A topic I have changed my mind on during my career is concentration in funds. I used to be strongly of the view that it was only worth taking active investment positions if they came with high conviction – usually in the form of concentrated positioning – otherwise, what’s the point? Although I came to realise I was wrong about this, I am aware that many people far smarter than me remain advocates of this type of approach. Why do I think it is a problem when they don’t?

Fund concentration is not the easiest concept to define. There are obvious examples such as very focused equity portfolios with large weightings in individual companies, but there is more to it than that. Concentration doesn’t have to be about position sizes in stocks, it can come through an extreme sensitivity to a certain theme, concept or risk factor. It is about our exposure to specific and singular points of failure.  Could one thing go wrong and lead to disaster?

The key concept to consider when thinking about the risk of concentrated funds is ergodicity. This is a horribly impenetrable term, but at its core is the idea that there can be a difference between the average result produced by a group of people carrying out an activity, and the average result of an individual doing the same thing through time.

Let’s use some simple examples.

Rolling a dice 20 times is an example of an ergodic system. It doesn’t matter if 20 people roll the dice once each, or an individual rolls the dice 20 times. The expected average result of both approaches is identical.   

Conversely, home insurance is a non-ergodic system. At a group level the expected average value for buyers of home insurance is negative (insurance companies should make money from writing policies). So, why do we bother purchasing it? Because, if we do not, we expose ourselves to the potential for catastrophic losses. The experience of certain individuals through time will be dramatically different to the small loss expected at the average group level.

Investing is non-ergodic. Our focus should therefore be on our individual experience across time (not the average of a group); this means being aware of how wide the potential range of outcomes are and the risk of ruin.

In concentrated funds, the prospect of suffering irrecoverable losses at some point in the future is too often unnecessarily high.

‘Risk is not knowing what you are invested in’

One of the most common arguments made by advocates of running very concentrated equity portfolios is that it is an inherently lower risk pursuit because we can know far more about a narrow list of companies than a long list. If we have a 10 stock portfolio we can grasp the companies in a level of detail that is just not possible if we hold 100 stocks, and this depth of understanding means that our risk is reduced. The first part of this is right, the second part is wrong.

The problem, I think, stems from the fact that there are two types of uncertainty – epistemic and aleatoric. Epistemic uncertainty is the type that can be reduced by the acquisition of more data and knowledge. Here the idea of portfolio concentration lowering risk makes sense. Conversely, aleatoric uncertainty is inherent in the system; it is the randomness and unpredictability that cannot be reduced. It doesn’t matter how well we know a company or an investment, we are inescapably exposed to this. The more concentrated we are, the more vulnerable we are to unforeseeable events.

While I think a neglect of aleatoric uncertainty is at the heart of unnecessarily concentrated portfolios, there are other issues at play. Overconfidence is likely to be a key feature. we may be aware that the range of outcomes from a concentrated approach is wide, but that may be desirous to us because we believe that our skill skews the results towards the positive side of the ledger. Given our ability to fool ourselves and the aforementioned chaotic nature of the system, this seems to be a dangerous assumption to make.  

Unfortunately, there is also an incentive alignment problem. A wide range of potential outcomes from an investment strategy becomes very appealing if we benefit from the upside but someone else bears the downside. This asymmetry is inevitably one of the reasons why high-profile macro hedge funds so often seem to be swinging for the fence with concentrated views. The often-severe downside of the negative outcomes are borne primarily by the client (a situation no doubt exacerbated when a hedge fund manager is already exceptionally wealthy).

Running a very concentrated investment strategy places an incredibly heavy onus on being right and also leaves us acutely vulnerable to unforeseen events unfolding that can have profoundly negative consequences. Exposing ourselves to such risks wilfully seems imprudent and unnecessary.

Investors are likely to overestimate how much they know, and underestimate how much they cannot know.

—

– Being wary of concentration does not mean increasing levels of diversification are always beneficial. There is a balance to strike.

– It feels important to note that the risk of concentrated strategies can be diversified by combining them, but we should still consider what the concentration levels say about the investor who is willing to adopt such an approach.

—

My first book has been published. The Intelligent Fund Investor explores the beliefs and behaviours that lead investors astray, and shows how we can make better decisions. You can get a copy here (UK) or here (US).

Given how volatile and unpredictable financial markets are, thinking in probabilistic terms is an essential skill.  It allows us to acknowledge uncertainty, express our level of confidence clearly and gives us more freedom to change our mind. There is a problem, however. Expressing ourselves in probabilities doesn’t come naturally and is often actively disliked. At best it is regarded as spurious accuracy, at worst evidence of an absence of conviction. We value bold and singular predictions about the future, not caveats and caution. How do we encourage probabilistic thinking in a world that doesn’t want us to?

One idea I happened upon in Julia Galef’s excellent book “The Scout Mindset”, is the ‘equivalent bet test’, which is a decision-making tool described by Douglas Hubbard in “How to Measure Anything”. 

The concept is simple.

Let’s take a fairly generic claim that a professional investor might make. They think that the US ten-year treasury yield will rise above 5% before the end of the year. That’s great, but what does it actually mean? Are they certain that this will happen? (Given the historical accuracy of bond yield predictions, I hope not) Or are they only 51% sure? The difference matters a lot, but we have no idea. How do we find out? By creating an equivalent bet, where we are certain of the odds.

We say to our forecaster. There is now $100,000 at stake. We will give you this amount of money at the end of the year if your prediction on treasury yields is right. Alternatively, we will give you the same amount of money at the same time if you can pick a blue ball from a hat containing six blue balls and four red balls. You can only choose one of the bets – the treasury yield forecast or the drawing the balls from the hat.

If they decide to delve into the hat, then we know that their confidence in their bond yield forecast is less than 60%.

We can then adjust the ball selection bet to a point at which the forecaster is ambivalent about the two options. We then we have a reasonable guide to how confident they really are about their prognostications.

This is clearly an imperfect approach, a hypothetical $100,000 will almost certainly provide a different decision making response to a real sum of money, but it is likely to be broadly consistent and incredibly helpful.

Not only does the equivalent bet test encourage the forecaster to think about their judgment in probabilistic terms, it also provides a far greater level of clarity about both how confident an individual is and how well-calibrated (or not) they may be.

Whether we like it or not, we live in an uncertain, volatile and probabilistic world. Our decision-making approach should reflect this.

—

Galef, J. (2021). The scout mindset: Why some people see things clearly and others don’t. Penguin.

Hubbard, D. W. (2014). How to measure anything: Finding the value of intangibles in business. John Wiley & Sons.

—

My first book has been published. The Intelligent Fund Investor explores the beliefs and behaviours that lead investors astray, and shows how we can make better decisions. You can get a copy here (UK) or here (US).

Industries in which capital has become abundant and optimism unbridled often end up disappointing investors. Echoes of these capital cycle pitfalls are also evident in active fund manager selection. In the ‘alpha cycle’ a certain area of the market delivers high returns, stories then emerge not only about the compelling opportunities to be found but also about the savvy investors who are exploiting them. This intoxicating mix of unusually strong performance and seemingly irrefutable narratives draws increasingly large flows from investors who either believe the tales being woven or are compelled to chase the momentum. The more extreme this performance dynamic, the more investors are likely to mistake luck for skill, and the cyclical for the structural. Leading them to make the wrong decisions at a terrible time.*

It is a curious phenomenon that investors seem to behave as if capital flooding into a certain type of fund (whether that be investment style or market subset) is a prelude to higher returns in the future. Of course, there is the potential to capture an ongoing trend, yet from a fundamental perspective abnormally high returns are likely to be the result of assets becoming significantly more expensive.

Exceptionally positive performance now is drawing returns from the future – we are not going to make them again. We just act as if we will.
 
At the heart of this alpha cycle problem is our propensity to believe that what is cyclical is in fact a structural shift.  We are prone to see an active manager delivering strong performance in an in-vogue area of the market as someone with durable skill, rather than simply benefitting from largely unpredictable tailwinds that will at some point reverse. The most dangerous situation is when we begin to believe that there has been a permanent change in markets (XYZ is the only way to invest) and a particular active manager is the exemplification of this approach.  Here we have the glorious opportunity to be wrong twice – about both markets and skill.

As flows into a certain style of active manager increase, so too does the clamour to invest and the belief that it is the obvious thing to do (performance doesn’t lie). As our prospective future returns dwindle, our conviction increases.

At points of performance extremes, whether an active manager has skill will become irrelevant. If they are investing in part of the market enjoying euphoric sentiment and following a period of unsustainably high returns, any edge will be overwhelmed by the almost inevitable reckoning.

Selecting active fund managers who have enjoyed prodigious tailwinds comes with twin challenges. First, we are likely to grossly overstate the presence of skill (we cannot help but conflate performance with skill). Second, even if they do possess an edge, it is unlikely to matter because the odds of investing successfully in an area that already has stretched valuations and delivered exceptional performance are poor. 

We might argue that a fund manager has the ability to navigate such situations adroitly. Rotating out of areas that have delivered spectacular results and into less glamorous segments. Perhaps. Yet most managers have stylistic features or characteristics that almost inevitably leave them exposed to certain market trends. Furthermore, the type of fund managers that will attract the most attention during a particular cycle are likely to be the purest representation of whatever has been working.

The problems of the alpha cycle do present a potential opportunity. Fund managers with an out of favour approach who are investing in an unloved part of the market are likely suffering from poor performance and holding assets with far more attractive valuations. Even if they don’t have skill, the odds of a good outcome might be compelling (certainly better than investing at peak cycle).

There is a problem, however. We are very unlikely to believe managers somewhere near the trough of the alpha cycle have any skill, they are also likely to be haemorrhaging assets and they may be in danger of losing their job. It is never just a cycle remember; these are always profound structural changes at play.

I have framed the problems that stem from the cyclical nature of investment returns as an issue with active funds – it is not. These same behaviours exist across asset classes, regions and sectors. It is just that active funds have the additional danger of skill being used to justify unsustainable performance.

Investors need to adopt the mindset that unusually strong returns are a prelude to lower returns in the future and behave accordingly.

—

* I use ‘alpha’ in this article in the broadest possible sense – general outperformance, rather than anything more technical.

—

My first book has been published. The Intelligent Fund Investor explores the beliefs and behaviours that lead investors astray, and shows how we can make better decisions. You can get a copy here (UK) or here (US).

One of the most common questions I have heard through my investment career has undoubtedly been: “what’s the catalyst?” I have definitely asked it myself a few times but have tried to abstain in recent years. Most discussions around catalysts are an effort to anticipate what will change the view other investors hold about an asset – a pretty challenging task. We appear to believe that because catalysts seem obvious after the event, they must be predictable beforehand. This is a dangerous assumption.

There are two types of investment thesis – one about something that is already performing well, here a catalyst is unnecessary because we just need to extrapolate; the other is about something that isn’t working right now, and then a catalyst seems to become essential. The desire to identify a catalyst probably stems from our bias toward believing that current trends will persist, coupled with our discomfort at being uncertain – we want to know exactly how and when something will change.

It is worth taking a step back to consider what it is we are typically doing when identifying a catalyst.

There are three elements:

1) Predict an occurrence.

2) Predict how other investors will react to it.

3) Hold other things constant.

There are problems in each of these components – we are poor at making predictions about future events, we aren’t great at forecasting the reaction of other investors, and things are never constant. Other than that, we are all set.  

Catalysts are asking us to specify in advance a specific driver of a change in the return profile of an asset or security. This is no easy task.

Although I am generally sceptical about catalyst predictions there are certain instances where it is more reasonable, particularly with micro-level decisions. For example, if an investor were to suggest that a company selling an underperforming unit could be a catalyst for higher returns to shareholders – this would seem specific and defensible (difficult but defensible).

The broader an investment view – I think US equities will start to underperform because X will happen – the more fanciful it seems. It just becomes far too complex. (I have looked back over my twenty-year career and found 476 catalysts identified for Japanese equity outperformance).

As with most things in investment decision making, the obsession with catalysts is, in part, a time horizon issue. If our horizon is appropriately long-term, we don’t need to predict what specific catalyst will change a certain investment’s fortunes, we can wait for the gravitational pull of fundamental factors to take hold (assuming we are right). The shorter our horizon the more sentiment matters – we are explicitly attempting to guess the behaviour of other investors, so we need a view on what might change that. (Short-term investing is hard).

It is safe to assume the success rate for identifying catalysts for potential shifts in the return profile of an investment is pretty low. We need to be right twice – about changing performance and the precise reason it will happen. The first one alone is hard enough.

If we are in the business of specifying catalysts, it would be prudent to keep a record of the judgements that we make through time. We may not like what we find.    

At the start of this piece, I said that “catalysts seem obvious after the event”. The critical word here is “seem”. In most cases it is incredibly difficult to confidently specify the exact causes of a change in the performance of an asset even after it has occurred. We can certainly tell a great (and simple) story as to why something has happened, but the truth is almost always more intricate. If we cannot do it with the benefit of hindsight, trying to do it with foresight feels unwise.

Will there be a specific catalyst that alters the return profile of an asset? Maybe. Can we identify it in advance? Probably not. Do we need to? No.

—

My first book has been published. The Intelligent Fund Investor explores the beliefs and behaviours that lead investors astray, and shows how we can make better decisions. You can get a copy here (UK) or here (US).