There are a few insightful questions we are often asked by clients on the persistence of active management alpha: 1) what structurally allows active managers to outperform? 2) If you are right, who is on the other side? 3) Which sources of Alpha meet the Bezos test? (no one actually ever asks about Bezos, but we like the mental model he provides).

We have previously discussed the structural differences between active large and smalls caps – predominantly driven by the different size of opportunity sets, the number of investors competing for ideas, and mature and stable vs emerging and dynamic industries amongst other drivers. What we focus on this month are some of the behavioural drivers as to why alpha exists.

Even when you know the odds are against you, your behavioural bias is a strong force to overcome. Gamblers know the odds are stacked against them, but still they gamble.

Market studies have shown that companies with the highest forecast revenue growth rates (either consensus forecasts or company guidance), otherwise known as glamour stocks, consistently underperform the market over the long term. By a lot.

One of the reasons is the compelling narrative ultra-high growth companies possess – which often means high prices relative to fundamentals. The other is that persistence in these growth rates is negligible – the historic evidence showing the very thing you are betting continues has a track record of not doing that.

“…sales growth rates have historically shown limited persistence and therefore a strong pattern of regression toward the mean. We found the correlations (r) for sales growth rates to be 0.16 over three-year periods and 0.14 over five-year periods for U.S. public companies from 1950 to 2025.”  – Competitive Advantage Period, The Neglected Value Driver. Michael J. Mauboussin & Dan Callahan

Even knowing this is true – like knowing when gambling you are likely to lose – doesn’t stop people from repeatedly making this bet. Human nature remains undefeated.

Sitting here today, we know from talking to investors and watching the market narrative around artificial intelligence evolve, that were many of these private companies public, they would a) fit squarely in the “highest forecast revenue growth” bucket and b) likely be irresistible to investors not wanting to miss out on “the greatest technology innovation since the Internet, if not ever”.

This is not theoretical. There is a closed-end fund in the US called Sunrise Innovation Fund (VCX US) which holds investments in some of the hottest private companies in the world. Think Anthropic, OpenAI, SpaceX, Databricks, Anduril. The stated NAV of this fund at the end of February was US$18.26 per share (per Bloomberg). Since that time some of the largest holdings have raised money at much higher valuations such as Anthropic (US$800bn up from US$380bn), OpenAI (US$852bn up from US$730bn) and SpaceX (est US$1.75tr IPO price). If we marked-to-market VCX US we would get a NAV today of closer to US$28 per share. If you think that is capturing peak values for these hottest private companies, then you may want to consider that the fund itself trades at US$95 per share – a 340% premium to this marked-to-market NAV.

This is exactly the sort of glamour investment unlikely to pay off in the long run, yet still putting capital at risk is exactly the human behavioural bias at play.

Outside of behaviourally biased investors, your counterparty could be an index tracking investor having to sell a deleted stock – although be careful here, the presence of very sophisticated of index-arbitrage funds mean this does not mean there is a free lunch.

Or it could be an investor with a different time horizon. Long term investors often accept they will never pick the exact top or bottom on their entry or exit for a 5 year investment. They may be buying from a short selling investor with a 6 week time horizon who is looking to capture the 10% fall they think will occur when the next result comes out. Say the short-term investor is right, and the stock falls 10% – but over the next 5 years the stock performs in line with the expectation of the long-term investor and increases 200%. Both investors are sticking to their process around investment horizon. Both made money. Someone else sold the very bottom to allow the short-term investor to cover. Was this another emotional / behavioural investor giving up? Who knows.

Another observation of active management crowding in even large stock universes is how stock screens and AI research tools are used.

Let’s assume you start with 10,000 stocks (Bloomberg World – large, mid, small index) and you decide the best way to look for opportunities that meet your criteria are to screen for companies that have an ROE >25%, net cash balance sheets and a P/E Ratio of less than 15x. You quickly get to a shortlist of ~200 stocks. But guess what, that is the same screen, and the same list of stocks thousands of other investors are also using.

The world has more sophisticated screening tools today. Witness the incredible complexity and capability of AI investment research. Despite increased sophistication and compute power, the end state will probably be like the stock screen above, with everyone looking for the same thing, just with a different tool.

The second issue is how quality is defined. Looking for high ROE stocks does make some sense, as ROE tends to be stable through time (Mouboussin again) however if the ROE does mean revert – and high ROE is a flame for the moths of competition in a capitalist system – then the investment returns are likely to be very disappointing.

In a world where AI is being used as a ladder to disrupt the moat of incumbents, going after the most attractive business models with the highest ROEs is the logical thing to do. Many will fail, as the vibe coded ladders are not long enough, and drop the disrupters into a moat of crocodiles, but for those that succeed the downside to the disrupted will be significant.

“I very frequently get the question: ‘What’s going to change in the next 10 years?’ And that is a very interesting question; it’s a very common one. I almost never get the question: ‘What’s not going to change in the next 10 years?’ And I submit to you that that second question is actually the more important of the two — because you can build a business strategy around the things that are stable in time. … [I]n our retail business, we know that customers want low prices, and I know that’s going to be true 10 years from now. They want fast delivery; they want vast selection. It’s impossible to imagine a future 10 years from now where a customer comes up and says, ‘Jeff I love Amazon; I just wish the prices were a little higher,’ [or] ‘I love Amazon; I just wish you’d deliver a little more slowly.’ Impossible. And so the effort we put into those things, spinning those things up, we know the energy we put into it today will still be paying off dividends for our customers 10 years from now. When you have something that you know is true, even over the long term, you can afford to put a lot of energy into it.” – Jeff Bezos, November 2012

When investing, what remains stable through time is human behaviour. Until such point all investment decisions are totally handed over to the machines, the behavioural aspects of markets are likely to remain true in the future. Having an investment process deliberately designed to deliver alpha from behavioural biases – like those discussed above – is also more likely to work in the future.