Why Do Fund Investors Neglect Base Rates?

Given the vast amount of time, effort and resource dedicated to fund research it would seem absurd to suggest that the most critical pieces of information are ignored or neglected, but it often seems that they are.  Fund research can be defined as an inside view activity; one where the precise details of a particular case can dominate our decision making. This might be the pedigree of an active manager, an impressive track record or the attractions of an in-vogue investment theme. The allure of these specific features leaves us prone to ignore the general lessons from similar situations (the outside view). This means we make decisions without even knowing where the ballpark is[i].

The most effective form of outside view is the formulation of a base rate.  Although popularised in recent years, the idea of employing a base rate in decision making can still seem a somewhat arcane activity.  But a base rate is simply an observation derived from an appropriate reference class of similar examples[ii]. If we knew nothing about the specifics, then the base rate would be our only guide.  Let’s say we wanted to estimate the operating margin of a certain airline. The inside view would be derived from the specific features and operations of that company, whereas the outside view / base rate would be based on the operating margins of all airlines.

For fund investors, we can think of the base rate as answering two questions:

What do similar scenarios tell us about the likely outcome?


What do similar scenarios tell us about the odds / probabilities attached to our decision?

Fishing for a base rate

Imagine we are asked to make a forecast of how many fish a fisherman will catch on a single day. We speak to the fisherman and find that they are incredibly skilful and experienced and utilise the most sophisticated equipment.  They are supremely confident that they can catch more than 10 fish. We have our inside view. Is that sufficient to make a robust prediction? No. We need an outside view.  On further investigation we find that over the past year fisherman at the lake catch on average only one fish per visit. This is our base rate.

We might refine our reference class and look at the average catch of fisherman at different times of year or in certain weather conditions, but a good base rate needs to balance sample size and specificity.  Once we have a base rate, we understand the history of similar scenarios and have an idea of the odds of certain predictions being correct (the chances of catching more than 10 fish seems low). We are now much better placed to decide.

The base rate is a vital piece of information. Imagine attempting to forecast the number of fish caught without it.  It does not provide us with the answer, but it does offer a starting point.  We can still make a decision that is contrary to what the base rate tells us.  For example, we might argue that fishing is an activity heavily influenced by skill, and therefore the skill of the chosen fisherman will render the base rate less relevant – but at least we are clear in the assumptions we are making. Underweighting the base rate in a transparent fashion is far better than ignoring it.

Fund investing and base rates

Unfortunately, there is no single base rate that we can take from the shelf to apply to our decision making. We need to define and identify an appropriate reference class. There also might be multiple angles around which we wish to apply a base rate to a decision.  There are a range of base rates fund investors might adopt. Here are some hypothetical examples:

Over the past 10 years X% of active managers have outperformed the benchmark. 

This is a relatively straightforward and easy to source base rate with a clearly representative reference class. We don’t have to be beholden to it, but we should be explicitly aware of it as a measure of the odds of success in selecting an active manager. Again, we might argue that the figure is unrepresentative of the future because, for example, the dominance of a select group of large companies made it unusually difficult to improve on the benchmark over the past decade. However we use such a base rate, we should be clear on what it is and how it impacts our choices.  

On average, funds that outperform by more than 10% annualised over 3 years, return X% over the subsequent three years.  

This is a different type of base rate more attuned to the tendency of fund investors to chase past performance and for fund excess returns to mean revert. If we are buying a fund with a stellar track record it is critical to understand how likely it is that this will continue; particularly as the inside view at the point of maximum performance will be incredibly compelling. A good example of this type of thinking is an article by Morningstar observing the performance of funds after they had risen by more than 100% in a single calendar year.[iii]    

Equity markets that have outperformed over the last decade and are relatively expensive, underperform by X% over the subsequent decade.

Base rates for fund investors are not simply about the classic, binary active versus passive debate. All investors are active in some form. Even if the underlying funds we use are trackers, the asset allocation involves active decision making – nothing is purely neutral. If we have benefitted from being heavily invested in US equities for the last decade by virtue of the construction of a particular index, we still need to apply base rate thinking to the implicit views we are expressing.

Why are base rates ignored?

Given how valuable base rates are to effective decision making, it is somewhat puzzling that they are not more widely and obviously used by fund investors. There are several factors that limit their employment:

Overconfidence: An exaggerated belief in our own abilities means that the base rate or average does not apply.  Odds of 10/1 against are meaningless to us.

Exceptionalism: Overconfidence means that we think that we are exceptional; in addition to this we tend to think that our specific case is the exception. This time / this one is different, so the base rate again does not apply. 

Salience and Stories: The inside view is far more interesting and persuasive than the outside view / base rate, therefore it tends to overwhelm it. The backstory to a specific manager or market is so much more diverting than looking at historic experience and averages.

Time: As with all incidents of (poor) investment judgments time has a central role to play.  Base rates often exert their influence over time but can be of little use over shorter time horizons. Without the right level of patience or appropriate environment the importance of base rates will diminish.

Difficulty: Although the concept of base rates is relatively simple, its application for fund investors is not easy. Aside from having access to the required data, it is not always obvious which element of a decision we wish to apply a base rate to and what the appropriate reference class is. There is no perfect answer here but doing something is a dramatic improvement on doing nothing.

Incentives: At times base rates will tell us information that we don’t want to hear, so we consciously or unconsciously ignore them.  

An attempt to incorporate an outside view and ascertain appropriate base rates should be integral to any fund investment we make.  Establishing a base rate not only improves our judgement, but it brings clarity as to why we have made a particular decision.  We need to allocate time away from the specific to better understand the general.

[i] https://www.mckinsey.com/business-functions/strategy-and-corporate-finance/our-insights/daniel-kahneman-beware-the-inside-view#

[ii] https://plus.credit-suisse.com/rpc4/ravDocView?docid=gIamqy

[iii] https://www.morningstar.com/articles/1017292/what-to-expect-from-funds-after-they-gain-100-or-more-in-a-year-trouble-mostly