quick framework from Nassim Taleb on building a portfolio that benefits from volatility rather than being destroyed by it, and why the AI boom has a maths problem that nobody wants to talk about.
The principle that
makes patience pay
Most investors understand mean reversion intellectually. Almost none of them can sit through it emotionally. The moment "nothing is happening," impatience sets in — and that's exactly when the principle is doing its work.
Mean reversion is why we at TMM focus on the principles and characteristics of investing rather than short-term calls. These principles work over time, not at all times. Understanding that distinction is the difference between skill and luck. It is the difference between now and the future.
The volatility gremlin
Most investors focus on making money. The best investors focus on not losing money. This is not a cliché — it is mathematics. Losses are asymmetric. A 20% loss requires a 25% gain just to break even. A 50% loss requires a 100% gain to recover. Each drawdown you avoid dramatically improves your compound return.
This is especially important when investors don't rebalance in overvalued markets. Focus on not losing money and you will generate higher returns in the future. It is more than a Buffett quote — it is the fundamental geometry of compounding.
What could mean revert next?
Seven themes are stretched relative to historical norms. Not all will revert this year. Some may take a decade. But the direction is clear — and knowing which way the rubber band is pulled tells you where to be patient and where to be cautious.
Fragile, Robust,
Antifragile
Nassim Taleb gives us a framework that cuts through the noise. Every investment — every company, every sector, every portfolio — belongs to one of three categories based on how it responds to volatility and shock. Most investors only think in two categories: up or down. Taleb adds a third dimension: what happens to the structure under stress?
Hates volatility. One large negative event is fatal — like a coffee cup dropped on a stone floor. Works fine in calm conditions, breaks irreparably under stress.
Absorbs shocks and bounces back. Like a candle — melt it down and you can recast it. Not destroyed by volatility, just temporarily disrupted.
Gets stronger from volatility. Like weightlifting — the stress is the stimulus. Panic in markets creates buying opportunities for the patient.
Mediocristan vs Extremistan
Taleb distinguishes two worlds based on how events distribute. Mediocristan is the world of heights and weights — adding one outlier (Shaquille O'Neill) barely moves the average. Extremistan is the world of earthquakes and financial markets — thousands of small inconsequential events, and then a single 6-sigma event that reshapes everything.
Here's the critical insight for portfolio construction: in the short term, markets can look like Extremistan (COVID March 2020, GFC). But over a long time horizon, markets belong to Mediocristan. A single company, however, belongs to Extremistan — it can go to zero. An ETF cannot.
Think about your personal situation. If retirement is close, you are effectively in Extremistan — one large drawdown at the wrong moment is hard to recover from. If you are young, you are in Mediocristan and short-term volatility works in your favour. The answer is different for everyone, and it depends on whether your portfolio is individual stocks or ETFs — and which sectors.
Building an Antifragile Portfolio
This is a preview. In the full masterclass, we go deeper on path dependence, ergodicity, the risk hierarchy, Kelly criterion allocation, and how to construct a long/short ETF structure that genuinely benefits from volatility — not just survives it.
- Path dependence and why the sequence of returns matters as much as the returns themselves
- Ergodicity: why the average investor's path and the market's average path are not the same thing
- The risk hierarchy: how fragile/robust/antifragile maps to individual stocks, bonds, and cash
- Long/short ETF construction as a tactical and strategic antifragile instrument
- CAPE-based valuation (f(x) = returns as a function of exposure to valuation)
The AI narrative
has a maths problem
The AI boom has been running for a few years. As companies like OpenAI move toward public markets, the problems are starting to add up. This is not a call that AI is useless. It is a call that investors should not succumb to narratives that the economics do not support.
- AI still hallucinates — the fundamental reliability problem remains unsolved
- The AGI narrative is overblown. Many experts say it is not possible with current LLM architectures
- AI is prone to sycophancy — telling people what they want to hear, not what is true
- Many companies are not seeing proportional productivity gains. Uber's COO said he's not seeing returns from increasing AI costs — and burned through their annual AI budget by April
- Data centres are deeply unpopular. 70% of both Republicans and Democrats oppose them
- Data centres consume enormous water and energy, often taken from local communities without consent
- The economics are structurally broken: $3 spent to generate $1 in revenue, before overhead
- OpenAI and Anthropic alone represent 40-55% of Microsoft, Google, Amazon and Oracle's entire revenue backlogs — circular financing disguised as growth
- Valuations are disconnected from the underlying economics by any conventional measure
This does not mean AI is a fraud. It means we are probably a long way from where the technology gurus believe we already are. It's the old Buffett framework: voting machine versus weighing machine. Right now, markets are voting. The weighing will come. In the end it is always about profitability and competition — and both are problems AI has not yet solved at scale.
What makes a business
worth owning forever
The characteristics of a worthy business
It's all about the ability to compound capital on a continuous basis. These companies are exceedingly rare. The combination of competitive moat, superior management, and the ability to grow dividends and capital over time — without relying on tax rates or macro conditions — is what separates a great business from a good one.
Remember: most market returns come from a handful of stocks. If you do not hold those stocks, you will underperform the index. This is why, for most investors, a range of ETFs allocated using the Kelly criterion is the most reliable path to superior returns.
All great investors suffer periods of underperformance. Even Buffett. The 3 Wells structure exists precisely to get around this problem — diversifying across strategies so that when one is in a trough, the others are working. The skill is recognising that the trough is temporary.