The investor’s dilemma – luck or skill?

Performance does not persist, as shown by the latest scorecard from S&P Dow Jones Indices

Neil Woodford and Crispin Odey, two of the UK's most successful money managers over the last three decades, are under serious pressure after enduring a torrid few years.

So too are some of their American counterparts, such as high-profile hedge fund managers David Einhorn and Bill Ackman, who have badly underperformed in recent times despite their stellar long-term records.

Are they unlucky? Is it wrong to say a fund manager has lost his touch just because he has lagged over a particular period? Or might it be the case that their original success was little more than luck, and they have now been found out? When it comes to investing, how can you distinguish luck from skill?

It's perfectly normal for even the most successful money managers to lag markets over a multi-year period

The aforementioned managers have underperformed for different reasons. Woodford, long regarded as Britain’s answer to Warren Buffett, is a long-term value investor, while Odey is an aggressive hedge fund manager who has been wrongly betting on a “devastating” market decline since 2015. Ackman has a very concentrated portfolio that has cratered due to a number of ill-timed bets, while Einhorn has lost out due to his scepticism towards some of the US markets’ most fashionable and expensive stocks.


Underperformance is normal

Whatever one thinks about the investors in question, it's clear a money manager should not be written off simply because they have underperformed over a given period. Index fund giant Vanguard recently analysed the performance of the most successful US equity funds over the last 15 years. Despite their success, more than 70 per cent underperformed for three consecutive years.

Vanguard's analysis is echoed by Aaron Reynolds, author of a 2011 report entitled The Truth About Top-Performing Mutual Fund Managers. Reynolds focused on 370 funds that delivered the goods for investors over a 10-year period. Of those, every single one experienced at least one year where they underperformed peers. More than half had a year where they badly underperformed by at least 10 percentage points. The vast majority – 81 per cent – suffered at least one three-year period where they underperformed peers.

In other words, it’s perfectly normal for even the most successful money managers to lag markets over a multi-year period. All too often, however, investors are not given the benefit of the doubt. A 2015 State Street study of 400 institutional investors found almost 90 per cent looked to replace a manager after just two years of underperformance.

Accordingly, most managers elect not to rock the boat, says Woodford.

“They take that route because where I am now is about as uncomfortable as it can get professionally. It is a horrendous place to be.”

Woodford's complaint is shared by Ben Inker of GMO, the fund giant founded by contrarian icon Jeremy Grantham. A GMO client who invested a dollar at inception would now have $18.40, Inker wrote in a report last month, almost double that of its benchmark index. However, many clients did not share in those gains because of their hiring and firing decisions, said Inker.

“More than half fired us in 1999 and 2000” after a few difficult years; “even more clients hired us between 2004 and 2007” following GMO’s “extremely strong performance off of the tech bust”; now GMO is in the firing line once again due to its underperformance in the current market cycle.

Past performance

While skilled managers may lose clients through no fault of their own, the corollary is also true: many unskilled managers attract an influx of money after enjoying a lucky run of outperformance. However, investors eventually learn the old cliché about past performance being no guide to future returns is true. Performance does not persist, as shown by the latest scorecard from S&P Dow Jones Indices.

It examined what would have happened if an investor had bought into a fund on the basis that it was in the top quartile of performers in the 12 months prior to September 2015. Disappointment awaited. Just 18 per cent of funds stayed in the top quartile over the coming year, even worse than chance would suggest. By September 2017, only 6.4 per cent of the original funds had maintained their initial outperformance.

The same phenomenon is evident over longer horizons. S&P looked at the top quartile of funds over the 2007-12 period and found only a very small minority repeated their performance over the following five years. In fact, the best performers were more likely to move to the bottom quartile than to remain at the top.

Clearly, buying a fund on the basis of recent performance is a mug’s game, but many investors are unaware of this. Accordingly, the industry responds to investor naiveté by continuing to launch new funds. By chance, some are bound to outperform and can be advertised accordingly, while the losing funds are withdrawn (roughly one-third of the worst-performing funds over the 2007-12 period had disappeared by September 2017).

The irony is that investors would often make more money by selecting the underperformers instead of the outperformers. One study that examined the hiring and firing decisions of 3,400 institutions between 1994 and 2003 found managers are hired after strong periods and fired after weak periods. Ironically, the higher the results were prior to hiring, the more returns subsequently fell. Fired managers, in contrast, were more likely to outperform over the following three years.

Skill or luck?

Clearly, you cannot judge a fund manager’s ability on the basis of their recent results. How does one judge an investor such as Crispin Odey, the Brexit-supporting hedge fund manager who has outperformed over the last 20 years but who has now lagged his peers over the last one-, three-, five-, and 10-year periods? With great difficulty. It’s incredibly difficult to distinguish luck from skill.

Nobel economist Robert Merton recently estimated that if you wanted to be 95 per cent sure that a manager is skilful, they would have had to average annual returns of over 30 per cent over a 20-year period. Few, if any, investors have ever managed returns of that magnitude.

Many investors might think Merton’s criteria is too strict, but it’s clear investors underestimate just how much data one needs to make definitive judgments.

When we are thinking fast, we tend to overweight the value of small samples

Say you have two coins: one is fair and the other has a 60 per cent chance of coming up heads. How many coin flips would you need to perform to give yourself a 95 per cent chance of correctly identifying the biased coin? Last year, quantitative expert and London-based Elm Partners founder Victor Haghani asked 700 financial professionals this question. The median guess was 40 coin flips. In reality, you would need to flip the coin 143 times to be 95 per cent certain, noted Haghani.

“When we are thinking fast, we tend to overweight the value of small samples,” cautioned Haghani, who noted that a full 30 per cent of respondents thought 10 flips or less was sufficient. “This built-in bias to overweight small samples results in a tendency to ignore the investing dictum ‘past performance is not indicative of future results’ when we clearly should not.”

To return to the original question, investors such as Neil Woodford, Crispin Odey, Bill Ackman and David Einhorn are entitled to protest that they should not be prematurely condemned. Unfortunately for their investors, it’s equally true to say their earlier success may owe as much to luck as skill.