Sticky Luck

 Two:

Greatest Investors and Efficient Market Theory

Efficient market theory, which suggests fund managers cannot beat the market, is rejected by chartists and intrinsic-value analysts.  Are Buffett and Lynch their proof?  Are there other gurus to add to the list?  An analysis of mutual fund performance shows no evidence of fund management talent.  A tour through the Greatest Investors reveals nothing that cannot be ascribed to sticky luck.  Any Buffett and Lynch magic appears almost certainly time-specific management talent, not arm's-length portfolio management.

The term Efficient Market Hypothesis (EMH) does not sound too interesting to the average investor.  Some may respond immediately with skepticism – a typical comment might be, “The market isn’t even rational, let alone efficient!”  Let me agree that the market is not necessarily rational, depending upon how we define the term.  But efficiency, as used in the term EMH, need have nothing to do with rationality.  In fact, the most common criticism of the EMH seems to be based on nothing more than a misunderstanding of what efficiency means in the market context.

Since there is no single definition of the EMH that would be appropriate for my purposes, let me attempt my own definition.  So as not to be accused of inaccuracy, I will refer to my version as the Sticky Luck Market Hypothesis or SLMH. 

The EMH itself was coined after the study of a great deal of market data, to provide a possible explanation for the patterns of that data.   The historic performance of stocks was being analyzed to see if there were any patterns that could then be used to predict future market performance – at least well enough in order to consistently beat the market.  There were always patterns aplenty to be found in any history.  However, if the investor started at the present time and then used only the patterns of the past to predict the future from that point, no consistent market advantage was found.  In other words, patterns always looked very useful in hindsight but, no matter how long they had persisted, there was no special financial advantage in assuming those patterns would continue into the future. 

We should take a moment to reflect on what a curious discovery this is.  How can long patterns be evident in the past but, as soon we face the future, cease useful existence?  Why should there be such a huge break between looking backwards and looking forwards?  A good part of this book is devoted to explaining why the stock market behaves this way, and how difficult it is for investors to internalize that fact.  The SLMH, as I am going to define it below, is one piece of that explanation:

The flow of information in today’s stock market, and the reaction of investors to that information, is too fast and complex for an individual investor to make anything but a “gut feel” immediate gain by responding to that information within seconds.  Moreover, after those seconds, the market has already fully reflected that new information, as far as we can ever detect it.  Any further price movement cannot be traced as a consequence of the information itself, though it may be a consequence of investors’ responses to earlier price movements, including the one related to the new information.

Here’s an example. A company releases its quarterly earnings, which are below market expectations.  Within seconds, the price of the stock has dropped 50c.  Thereafter it may recover, or fall further, by any amount.  But these subsequent price movements are not a consequence of the earnings release itself.  It’s true that some investors may only learn the news later in the day and decide to sell their stock belatedly, helping to drive the price down.  But there is just as much chance that other investors, seeing the price has fallen, decide that the stock is now a good buy.  There is also a good chance that some investors, seeing the price drop, anticipate further price drops and decide to sell.  In other words, the market movement that can be explained only by the actual news release has ended within seconds.  Thereafter, it is impossible to separate out any residual effect of the news from the effect of the price movement itself.  Moreover, other market news that relates to the company, it competitors, their industry and even the economy emerge all the time and quickly make any further “news-effect” untraceable.

Nothing Rational About Efficiency

Note that I have not used words like “rational,” “efficient,” “proper,” “accurate” or “correct” which are so often used in attempted EMH definitions (frequently by its detractors) and which can be so misleading.  At the end of the trading day, our stock may have actually risen above its starting level and those who sold after the initial drop may suffer remorse.  With the benefit of hindsight, commentators will say they “panicked.”  Those who bought at 50c off believed the sellers panicked, and feel smug that evening.  But then maybe the stock drops a dollar the next day and there is no discernable news to explain it.  The smug and remorseful are reversed. 

It is usually the buyers who then throw up their hands and say, “and people like you believe the market is efficient!”  But in the sense that the word efficient is used in a proper EMH definition, this strange event is in fact a good illustration of the hypothesis.  From the evening before, looking into the future, no one could use the earnings release as a predictor of market movement.  For a whole day, the market of millions of investors had shrugged off the bad news.  The greatest genius in the world could only guess at the next day’s behavior.  There are no inefficiencies which would allow some math whiz to expect to beat the market based on an earnings release and price movements more than a number of seconds old.

I have allowed for the “gut feel” of gifted traders who, on getting immediate access to the latest news, have a sixth sense of the broader market reaction and pounce on it.   I believe such successful individuals must exist, and their skills must be well rewarded, in order to drive market efficiency.  But the benefits of their skills will flow only to them and their sponsors (typically financial institutions).  Though we can invest in some of those sponsors, the future expected earnings from this activity is already reflected in the sponsors’ market price.  Hence any “beat the market” strategy through such trading is unavailable to investors who are not both directly involved in such constant, very short-term trading activity, and who do not possess the skills, huge infrastructure and risk capital needed to profit from it.  In particular, it will not be provided through any mutual fund – the risk-reward structure of such funds provides almost no incentive to even seek it.

Thoughout this work I will refer both to efficient-market theory as is commonly recognized today, and also my own SLMH which avoids any assumption about rational investors (or a pivotal role of arbitrageurs, as we will discuss later.)  I will generally refer to them both under the term efficient-market theory, using the term theory in its looser sense.  Since I believe that the SLMH incorporates the appropriate elements of behavioral finance, I will not use efficient-market theory as a opposite concept , but as one that can embrace all appropriate elements of both schools of thought. 

Hair-Triggers for Any News

For those familiar with the working of financial markets, there is a lot of commonsense support for efficient-market theory.  It is difficult for outsiders to fully grasp the vast scale of financial markets, or the sophistication of both the players and the software programs that analyze all available data and respond to it with buys and sells.  If a company is traded, then every little factoid it releases is incorporated into multiple analysts’ databases and thereby triggers market action.  Once that immediate action is complete, it requires another factoid to trigger further action.  That factoid could come from the price movement resulting from the original factoid, from other companies, other markets, general statistical data releases, rumors, weather patterns or other acts of God, or the release of some analyst’s report which he or she has been working on for weeks.  That release may be confined to a single office (i.e. be no discernable news to the market as a whole) but may trigger a large buy or sell which becomes another actionable factoid.

An analogy would be the workings of a human body.  If we were to get stabbed with a dirty needle, we know instantly that we will experience pain for a period of time.  But whether or not there is further, related pain in the coming hours, the next day, the next month or next year, will depend upon billions of chemical reactions about which we do not have enough information to make a prediction.  Like the poor earnings news, the needle stab is a bad sign, yet our body may shrug it off and even be stronger from the antibodies.  The shortcoming of this analogy is that the body is mechanical and we could imagine, at least in theory, some technology so sophisticated that it could trace and predict every cell reaction in the body.  This would not be possible even in theory with the stock market, as future movements depend upon the future decisions of free-thinking individuals.

The EMH was the dominant working theory of the market since its inception in the 1960s until roughly the early 1980s, when it started to come under increasing criticism.  Part of the popular criticism, as I mentioned, stems from a basic misunderstanding of the term efficient.  However, serious scientific inquiry also suggested persistent anomalies in pricing, and gave birth to the behavioral-finance school.  Certainly, a body of convincing evidence arose to support the idea that emotion is a powerful force in the market.  The next step in the logic was to suggest that experienced practitioners could take advantage of anomalies and emotional swings in order to achieve consistently superior investment results.

This last conclusion is fortunate for the industry that can be called money management, or fund management, or investment management.  The primary purpose of this industry is essentially to make decisions about purchasing stocks or other securities that are listed on exchanges, in order to maximize the investor’s returns.   Despite the importance I attach to risk management, expenses and tax efficiency, I think it is fair to say that the majority of investors believe they pay fund managers to maximize returns.  More crudely, they are paid to “beat the market” or beat other managers.

But if efficient-market theory is actually true, how can those fund managers be expected to perform better than a randomly selected portfolio of stocks?  All known information is reflected in current prices.  Even if we think the market has over-reacted to a piece of information, we cannot know if and when it will correct that over-reaction.  And how certain can we be about an over-reaction, since we may not have full access to all information that caused other buyers and sellers to act, which may prove that the actual situation was worse or better than our current information suggests?

So What is the Value of Active Fund Management?

In fact, efficient-market theory suggests that the many millions of dollars paid by investors in fees to fund managers do not buy superior performance.  If the past provides no reliable insight into the future, what is the purpose of paying for a manager’s guesses which will be no better than our own, regardless of how much data he possesses and analyzes?  Though we do need people to manage funds, this could simply be an administrative task and the pay for such work could be a small fraction of the current levels.  While it is true that fund management is a profession fully open to competition, it would appear that the highly-paid superstars have just been lucky in the past, and are no more likely to be lucky in the future than anyone else. 

Naturally, the fund-management and advisory industry has provided counter-arguments, and they have been increasingly successful in the debate since the early 1980s.  This debate has not been limited to behavioral finance.  The runaway success of equity investment since 1982 has spawned vast volumes of investment advice for the general public, which is under no constraint to follows the rules of scientific inquiry.  Anecdotal evidence is amassed, together with specific “lessons from history,” which have a cumulatively persuasive power.  This incremental approach, together with its message of hope and its clear parallels with all other walks of life (find the wise guide and you, too, can beat the averages – they’re just averages anyway, right?) appeals to the way the ordinary man or woman makes decisions about unfamiliar matters, whether it is major family events or political choices.  Wise-sounding aphorisms abound, such as “know what you buy” – the more I learn, the better I’ll be, though maybe never as good as those professional fund managers.

How do people get good at investing?  I will divide all the arguments of the industry into two general camps:  the “chartist” camp which believes we can learn the future by mapping the past, and the “fundamentals” camp which focuses on “intrinsic value.”  Later I will argue that, from the perspective of efficient-market theory, these are really both the same conceptual camp after all, but look different because they approach the same idea from a different direction.  However, in this chapter, before we have dug deeply into theory, I will treat them in their conventional forms.   I should also mention that these are not exclusive camps.  Many managers and advisors claim to use tools from both.

Chartists and Fundamentals

The chartist school predicts the likelihood of future price movement from past price movement, rather than anything related to underlying causes such as the company’s or the economy’s detailed statistics.  It relies upon patterns it can detect in price movements that, its advocates claim, have a tendency to repeat themselves.  This school also includes those who believe that the price movements of certain securities are linked, but in a staggered fashion so that a movement in one set of securities (e.g. transportation) can be an early-warning sign for a movement in another set (e.g. the market as a whole).  A classic example of this theory in action was triggered by the steepening of the yield curve in early 2002, which many market commentators took as a bullish sign – a badly wrong call, in that particular event. 

There is no question that repeated patterns can be seen in historic data.  The vital question is whether or not the study of such patterns can be expected to improve investment performance.  To the casual reader, especially the non-mathematician, the answer to that question would seem already contained in the first observation – if there are patterns, surely we can make good use of them.  A simple retort is that we can see patterns every day in clouds, but that makes us no better at predicting the shape of the next cloud.  I intend my main thesis to be a more convincing response.

The fundamentals camp believes that it is possible, based on detailed analysis of a particular company, to determine its “intrinsic value.”  Intrinsic value is the total of money it will generate for an investor into the future.  If we apply a factor for the “time value of money” to the company’s future earnings stream we can arrive at an “intrinsic value” in terms of today’s stock price.  If we can buy the stock more cheaply on the market, we have gotten ourselves a bargain.  In fact, even if we don’t have enough money, the logic of fundamentals tells us that, provided the right kind of credit is available, we should borrow money to buy such a stock.

The fundamentalist (if I may use such a term in this context) pays little attention to short-term market movements and their causes.  The intrinsic value of a stock is heavily dependent upon the performance of the company over future years, let alone weeks.  A short-term price dip is simply an opportunity to make more money, provided the fundamentals have not changed.  However, unless the fundamentalist is determined to hold the stock forever, he or she must believe that the future business performance they predict will be eventually rewarded by an exceptional price increase, as other investors see the emerging future good news (unexpected to them now as they have not done the fundamentalist’s analysis) and bid up the price.  In other words, the fundamentalist doesn’t really rely upon news to make investment decisions.  He or she may spend months understanding a company.  They then set a “target price” – the price if the rest of the market were to recognize the stock’s potential vis a vis other investments.  Sooner or later – often there’s no prediction on timing – those other investors who rely upon more obvious evidence such as news will keep buying until the target price is reached.

As we’ll discuss later, belief in the virtue of fundamentals meets some important psychological needs of investors.  The focus on value gets us away from the short-term, predatory mania often associated with stock markets, and the perceived villain of bubbles and busts.  It suggests a steady, unruffled expert who does hard, big-brained work and distains fashions and the herd mentality.  Since most peoples’ investment goals stretch over generations, the focus on long-term success is also appealing.  Few people encapsulate this image of vision and wisdom more than Warren Buffett. 

Buffettology

Buffett has proven a valued ally to the fund management industry.  First, he has himself criticized efficient-market theory, saying (with characteristic modesty) that he would be penniless if it were true.  Second, he has proven highly critical at times of the fund management industry in general, proving that he is an independent thinker and not part of some industry cover-up.  Third, his performance has so outstripped that of the rest of the market over forty years that he would seem to be the conclusive proof that good fund management does add higher investment returns.  If Buffett exists, it would seem to follow that other talented fund managers must also exist, who have the same ability to understand fundamentals.  Buffett may be at the front end of a bell curve, but there must be plenty of good guys behind him before we get to the average manager.  We just need to identify them, and there must be plenty of data available to do that.

We’ll let Buffett’s performance speak for itself.   In its 2005 annual report, Berkshire Hathaway showed that it had earned an average return of 21.5% over the period 1965-2005.  This compared with an average return of 10.3% on the S&P 500, which is a reasonable gauge of the return on the total US stock market.  The cumulative effect of this difference is breathtaking:  whereas $100 invested in the S&P500 in 1965 would have accumulated to some $5,600 – no mean feat – the same amount invested with Buffett would have accumulated to a mind-blowing $293,000.   The annual report also points out that the S&P500 return is before-tax, whereas the Buffett return is after-tax.  The case seems overwhelming.  How could an average additional return of more than 11% per year for a period of forty-one years be ascribed to luck? 

We should first differentiate between Buffett’s ability to pick winners among publicly-traded companies on the one hand, and his ability to manage an insurance company and negotiate and manage acquisitions on the other.  I believe that a large part of his success has come from these latter investments, and I would not challenge his superior skills in these fields.  Few would question the fact that business geniuses exist.   Doing deals, managing people and making judgments about information to which a limited number of people have access are each skills that can differentiate the best executives on a consistent basis. None of these activities, however, are equivalent to the selection of freely traded securities to which all investors have equal access in terms of price and information.  Buffett’s activities that involve significant management and board influence are a business talent that can be compared with any highly successful CEO.  These activities, because of their personal business influence, do not equate to portfolio management.  I strongly suspect Buffett’s returns as a pure portfolio manager – if any significant portion of his activity could be described as such – are substantially less than his performance as a whole.  But for the moment I will assume they are also quite  impressive. 

Ericology

Now I will introduce my friend, Eric.  Eric started investing around the same time Buffett focused on Berkshire Hathaway.  He had inherited $1,000 from his grandfather’s estate and resolved to invest it until his retirement, scheduled for 2005.  For almost nine years be invested in an equity mutual fund which did a bit better than average.  Then came the Yom Kippur War in 1973, and Eric was convinced it would escalate into World War III.  He panicked, selling his mutual fund and investing the proceeds in T-bills.  After about a year he felt a little foolish about his prediction.  In the office one day, the boss arranged a tour of the mainframe computer his company used.  Eric was transfixed by this gleaming machine.  He had a vision of a future where there would be almost no work except the building of such huge machines, which would do all other jobs for us.  He promptly invested his money in a hi-tech mutual fund and determined to leave it there until retirement.  For the next 20 years he had several severe temptations but did not move his money.  Then Y2K emerged as a threat.  He had visions of many computers failing on that fateful day.  Despite the preparatory hi-tech spending, he thought, it would be dwarfed by the activity after it struck.  Then came the anticlimax.  As the world celebrated, he became morose and concluded that the Internet was a fraud, a source of manic behavior which would soon be abandoned.  He hesitated for a couple of months  then sold his tech holdings.  In a future low-tech world, he decided, real estate would be the only item of true worth, so he dumped his money into REITs. 

Eric was never known by his family or friends for his sense of judgment, which was usually poor.  He could never hold down a job for very long.  Yet he retired with a sum of over $700,000 from his original $1,000 stake, with an average annual return of almost 18%.

My fictional friend Eric made just three very timely decisions, largely for reasons that could not be described as profound, if not outright wrong-headed.  Few people will be as lucky as Eric but, among the world’s hundreds of millions of investors, I would doubt that he is unique.  In fact, if we start with the assumption that all investment choices are a matter of luck, it’s easy to accept that there will be a few Erics around.  Why not also a few Buffetts?  In other words, why should the existence of Buffett and his track record prove that any particular approach to choosing freely traded stocks actually works?  As already mentioned, if we are to consider him as a fund manager rather than as a direct managerial influence on business, much of his activity would not qualify for comparative measurement.  The annual 21.5% will largely be the result of hands-on company relationships.  By how much will this figure drop if Buffett plays the market anonymously, just like the rest of us?   I don’t know the answer but, as we lower our sights from 21.5% per year, and allow for pure luck – let alone the sticky luck we will describe – Eric’s 18% looks uncomfortably close.

I realize that the fundamentalist will be irritated, but not dissuaded.  Yes, similar luck was achieved in even shorter periods by investors in Microsoft, Yahoo!, Cisco Systems, Google and others, particularly if we include their private equity.  And only one timely decision was necessary in those cases.  But the point is that Buffett actively manages his portfolio, making thousands of investment decisions, all of which could not have – statistically – been a matter of luck.  Anyway, he uses a rigorous analytical method which, if we were able to sit down with him day to day, we would find left little to sheer chance.

Note that this defense abandons the argument that his record speaks for itself.  We are now forced to accept that his (non-interventional market-trading) record may be shared with dozens, maybe hundreds, of lucky people – including, incidentally, early retail purchasers of Berkshire Hathaway stock who have little but blind faith in the man.  Instead we have to fall back on the elegance of the technique, rather than its clear result.  We know what Buffett has told us about his technique, of course, but there isn’t enough information provided to understand his critical decision points, let alone begin to reproduce his work.  He would be the first to admit that he has made many mistakes, so I don’t think consistent performance among his thousands of decisions is a telling point.  His range of experience is probably like the rest of us – a number of spectacular successes covering for a number of dogs.  

I have no doubt that Buffett genuinely believes his form of analysis is the best way to determine the value of businesses and that, to date, his approach has proven to be well ahead of the market.  But Buffett – or anybody – can do no more than model with human tools the vast complexity of the future.  He is like an excellent doctor who is renown for his prognoses.  But for the analogy to work we also have to imagine a world in which the human body is rapidly evolving.  His prognosis models have worked splendidly for the past forty-three years, just as the hi-tech industry performed spectacularly for almost 26 years between 1974 and 2000.  Unfortunately, both of these are what I refer to as sticky-luck events. 

Despite Buffett’s exceptional record, the market chooses to award him just a tiny fraction of its available funds to invest.  No matter how good his track record, we sense that he has just as much chance of stumbling next year as anyone else.  The fact that we can think in this way, in the face of such a consistent and successful history, suggests that we have a gut feel for sticky luck.  Unfortunately, we do not extend this gut feel to an explicit investment strategy, eliminating the counter-productive behavior evident with many investors.  Though the market as a whole can conclude that Buffett, left to rise to his free-market price, is not some unique investment, its individual investors still seek the latest fount of future wisdom.

There is never a shortage of claimants.  Whether through chartism or intrinsic value, the financial industry exudes a confidence in its fund managers and the “value add” they bring the investing public despite their significant fees.  Yet those market forecasters who are paid purely for their advice – at least the ones that we public have access to – tend to offer a bag of very mixed messages.  Is it possible that fund managers get more consistent – and maybe better – advice, or have some knowledgeable way of sorting through the white noise and getting to the truth?

Mutual Fund Management

I expect the reader will have knowledge of studies that show mutual fund managers, as a whole, to do a mediocre job.  In A Random Walk, Malkiel states that, for the period 1986-2005, actively managed large-cap funds underperformed the S&P500 by almost 1.5% per year.  The percentages of underperforming funds were 68% over three and five years, 79% over ten years and 82% over those twenty years.

Such data seems dismally convincing.  However, I thought I would present my own results, using the type of data that would be quite easy for the reader to reproduce.  We all have easy access to mutual fund investment tools which show performance track records.  Moreover, I wanted to go beyond large-cap funds.  In this case, benchmarking is less obvious, so I wanted to be transparent about my process and let the reader judge for him or herself.

I went to my Ameriprise Financial brokerage account and selected all available diversified equity mutual funds which were able to provide a ten-year average performance.  This option gave me 503 funds.  I further selected only those which had an inception date of at least fifteen years prior to the date of performance measurement, which reduced the total to an easily manageable 152.  Note the apparent explosion of mutual funds between 1992 and 1997.  There is little doubt that many more were started during that period to capitalize on the rising market, but a lot of these new funds may have replaced older ones that were retired or merged.  Those that had survived since at least the start of the 1990s economic expansion are clearly a select group.  Poorly performing funds are inevitably closed, as they find it more difficult to attract new investors.  It is a simple task for a mutual fund company to add a new fund name and move around its managers to give the appearance of fresh faces.  The 152 that still compete with some 351 newcomers clearly have something to boast about.  They are truly the fittest in their survival through a perfect Darwinian process, where performance is both transparent and also overwhelmingly the reason for success. 

The performance we are going to measure has been, according to language attached to the table, assembled by Morningstar.  Morningstar makes the expected disclaimers about their material not being warranted to be accurate, complete or timely.  So who is responsible for the authenticity of the performance data?  It is not clear, but we will trust the fact that the system works honestly.  We also tell ourselves that the funds are most unlikely to sell themselves short on their performance, or to let Morningstar do so.   Of the 152 older funds, five showed a blank for the performance since inception.  By eliminating these funds from my analysis, I think I am unlikely to be biasing my results downwards.

To measure the quality of performance, I used Morningstar’s SBBI (Stocks, Bonds, Bills and Inflation) 2007 Yearbook.  This publication shows the total return generated by a number of different sectors of the market for each month since 1926.  The book divides stock returns between large and small capitalization.  It does a less detailed analysis of growth and value stocks.  Each of the funds surveyed was categorized by Morningstar as one of large, mid or small cap, and one of growth, value or blend.  For the periods since inception, I matched the market index returns for large and small caps with the funds in these classes, and an average of the large and small caps with the mid-cap funds.  The numbers of large, medium and small cap funds were 100, 34 and 18.

The Growth-Versus-Value Phenomenon

The SBBI Yearbook provides its own analysis of growth and value only since 1968 and only by year, rather than by month.  Nevertheless, I believe it is important to take the difference into account.    A series going back to 1928 was constructed by Fama and French, using hand-collected data, and provides a consistent picture for this earlier period to the post-1968 data.  Since the post-1968 period dominates in our survey, I think it is reasonable to apply its relative averages to the averages from our funds.

The importance of the growth-value adjustment stems from the remarkably different performance of the two classes over the 39-year period.  The average annual returns of value stocks exceeded growth stocks by some 2% in the case of large caps, 4% in the case of medium caps and 6% in the case of small caps.  The surprising size of these differences warrants comment.  The SBBI methodology is based on book-to-price ratio.  However, there is no single definition of book value for this purpose, and the Yearbook describes several different methodologies.  More importantly, fund managers may have a quite different view on what the terms “value” and “growth” means.  We know that Morningstar has done the mutual fund classification, and no doubt they used a consistent methodology to their Yearbook, but it is very possible that different funds sail in and out of their definitions inconveniently.    Here is a quote from a growth fund prospectus:

“The Fund’s Advisor looks for companies with stocks that are believed to have strong earnings growth potential and are reasonably valued at the time of purchase.”

And here is another fund’s summary from Yahoo Finance:

“The investment seeks capital growth over the long term. The fund normally invests at least 80% of assets in a broad portfolio of common stocks of companies with market capitalizations equal to those within the universe of Russell Midcap Growth index stocks.”

These fund descriptions sound a long way from Morningstar’s definitions.  Even if we assume Morningstar is able to monitor the fund’s actual holdings – very unlikely historically – these definitions allow the funds to be both value and growth at different times.  More likely, Morningstar follows the fund’s own description, which may again turn out to be value or growth from a book-to-price perspective, and vary with time.

So we have a significant problem.  We know that the book-to-price ratio makes a big difference, especially for smaller stocks, but we don’t know whether the mutual fund classification is quite so relevant.  We suspect strongly that fund managers tend to make full use of the latitude they give themselves in their prospectuses.  And a chip maker with expensive fabrication facilities during a bear market will have a value-stock ratio but may well be seen by the manager as a growth stock. 

Since our selection of mutual funds contains 69 growth stocks, 32 of them mid or small, and only 39 value stocks, 6 of them mid or small, we are not in a neutral position.  We also suspect that many funds have classified themselves as growth as it is more appealing, most investors being unaware of such stocks’ laggard performance.  This classification may well be more of a marketing posture.   Since Morningstar divides all stocks into growth and value by market capitalization, it would be strange for this set of all long-lasting funds to be truly so lopsided by this definition.  But if we force ourselves to assume that Morningstar got their fund classification precisely in line with their Yearbook definition, it would make a difference of 0.4% per year to the overall benchmark for the average fund performance.  As the growth-value analysis goes back only to 1968, this also involves assuming a similar adjustment for the 23 funds started before 1968.

The Results

Using only the large, mid and small cap indices, I calculated the geometric average annual return for each fund and its related index.  I then took the arithmetic average of all the funds’ average annual returns and also of the related indices.  The average fund return was 11.7%, compared with an index return of 12.2%.  If we reduce the index return by 0.4%, to allow for the fact that it represents a blend rather than the greater weight towards growth that could be argued for the funds’ investment strategy, the two figures are startlingly similar.  Our average fund returned an annual 11.7% compared with an adjusted index of 11.8%.

Fund managers would immediately point out that their figures are net of transaction costs, while those of the Morningstar indices are not.  So it is here that I would be inclined to put some pressure on the 0.4% adjustment.  If reduced to 0.2%, it would allow 30 basis points for expenses.  This may be low historically, but half the funds under consideration were started after 1985.  It is interesting to weight the funds by years since inception, so we give equal weight to every fund performance year rather than every fund.   This increases the lag of fund performance behind the indices by some 50 basis points.  It is reasonable to explain this result by the greater emphasis upon earlier years when transaction fees were much higher.  Yet we are still dealing within a very narrow range.  Another 20 basis points for transaction costs would still make our average fund performance much too close to the indices for mere coincidence.  The difference remains well within the margin of error of the measurement process I have used.

We only have to consider the spread I quoted earlier between small growth stocks and value stocks – an average difference of 6% per year over 39 years – to put the average fund performance in perspective.  The average term of these funds was some 29 years, and over that time period we have seen dramatic market changes.  Why didn’t these funds, these survivors of the fittest, manage to show significant daylight between themselves and the indices? 

We often hear accusations that most actively managed funds are closet index funds, but this is not at all clear from the distribution of results.  The bottom quartile of performers was some 2% per year below their indices, even after the full growth-value adjustment, and over 3% below without it. The bottom decile was 3.5% per year below their indices with the adjustment, and 5% below without it.   It appears that funds did make significantly different investment choices.  It is simply that the net effect of these different choices for the complete group was effectively zero.  To the moderate extent that funds beat or underperformed their indices, there would be no reason to suppose that luck was not the entire source of difference.

Who Benefits?

That is, except for the volume of work that must have gone into the day-to-day management of these funds.  With an average fund life of 29 years, this group represents a total of 4,224 years of fund management.  Let’s assume that each fund employs on average five people for a full 2,000 hour year.  This assumption seems very modest, but it implies 42 million hours of work.  With the benefit of hindsight, we can now confidently tell all the investors that they would have been better advised to have put all their money in index funds (assuming they had existed in the earlier years.)  After all, I think most investors would rather be assured of the index return than a random result either side of the mark over which they have no control.  We are faced with the appalling notion that 42 million hours of work amounted to the well-paid digging of ditches and their subsequent refilling.

I should once again emphasize that these equity funds are the mutual-fund industry’s pride and joy, and its poster-children for the benefits of long-term equity investment.  Undoubtedly, advertising returns of 12% per year over an average of 29 years is enough to get the investing public enthused about the merits of equity mutual funds.  However, what is the purpose of studying the past performance of the individual funds themselves?   The funds themselves always tell us that past performance is not a guide to the future, and now we are beginning to understand why.  There is no dishonesty involved here.  The investing public wants track records, and uses them in their selection criteria, and so the industry provides them.  If we listen to the words of each advertisement, we find no one in the industry claims to be able to beat the others in the future.  They tell you if they have beaten others in the past, because that is factual.  You cannot blame them for being highly selective about the time periods which best suit their comparisons.  Other than that, the ads focus on the firms’ investment philosophy, their attention to their clients and – maybe most important in the public’s eyes – their longevity. 

The fact that people have entrusted their money to a management company for decades is a reassuring thought.  Yet my survey shows that longevity appears to be mildly negatively correlated with performance.  I mentioned the larger underperformance against the indices when averaging fund years rather than fund.  Another perspective on the same phenomenon is a 33-year average life for the bottom half of the performers, versus 24 years for the top half.  Again, transaction costs may be the source of this difference, but it becomes very hard to argue that there is evidence that experience matters.  It becomes clearer that the industry is relying heavily upon the good fortunes of the stock market itself, rather than through added value through stock selection. 

Malkiel takes the “poster children” concept further.  Of course, surviving funds could be expected to do better than ones that previously folded.  But he also points out mutual-fund companies are known to start many new funds at the same time and keep only the best-performing ones.  Since some will, at random, do better over the first few years of existence, it follows that the survivors will always start their record with exceptional performance.  His own study of data showed that survivor funds on average exceed the total fund group for a given year by 1.5%.

Enough Evidence?

We often hear the argument that the performance of the stock market throughout the 20^th Century is evidence of the durability of its investment advantage.  The SBBI Yearbook uses this same argument with its data covering the period 1926-2006.  The mutual funds I have surveyed cover the same period, and are yet to show significant added value.  So it we accept the idea that an 80-year period provides conclusive evidence, why should we have confidence that fund management will ever be capable of superior performance?

I do not intend to pick on mutual funds.  The reason for this focus is the easy availability of objective data.  If we can fault their performance, we cannot fault their honesty.  They have met a need that the public has demanded.  Until recently, such funds were the only way that many small investors could participate in the success of the market.  I do not think the fund-management industry “fooled” the public into believing it could beat the indices.  It is clearly every manager’s goal to beat the indices, and each will lay out his or her strategy for doing so.  I think most managers believe that they can add great value, and I am quite sure that virtually all those who end up on the positive side of the indices believe that they were responsible for getting there.  Financial journalists appear to believe that, as they publish vast numbers of league tables and interview the successful managers.  Such beliefs of the professionals are sustained by the widespread belief of the investing public that winners must exist, and that they can be found.  As we will discuss later, such a belief appears to be rooted deeply in basic human learning and survival psychology.

Beyond the hidden genius of Warren Buffett, and the transparent mediocrity of mutual funds, what objective performance resources are readily available to us?  So far, efficient-market theory appears to be standing up well.  Remember, as defined, it does not claim a rational market, or that investors will not beat the market.  As we will see when we examine sticky luck more closely, it does not even claim that some investors will not beat the market with impressive consistency.  Instead, it claims that no one can expect to beat the market.  For now, mutual funds have given efficient-market theory a very easy time of it, not even showing a blip of above-index performance that would need to be explained by the less intuitive aspects of sticky luck. 

Greatest Investors

For more challenges, I went to Forbes Media’s Investopedia, and their section entitled “Greatest Investors.”  The introduction is as follows:

“Becoming a successful investor takes education, patience and maybe even a little luck. Historically, the market has returned a solid 12% per year on average. The icons we'll present here represent the pinnacle of the financial world. Each one has dramatically exceeded market performance. They have all made a fortune off their success and in many cases, they've helped millions of others achieve similar returns. 
These investors differ widely in the strategies and philosophies they applied to their trading; some came up with new and innovative ways to analyze their investments, while others picked securities almost entirely by instinct. Where these investors don't differ is in their ability to consistently beat the market.”

There are eighteen profiled investors.  It obviously includes Warren Buffett and his mentor, Bill Graham.  However, other than Buffett, only six of the investors have specific track records quoted.  (Quotes of track records for others can be found on the internet but are unsubstantiated, and are also generally in a context that invites skepticism.)  Some six of the other people are better known for their investment writing or establishing fund-management companies.  Others are described as highly successful investors, which unquestionably they are or were, but we have no objective way of judging the quality of their actual performance.  Moreover, those with the more spectacular-seeming records – George Soros, Julian Robertson and Michael Steinhardt – are hedge-fund managers.  For that reason, I need to take a very quick detour on hedge funds. 

Hedge funds are intended to make money in any legal way possible.  Part of this may involve positions in freely traded securities and, to the extent that they do, the funds form part of the investor universe I am discussing.  However, to the extent that the funds take leveraged positions on interest rates, exchange rates, other non-stock-market-related risks, private equity and even foreign equity markets, it makes little sense to compare their performance with benchmarks like the S&P 500.   A big play with a one-time 500% return will guarantee an added 20% average return for ten years.  Soros’ “breaking” of the Bank of England in 1992 was much more about the type of skills required for poker.  Hedge funds remain interesting in the context of this book because of their explicit attempts to exploit sticky luck in areas outside the stock market.  The current Yen-carry trade could be the finest example of sticky luck in the world today, and we will touch on it later.  For the moment, however, we will illustrate the special nature of hedge funds with Michael Steinhardt, who is said to have earned an average 24% per year over 1967-1995 and so is the only non-mutual-fund manager with a specific track record mentioned.  Wikepedia includes the following entry, which I think is sufficient reason for not including him in our performance analysis:

“After decades of successfully managing the fund, Steinhardt and his firm were investigated for allegedly trying to manipulate the short-term Treasury note market in the early 1990s. He personally paid 75% of $70 million in fines as part of settling the case with the SEC and Department of Justice. His firm made $600 million on the Treasury positions.”

Of the five mutual-fund managers with quoted track records, four would not appear unusual at the better end of my survey.  Their out-performance of the S&P 500 is quoted for three of them as 2.7%, 3.1% and 4.2%, and the fourth can be calculated to show a since-inception return of 2.2% over.  The only manager’s performance period significantly exceeding the survey average – his period being some fifty years – is the 2.7%.  Not only do these results fall within the random-error range of my survey, but they are within the value-to-growth spread of all but the large caps.  It is difficult to imagine these funds not taking advantage of their flexible prospectuses, to their good fortune. 

Peter Lynch Alone

We are left only with Peter Lynch.  During the thirteen years that Lynch managed the Magellan Fund, the profile states, “Lynch reportedly beat the S&P 500 benchmark index 11 of those 13 years, achieving an annual average return of 29%.”  Since the SBBI shows that large caps earned just under 17% per year on average over the period – exceptional enough – it’s curious that the 11/13 score was even mentioned along with the 29% figure.  The reference to “reportedly” creates some doubt about this data. The Fidelity website is no assistance – perhaps they consider such history of no commercial value for a closed fund.  Several other websites curiously state, “according to a secondary source quoting Valueline…”   Though 29% is mentioned in many other websites, almost all are of the “Secrets of the Masters!” genre, which I suspect copy their figures from other non-referenced sources without any form of verification. 

Yet, based upon the available evidence, I believe that efficient-market theory should accept the challenge of something close to 29% per year for the period 1977-1990.  My first observation is that we have only Peter Lynch among mutual-fund managers in such an exceptional category.  All other contenders for such outstanding consistency do not provide us with objective, auditable data and are also free to place large bets well outside the quoted-security environment.  One response to this comment may be that such extraordinary talent would not normally be content with mutual-fund management, so we could not hope to find the best among this group.  But I think that argument is implausible.  Even assuming hedge-fund management offered vast additional earnings that a manager thinks he or she could use for a happier life, there is no reason to assume that investment skills are always accompanied by equal greed – this is not the case for any other human skill.  If excellence in mutual-fund management truly exists, there would surely be many individuals who would seek to be the best in that area, perhaps because of its distinctive skill-sets, visibility and public appeal. 

This odd fact should lead us to question whether Lynch is the proof of how the stock market can be consistently beaten, or if it points to some other phenomenon.  Another fact to consider is how short Lynch’s career as a fund manager was – a mere thirteen years compared with the thirty years or more of most of the others.  I think it’s helpful to review his own thoughts on his art, and three quotes given in his profile are illustrative:

"Absent a lot of surprises, stocks are relatively predictable over twenty years. As to whether they're going to be higher or lower in two to three years, you might as well flip a coin to decide."

"If you spend more than 13 minutes analyzing economic and market forecasts, you've wasted 10 minutes."

If you stay half-alert, you can pick the spectacular performers right from your place of business or out of the neighborhood shopping mall, and long before Wall Street discovers them."

Lynch was simply wrong on his first point – anyone who began investing around 1960 barely kept up with price inflation during the next twenty years.  In fact, twenty-year cycles were about to make a dramatic upswing around the time of his retirement.  But at the time it looked as if the market had just enjoyed an exceptional ride and was due for some lengthy pedestrian times, if not a temporary down-surge.  Though large caps had averaged 14-15% per year during his fund management, “value” small caps had averaged almost 22% per year.   With such wide gaps between the performances of whole sectors like this, Lynch’s 29% does not seem quite so unreachable from a sticky-luck perspective.  On the edge of recession and having no particular confidence in the next few years’ performance, perhaps Lynch was concerned about sticky luck running out. 

But, to his credit, the last comment quoted above may be the most helpful.  Lynch claimed that anyone could spot the small, value-type business that was due to explode into wealth.  I recall his book talking about the taste of Krispy Kreme’s coffee and his instant decision that this new franchise had to be a winner.  I am sure that the reader has experienced great service from other franchises which nevertheless failed to flourish.  Lynch’s argument that investing $1,000 in such an enterprise can yield $9,000, yet exposes you to only a $1,000 downside, is an arithmetic fantasy that most schoolchildren can see through, particularly if they watch their parents play the lottery or gamble.  I do not know if Lynch’s writings are naïve or dishonest, but I can calculate that he would have needed only 1% of his fund each year to become “ten-baggers” as he calls them, in order to bridge the gap between his reported 29% and the small-cap value index. 

Lynch’s capacity to spot emerging success in the mall would seem to go beyond luck.  As with Buffett, it is probably a business genius’s capacity to collect and fully appreciate personal data that is not shared by the market as a whole.  Experiencing the business first hand and understanding what makes it different is akin to the job of any talent scout, whether at sports, music or drama.  For a period, Lynch soundly beat the market by being able to step into the shoes of entrepreneur, employee and customer of a number of small, retail companies he studied, and at a time in US economic history when their opportunities turned out to be very favorable.  It is an extremely rare gift, but may well depend crucially upon a certain type of business is a certain business era.  One key aspect of sticky luck is that, by chance, certain individual’s simplified business models can mirror a much more complex reality remarkably well for a limited period.  However, such models are liable at any point to collapse and never repeat.  Lynch did not wait to find out about the durability of his model, but instead chose to write about it as guru. 

By Comparison, Sticky Luck is a Tsunami

We are left with virtually no historical evidence that fund managers can significantly beat the market averages over the long term, using just public information as their resource.  The averages themselves are sufficiently wide-ranging that all known performance can be explained by one of them, and the fact that managers are free to explore the full range of their prospectus strategy, no matter what Morningstar chooses to label them.   The degree of deviation from a particular average that suits the manager’s actual investment path is easily explained by random variation of stock performance within the market sector.  Note that, in fund analysis, we have had to call only on sectors classified by size and growth/value.  As shown by Eric, much greater room for sticky-luck results is possible from foreign investment, fortuitous market timing, industrial sectors and commodities.  Because of the nature of sticky luck, such differences do not average out year to year but accumulate over many years as each sector goes through its bull and/or bear cycle. 

Ironically, it may be because of such pervasive average performance that little attention has been attached to the phenomenon of sticky luck.  When efficient-market theory was understood and adopted by economists in the 1960s and 1970s, an explanation of market randomness seemed sufficient.  As the market traced its unhappy way between 1966 and 1982, it would be characterized in the popular mind as a random walk to nowhere.  But then, the popular mind was not particularly interested, and pension fund trustees accepted the men of market-science as the experts.  After the 1981 recession came the glorious one-way ticket to wealth and public access through mushrooming mutual funds and low-cost trading.  Now the public asked for better understanding, and random walks could not sell copy.  There had to be a more compelling explanation of this robust upward trajectory.  Images of Buffett and Lynch fitted the bill.  If you just did enough homework, if you could see the pattern in the mass of data, if you could perceive the hidden value that would inevitably burst forth… you could truly make it to riches.  If they could be billionaires, surely getting a little of their know-how was worth a few million?  Chartism and intrinsic-value are just techniques, to be studied and put into practice like a software-training course.  We wish.