Via Negativa (What Should Fund Investors Not Do?)

As fund investors we spend a great deal of time deliberating the positive steps we can take to achieve better investment results. This is undoubtedly an important endeavour, yet there is something easier and more effective that should come first – deciding what we should not be doing.   

Nassim Nicholas Taleb highlighted the benefits of a focus on eliminating errors in his book Anti-Fragile where he describes the theological idea of via negativa, which is a means of explaining God by virtue of what it is not, rather than what it is. Taleb broadened this concept to contend that it is easier and more beneficial to stop negative activities than to attempt to identify new, constructive behaviours:

“You know what is wrong with more certainty than you know anything else.” [i]

We can eliminate the mistakes that we know are damaging and costly far more easily than discovering positive behaviours that might improve our fortunes. The more complex and unpredictable the environment, the more likely this is to be true.

This is an approach that should be adopted by fund investors, who face an unfathomable array of choices and decision points. Rather than obsessing over how to define the precise allocation to the right funds at the right time – an incredibly difficult task – it would be more productive to first concentrate on the actions we should avoid. Prioritise omission over commission.

So, what is it that fund investors should not do?

1) Don’t buy into a fund after extreme positive performance: Abnormally strong performance on the upside is highly unlikely to persist, investing after a spell of stellar returns is a horribly asymmetric bet.

2) Don’t be concentrated by fund, manager, style or asset manager: Concentration is the surest path to severe losses, it implies we know far more about the future than we actually do.

3) Don’t predict short-term market movements: We cannot predict the short-term behaviour of markets or funds that invest in them. We should not make decisions that suggest that we do.

4) Don’t check short-term fund performance: Short-term fund performance is typically nothing more than random noise, checking it frequently encourages poor decisions.

5) Don’t use performance screens: Given that strong fund performance tends to mean revert, it is hard to think of a worse way of filtering a universe of candidate funds than by ranking on the strength of historic returns.

6) Don’t keep selling underperforming funds to buy outperforming funds: A common behavioural trait which feels good at the time we do it, but compounds into a pernicious tax.

7) Don’t buy thematic funds based on strong backtests: If a fund is being launched based on an in-vogue theme with a stellar backtest the chances are we are already too late.

8) Don’t invest in active managers if we cannot bear long spells of poor performance: Even skilful active managers will underperform for long periods, if that is unpalatable invest in index funds.

9) Don’t invest in funds if you don’t understand how they make money: Investing in things we don’t understand is a recipe for disaster.

10) Don’t persist with an active manager when they start doing something different: A circle of competence for a manager is usually incredibly narrow, if they are venturing outside of this we should avoid them.

Each of these prohibitions is attempting to do the same thing – establish a simple heuristic that is generally effective in a highly uncertain environment. Although most investors prefer to embrace positive actions, putting a stop to poor behaviour is a much easier win.

[i] Taleb, N. N. (2012). Antifragile: Things that gain from disorder (Vol. 3). Random House.

Extremely Bad Decisions

There are undoubtedly countless investors nursing losses because of a choice made in recent years to abandon value-oriented strategies to fully commit to the rapidly accelerating growth bandwagon. Although the brutal reversal in the more speculative end of the growth universe can be read as a salutary lesson on style and factor rotations, it is not. It is about the dangers of making pro-cyclical decisions when returns reach extreme levels. Rather than worry about such excesses, investors willfully and gladly embrace them. We put more chips on the table as the odds of success deteriorate. *

Whilst it is impossible to precisely define ‘extreme’, our focus should be on two measures – performance and valuations.

Any investment strategy with unusually strong performance should be a cause for concern. If a reasonable expectation for excess returns from a (highly) skillful active manager is 2% per annum and a certain fund delivers 10% annualised outperformance over five years, this is not a time for adulation of a new star manager but a potential sign of a dangerous extreme. Things that cannot go on forever won’t go on forever.

There are times when performance alone can be misleading. For example, an investment fund could deliver what look to be abnormally strong results, but this might be a recovery from a tumultuous prior period. This doesn’t mean there is no risk, but it is a different type of situation. Utilising valuations in combination with past performance is the best signal for unjustifiable extremes.

Valuations can be considered in a multitude of ways but the most effective is to observe the relative valuation of a strategy or market segment through time – how expensive does it look compared to its benchmark through history? The more richly priced, the more uncomfortable we should be.

Historically strong relative performance and elevated valuations are unlikely to be sustainable and create a perilous situation where reversion or normalisation can lead to severe losses. The more extreme these measures become, the greater the risk of bad outcomes – both in terms of magnitude and likelihood.

The added problem for investors is that as returns and valuations soar, the pressure to invest will intensify. A lack of exposure to the in-vogue area will lead to painful underperformance, which will become increasingly pronounced as the trends continue. Everyone will be asking why we are not participating.

The strength of returns won’t be treated by most as a problem or a threat, but a validation of the skill of active managers in the sweet spot or proof of some new paradigm.  The more extreme the situation, the more persuasive it becomes.

If we do invest heavily into a strategy exhibiting extreme performance and valuations, we will inevitably find ourselves in the perverse situation where we have made a bad decision with the odds stacked against us, but where everyone will be congratulating us for making it. Prices become extreme for a reason – most people will be believers, and we will be viewed as having made a smart call.

It will also make us feel better. Buying in at extremes will come as a relief as we will no longer have to discuss why we’re underperforming or don’t hold enough of the fashionable areas or funds.

What happens after we invest in a strategy exhibiting extreme valuations and performance?

Well, they will continue to work for a time.  We won’t be so unlucky as to call the peak. Prevailing trends might continue to run for months and years. There will probably be sufficient time for us to take a victory lap for our decision, but the likelihood is high that at some point there will be a painful reckoning.

We cannot predict when extremes will correct. They will go on for longer than we ever expect and reach levels, which we thought were unobtainable. It is not a viable investment approach to simply sit on the sidelines and complain about the unfairness of it all.  We should, however, exhibit great caution in the decisions we make when valuations are rich and performance is remarkably strong, as this is the time where we are most vulnerable to costly mistakes that might take years or decades to recover.

This type of post is easy to write after there has been a reversal and some prior extremes have been extinguished (or at least dampened a little). Extremes are easy to spot and bemoan after the event. Yet this is about the general, not the specific. Any investment approach can be vulnerable to extremes. So how should investors deal with them?

During the excitement of extremes everyone will obsess over the inside view and be desperate to justify the returns and valuations of the current situation (this time it’s different). We should ignore this and instead focus on the outside view – what are the lessons from history about returns following similar excesses? 

When we reach extreme heights, it probably won’t matter how skillful a fund manager is, how good our research might be or how persuasive the narrative is; this will all be overwhelmed by an erratic but inescapable gravitational pull.  

If there is a sign of an extreme, we should be asking – what is the base rate for success of investing in funds or strategies with valuations and performance at these levels? How often has investing at such a time worked out well in the long-run?

The force and salience of extremes means that they can easily dominate our behaviour. The risk of investing too much, in the wrong area, at the wrong time is never greater.

Beware of extreme decision making.

* I have focused here on the risks of making pro-cyclical decisions when valuations are historically stretched and performance has been remarkably strong. The reverse of this is that there must be opportunities in being counter-cyclical at negative extremes – this is true but difficult and dangerous. It is easier to discuss what not to do, than what to do!

Why Can’t We Stop Making Short-Term Market Forecasts?

Our fascination with forecasts about short-term market movements is a puzzling trait.* Investors continue to make market predictions even though we are consistently wrongfooted. When other people make an incorrect forecast, we don’t disregard what they tell us next but instead say to them: “you were entirely wrong before, what do you think will happen next?”

This is not just a diverting human quirk; it is an incredibly damaging and costly behaviour. To save ourselves we need to stop listening to short-term market forecasts, and certainly not make them.

It is true that even the best investors make plenty of mistakes, but that doesn’t mean we should cut market forecasts some slack. There are four critical aspects, which mean that they should be readily ignored:

1) Most people are terrible at making market forecasts: The track record of market predictions is bleak. From tactical asset allocation to macro forecasting, history is littered with experts and amateurs alike who were undone by developments that they failed to anticipate. If everyone struggles to do it well, why do we pay attention to predictions, or believe that we will be the exception?

2) It is not reasonable to expect people to be good at making market forecasts: Creating accurate forecasts about a noisy, complex system is close to impossible. It is entirely fanciful to believe that anyone can consistently succeed in it.

3) Forecasts can have major consequences: Most market-related forecasts tend to be bold with severe consequences when mistaken. A fund manager being incorrect on a single stock is an entirely different proposition to a stark forecast about an impending market decline. When these predictions are wrong (which they usually are) the impact is typically harsh.

4) Market forecasts encourage poor behaviours: Forecasts about markets encourage the worst behaviours in investors such as short-termism and over-trading.  Also, if we believe that we can forecast markets it diminishes the perceived need for diversification.

Markets constantly remind us that attempting to forecast an inherently complex system is a fool’s errand yet, despite this, we cannot seem to resist it.     

Why do we keep making and listening to forecasts?

Our willingness to ignore the incontrovertible truth about market forecasts is driven by a range of factors from our own behavioural limitations to the structure of the asset management industry:

Overconfidence: Engaging in forecasting is a perfect example of how overconfidence can influence our behaviour. The trait encompasses three distinct forms: overplacement, overestimation and overprecision[i]:

Overplacement – it doesn’t matter if most people are abject at making market predictions, we place ourselves in the upper echelons. We are better than most people.  

Overestimation – we not only think we are superior to others, but we also significantly overstate our own skill and judgement.

Overprecision – we are far too certain that we have the right answer. This can be particularly damaging for investors who are prone to under-diversify and take injudicious decisions based on forecasting prowess.

Hindsight bias: Once something has happened, we cannot help but view it as an inevitable outcome. As everything becomes obvious after the fact, it feels like it was eminently predictable before it. Of course, it is not. What we witness is never inexorable, simply one path taken amongst many other possibilities. 

Incentives: Professional investors are incentivised to forecast market movements. Fees and careers depend on it. Giving clients the belief or sense that market, political and economic developments can be confidently predicted is an incredibly attractive sales pitch (even if it is an empty one).

Comfort: It is difficult living and investing in an uncertain world. Forecasts about the future give us comfort amidst chaos and unpredictability. Even though they are likely to be inaccurate, at the time they are made they make us feel better.

Fooled by randomness / survivorship bias:  If we have a large group of people making market forecasts then inevitably there will be some who are correct; this would be the case even if everyone was making guesses at random. That there are always some people who have called it right makes us believe that it is possible – “somebody got it right, why didn’t we?”. The mistake we make is to assume that some successful forecasts within a group at one point in time (the survivors) means that certain individuals can consistently make good predictions through time.

False Expectations: There is an expectation that investors should have an opinion about ‘markets’ and saying “we don’t know” is incredibly difficult to do (and generally not a good career move). The default is to have a view even when we know that it is likely to be unfounded.  

Despite it undeniably being a loser’s game, it is far easier to make predictions about markets than not. Part of the reason it remains such an ingrained feature of the industry is that most participants are more willing to keep playing the lucrative game, rather than take the more difficult path of realism and education. 

Why Do We Learn the Wrong Lessons?

Given that our forecasts and predictions are consistently mistaken it would seem we have ample opportunity to realise the error of our ways, but we don’t take these. Instead, we learn the wrong lesson over and over again. Whenever we are blindsided by an event or occurrence our tendency is to review what happened and ask – what did we miss? Or – how can we adapt our approach to make better forecasts in the future?

This type of reflection seems sensible but is far from it. The lesson that we should learn (but never do) is that short-term movements in financial markets (and everything related to them) are far too difficult to accurately predict, and we should stop attempting to do it. We need to make our investment decisions on the basis that we cannot forecast most things, not that we can.

But perhaps this statement is too definitive, there are better approaches to forecasting. So, should investors try to improve before abandoning the activity altogether?

Can Investors Superforecast Markets?

Philip Tetlock’s work on superforecasting has risen to prominence in recent years and given credence to the idea that although forecasting is generally done poorly, there are certain individuals that can consistently beat the average and experts. 

Although superforecasting is often framed as a group of particularly skilled and well-calibrated people; the real lessons are in how they make their forecasts and why this leads to better results.[ii] 

The key tenets of superforecasting include an ability to make probabilistic judgements (which allow for the expression of uncertainty), an awareness of base rates (appreciating the odds) and Bayesian updating (adjusting probabilities based on new information). If we are going to indulge in market forecasts, these should be central pillars of our approach. Despite the evidence that forecasting can be improved however, when it comes to markets investors should still avoid it.

Even if we take proven steps to enhance our forecasting abilities the challenge for investors remains too great. Financial market predictions are an exercise in extreme complexity. Not only are we attempting to anticipate how certain issues and events will unfold (known unknowns), but any market forecast will also be implicitly incorporating situations that have not yet occurred (unknown unknowns). To make matters worse, our market predictions are always about second order effects. Projections about how investors in aggregate (the market) will react to other developments. Superforecasting might help a little, but it is nowhere near sufficient to solve something quite so problematic and noisy.

Superforecasting extols the benefits of combining a probabilistic approach with Bayesian updating. This means when we receive new information, the odds should shift based on our assessment of its implications. Although undoubtedly a better approach to forecasting, it is hard to see how investors could realistically implement such an approach. Not only is it a herculean task to correctly judge information and ascribe probabilities, but these would somehow need to be converted into positions and trades. Critically, the inherent uncertainty in financial markets also means that frequent adjustments would be required resulting in high turnover and trading costs. It is simply not feasible.

Predictions about the future are sometimes inescapable and where investors must make them, we can seek to learn from the success of the superforecasters. Unfortunately, however, there is no magic formula to anticipating the short-term fluctuations of financial markets.

One of the problems of market forecasts is that they are so easy to make. That they can glibly roll off the tongue, belies how fiendishly complicated the activity is. When we see or hear one, it is worth taking a step back and thinking through exactly what foresight is being claimed and quite how absurd it is to believe that anyone possesses it.

* It is difficult to precisely define what short-term is, but the shorter the horizon and the greater the specificity in a forecast the more problematic it is likely to be. Forecasting that equity returns will be positive on a twenty-year view, is an entirely different proposition to predicting how the Chinese A Share market might perform over the next six months. Anything inside one year is incontrovertibly very short-term.

I have written about how and when we should make forecasts here.

[i] Moore, D. A., & Schatz, D. (2017). The three faces of overconfidence. Social and Personality Psychology Compass11(8), e12331.