The Unintended Consequences of "Sophisticated" Performance Measurement

Investment performance measurement is a striking example of an originally good idea made bad by its success.  It is true that a standardized way of measuring performance is necessary to weed out money managers’ claims that are outrageous, plain dishonest, or even merely “selective.”  But unfortunately, too many imitators embraced the original concepts; and as the size of the performance-consulting industry grew exponentially, more gimmicks were incorporated in addition to the simple indicators used early on by measurement pioneers.  Today, these augmented concepts are very widely used, often inadequately so, and in my opinion have frequently become counterproductive.


Most exceptionally successful long-term investors have proclaimed, at one time or another, their skepticism about investment consultants and the growing use of performance benchmarks aimed at splicing performances among investment styles, geographies, company sizes, sectors, etc.  The skeptics have included the likes of Warren Buffett, Charlie Munger, Peter Lynch, Martin Whitman, Jim Rogers, Seth Klarman, Georges Soros, Howard Marks, et al.  Recently a new Oxford University paper joined this critique.  The Financial Times of September 22, 2014, cites a study by a team of from Oxford University’s Saïd Business School, which analyzed consultants for more than 90 per cent of the retirement market.  It concludes, “On an equal-weighted basis, U.S. equity funds recommended by consultants underperformed other funds by 1.1 per cent a year between 1999 and 2011.”

Why should the record of consultants be so disappointing?  Certainly not because they are stupid or ignorant.  In fact, most are not only highly educated, but also are very proficient in math and statistics as well as highly articulate in their presentations.

Instead, I believe that their problem is rooted in the fact that what was once a practice has become a business.  This business requires consultants to foster a growing appetite for their services among clients, by creating a need for more frequent measurement and decision-making.  Thus, measuring and critiquing performance has become a quarterly practice – and at times, more often than that.

Unfortunately, in investing, a quarter or even a year is almost always a totally irrelevant period.


When I referred above to successful long-term investors, I did not mean successful over three, five, or even ten years, which in investment history amount to little more than one fashion or style season.  I mean a stretch of years encompassing several cycles, with bull and bear markets as well as many fads and fashions and their aftermaths.

If one overlooks the “noise” of superficial hiccups and false signals, economic events really progress at a near-tectonic pace that does not require constant monitoring.  Except for occasional accidents, corporate fortunes do not change in three months either, and new business strategies often take several years to bear fruit – or not.

Over shorter periods, it is mostly crowd psychology that moves markets.  Thus, trying to measure how well a portfolio has performed over three or six months really amounts to measuring how well a manager has participated in the mood-induced ups and downs of a bipolar group (the investing crowd).  Once we realize this, the risks of quarterly performance-measurement become clearer.


Over time, as the performance of a majority of hedge funds and the consultants who recommend them has proven disappointing, it appears that the selling argument of this relatively new industry has shifted from “superior returns” to “acceptable returns with lower risk.”  This too, I am afraid, reflects a dangerous misunderstanding.

Consultants – and the academics who gave them their theoretical arguments (including Nobel laureates) – cannot quantify risk.  This is because risk can only be measured for individual investments rather than groups or indexes; it necessitates exhaustive analysis; and it cannot be summed up easily in a single statistic.  So, a consensus developed among consultants to use volatility as a substitute for risk in their calculations.  And, eventually, they even began to call it “risk.”  But this is a serious misinterpretation.  Financial risk refers to the possibility of permanently losing some or all of your equity in an investment.  Volatility merely refers to the amplitude of price fluctuations within given periods.

Since the principal determinant of short- and medium-term price fluctuations is the mood of the investing crowd (as opposed to changes in the fundamental value of companies, for example), most successful investors welcome volatility.  Over their longer time horizon, they recognize short-term volatility as a periodic opportunity to find unrecognized value or growth.

By definition, for an investor to be better than the majority, he or she needs to be different.  This implies that he or she can also, occasionally, be worse.  Bernie Madoff’s notorious Ponzi scheme, where he sold billions of his funds to investors by producing made-up performance statistics (the investments did not actually exist), was particularly smart in one respect:  The performances “produced” and implicitly promised were not outrageous.  Good investors achieved similar results over the years.  But one giveaway was (or should have been) that this performance showed no volatility:  Almost the same results were achieved period after period.  A good analysis should have told investors that this kind of performance could not be achieved without occasional shortfalls.  Merely statistical methods did not.

Most performance consultants try to minimize or eliminate volatility.  In doing so, they all but abandon the possibility of being durably better than the majority; but, in my view, they do not reduce true risk, which can only be avoided through extensive fundamental (not just statistical) analysis of specific investments.


Another criticism aimed at consultants by highly successful investors is their primary focus (also out of business necessity) on performance relative to various fabricated benchmarks, rather than on absolute results, which would answer the question, “Am I becoming richer after inflation and taxes or not?”

Seth Klarman, founder of the Baupost Group, is at 57 years old often regarded as one of the very best investors of his generation.  In his 1991 book, Margin of Safety (which is out of print but can be bought for around $2000 on the Internet), he makes the argument that I summarize here:

Most institutional investors have become locked into a short-term, relative-performance derby.  Their short-term orientation may be exacerbated by the increasing popularity of pension-fund consultants.  Money managers motivated to outperform an index or a peer group of managers may lose sight of whether their investments are attractive or even sensible in an absolute sense.  They then really act as speculators:  They try to guess what others are going to do and then try to do it first.

Not only are money managers thus forced into becoming short-term traders, but when obliged to invest against a benchmark, suddenly everything for them becomes relative.  Instead of asking, “Is what I am buying cheap?” the manager begins asking, “Is the asset I am buying cheap relative to the benchmark?”  As performance measurement increasingly deviates from the original goal of money management – absolute returns – the ultimate nonsense may become having to reward a manager for losing only 30 percent (for example) if his benchmark (often chosen by the consultant) was down 50 percent!

Michael Edesess, a visiting fellow with the Centre for Systems Informatics Engineering at City University of Hong Kong, recently reviewed a paper by three academics from Boston University and the London School of Economics.  The paper, entitled “Asset Management Contracts and Equilibrium Prices,” argues that measuring and evaluating manager performance by comparing it to a market index may be distorting prices across the whole market.  Edesess goes one step further:

One inescapable conclusion is that the practice of evaluating managers by monitoring their performance against an index benchmark should be jettisoned – even if there’s no immediately obvious alternative to replace it.


What makes a good portfolio manager?  First, the choice of a discipline that makes sense based on both theory and experience; and second, the faithful application of that discipline through economic and market cycles.  Performance-measurement services are not readily useful for that purpose because, as famous finance author Charles D. Ellis reminded us, they do not report results; they only report statistics.

In fact, from a client’s perspective, investment performance should be assessed more as a manager’s navigating ability than as a crude statistic:  How has a portfolio performed in different economic and market environments; did it perform as expected from the chosen discipline or not; and if not, why?  Assessing a portfolio manager’s “seaworthiness” thus requires a multi-cycle perspective and a more subtle understanding of where any over- or underperformance came from.  For example, a good statistical performance attained by not following a promised discipline usually is a warning of future trouble.


I will not go so far as Warren Buffett in saying that diversification is what you use when you are not sure of your choices.  The fact is that one is never a hundred percent certain of one’s choices.  So, it seems desirable to diversify, though there is much disagreement among practitioners over the optimal degree of diversification. The debate is an old one: is it better to spread your eggs among many baskets, to reduce the risk of breaking, or to put all your eggs in one basket and carefully watch that basket? Personally, I favor one basket, albeit a large one, because I am concerned that too many investments may wind up working at cross-purpose. Similarly, over-diversifying among managers appears counterproductive to me, especially since it often results either in overlapping portfolio positions (which defeats diversification efforts) or, on the other hand, in contradictory positions in different portfolios, which tend to offset each other’s impact on performance.

Another argument against diversifying among managers is that, to my knowledge at least, most of the firms that have had superior long-term records have not done it.  However, very often, their funds are not managed by an individual, but by a very close-knit team.  This is an intelligent way to critically vet both new investment ideas and new talent, thus enhancing portfolio management while ensuring the continuity of the organization and its disciplines.

This promise rings particularly true when a managing organization’s culture closely aligns its managers’ fortunes to those of its clients.  Seth Klarman pointed out that a client probably would not choose to dine at a restaurant whose chef always ate elsewhere.  He thought an investment client would no more be satisfied with a money manager who does not eat his or her own cooking.  Many investment managers treat OPM – “other people’s money” – differently than their own.  If, in contrast, your manager has all his or her eggs in your common basket, you can at least feel assured that the basket is being watched.

François Sicart
January 14, 2015

Disclosure: This article reflects the views of the author as of the date or dates cited and may change at any time. The information should not be construed as investment advice. No representation is made concerning the accuracy of cited data, nor is there any guarantee that any projection, forecast or opinion will be realized.

© Tocqueville Asset Management

Read more commentaries by Tocqueville Asset Management