II. Chasing Performance

Key take-aways

  1. Investors often experience returns that fall short of the funds in which they have invested, because they move in and out of funds at inopportune times, due to attempts at market timing and switching funds. They zig when they should zag and zag when they should zig.
  2. Picking “winning” funds based on recent return measures is counter-productive. While these measures contain information about manager skill, this is swamped by how the fund’s investment approach has jibed with recent market dynamics, which move in cycles, and by pure luck.
  3. A key ingredient in market cycles and the root cause of the shortfalls investors suffer relative to the funds in which they invest are one and the same thingperformance chasing.
  4. The best way to select investment funds is to seek out funds that commit to using sensible, proven investment principles and techniques, a value discipline above all. Your bias should then be to stick with them.

Investors often experience returns that fall short of the funds in which they have invested.  The graph below, drawn from data in a 2015 study,(1) illustrates this puzzling dynamic.  The grey bars show the average compound annual return investors would have had by buying investment funds starting in 1993 and, from the initial purchase of each, simply holding them to 2013.  The green bars show the returns the investors experienced based on how much cash they actually had invested in the funds, given the buys and sells they in fact made along the way.

Source: Research Affiliates using data from the CRSP Mutual Funds Database and Morningstar Direct

The graph shows that this shortfall happens across all funds and for different categories of funds—whether by investment style (e.g. growth versus value), size of company (small capitalization versus large capitalization) or by approach to stock selection (index versus active).

The chief culprit in this tragedy is our behavioral tendency to move in and out of funds at inopportune times.  Much of the shortfall is another illustration of the counter-productive effects of moving in and out of the market as we discussed in The Dangers of Market Timing.  These same type of instincts and practices also cause us to zig when we should zag and zag when we should zig in another way—switching among funds.  How so?

What Motivates Us to Switch Funds?

First, as I noted in the Taking Advantage of Time Frame section earlier, we all look to the past for insight into the way things work in human affairs; and in investing, where forecasting is very difficult, this is in fact the best thing we can go on.  But in doing so, we need to look at what happened over long stretches of time, the longer the better.  Unfortunately, in evaluating funds, we tend to look at and emphasize the recent past—specifically, how the fund has performed over the past few years.  Our inclination is to look for “winners,” those funds that have posted above average returns over the last three to five years.  More often than not, this is a mistake.

Second, when something doesn’t work out in life, our inclination is to hold someone to account.  We kick ourselves for making a bad choice, while at the same time we want to make ourselves feel a bit better by apportioning blame elsewhere.  We also might conclude that it simply would be best to make a change.  With investment funds, this, too, is more often than not a mistake.

Recent Performance Measures (Returns): Market Cycles or Manager Skill?

Some investment funds are undoubtedly better than others—by virtue of their strategy, their process or the flat-out skill of their managers.  And, of course, we all would like to find the best ones and cast away ones those we come to believe are lacking.  The problem is that recent returns—again over the last three years or even stretching to five years—are a poor measure on which to make such judgments.  Because, while recent performance measures do contain information about manager skill, this is swamped by sheer luck (chance or randomness) and something else these measures pick up—that is, how funds actually approach investing and how well this approach has worked within recent prevailing market dynamics.

Let’s elaborate.

488690711-squareStock markets as a whole move in cycles—there are bull markets followed by bear markets followed by bull markets, and so it goes. Within stock markets, there are other types of cycles.  Stocks of companies in certain sectors of the economy—industrials, financials, health care, technology, for example—do well at certain times and relatively poorly during other stretches.  This likewise ebbs and flows and ebbs and flows.  In addition to what sectors they are in, stocks are often grouped according to other shared attributes that can be quantified, commonly referred to as “factors.”  These factors or attributes include how fast companies grow or are expected to grow, how big the companies are, how the companies are valued and even how their stock prices have been changing in the market.  The stock prices of companies with such shared attributes likewise tend to move in cycles.

Let’s take a look at a sector example.

Constancy Investors analysis: Growth of $1 invested in stocks of companies in the Russell 1000 Index in the Industrial sector divided by growth of $1 invested in the Russell 1000 Index. Data source for monthly Total Return data is Bloomberg.

Industrial companies fell out of favor during the tech-telecom bubble of the late 1990s, as they were viewed as horribly “old economy.”  Their stock prices accordingly fell relative to the stocks in the Russell 1000, a broad stock market index containing the 1000 largest publicly-traded U.S. stocks, lagging behind by as much as 14% in late 1999.  As the bubble burst and tech-telecom stocks fell to earth and industrial businesses moved back to fair valuations, this relative performance reversed.  Note that Industrials, as businesses that feel the effects of the general business cycle, also tend to have lagging price performance during economic downturns, as was the case in 2008, which then reverses during recoveries.  A fund whose approach tends to result in a strong weighting in industrial stocks is going to reflect such ebbs and flows in its near-term performance, though over the longer-term they wash out.

Here’s another sector example, that shows the ebbs and flows of sector performance even more dramatically.

Constancy Investors analysis: Growth of $1 invested in stocks of companies in the Russell 1000 Index in the Financials sector divided by growth of $1 invested in the Russell 1000 Index. Data source for monthly Total Return data is Bloomberg.

Like Industrials, the stock prices of companies in the financial services industry lagged those of go-go tech and telecom companies in the late 1990s.  But then Financials soared as the housing market and the loans that allowed those home purchases took off.  When the housing bubble burst, financial services businesses suffered most severely in the ensuing financial crisis, and their stock prices cratered.  They have only gradually found a footing and started to regain ground relative to the overall market, as measured again here by the Russell 1000.

Finally, a factor example, looking at valuation as an attribute.  We specifically look at companies with low valuations (bottom or cheapest 20%) as measured by the ratio of their stock price to their earnings per share over the last twelve months, as measured each month as we move through the study period.

Constancy Investors analysis: Growth of $1 invested in stocks of companies in the Russell 1000 Index that are in the lowest (cheapest) 20% on Price/Last-Twelve-Months Earnings (LTM P/E), rebalanced monthly, divided by growth of $1 invested in the Russell 1000 Index. Data source for monthly Total Return data is Bloomberg.

Selecting stocks with low valuations is one of the few time-tested disciplines for producing higher returns, and this strategy worked very well over this 20-year period.  With the methodology used in the study graphed above, the cumulative return over these 20 years has been nearly 2.4x that of the Russell 1000, the broad market index used here.  But value stocks did terribly in the late 1990s, under-performing the index by over 40% over the span of two years, again as “old economy” stocks were abandoned in the tech-telecom bubble.

Here’s the key take-away.57432663-square

A key ingredient in market cycles and the root cause of the shortfalls investors suffer relative to the funds in which they invest are one and the same thingperformance chasing. Certain industries or stock factors move into favor due to a new development—exciting new technologies, new growth in an area like housing, new investment products like exchange-traded funds (ETFs).487487825-square  Investors plow into the stocks that are on trend and pile out of those that are not.  We enter the self-fulfilling prophecy phase: “the popular stocks are going up and the unpopular stocks are going down, so of course this is the right thing to do.”  More people plow in.  The whole thing becomes overdone.  Valuations become ridiculously expensive for the popular companies and ridiculously cheap for those left by the wayside.  Then things reverse.  Because valuation ultimately matters.  It is an irresistible force, like gravity.

Note that the focus on near-term performance, especially in the self-fulfilling prophecy phase, is an accelerant in this ultimately self-defeating chain reaction of performance chasing.  So, let’s return to the topic of near-term performance.

Passive funds (whether mutual funds or ETFs) follow set, predetermined criteria and rules—the great bulk simply buying the stocks in the designated index they track.  If an index fund’s performance declined in recent years, it would be absurd to attribute that to a loss of the fund’s skill.  An index fund is a mechanical, administrative exercise; and the fund has continued to do exactly what it has always done.  So, rather, we should conclude that the potential returns for the index had simply dried up.  And this likely happened because the general stock market environment had previously favored the index fund and it had grown popular, such that the prices of the stocks in the index became so high that their expected, forward return had been eroded.  By way of example, take a look at any of the technology ETFs that were born in the late 1990s and doubled or tripled in a year or two only to plummet 85% over the next two years.

What is often lost on investors is the fact that the same dynamic very much applies to actively-managed funds as well.  Most active funds adhere to some investment strategy or discipline which they describe up-front.  For example, an active fund managed according to a “value” philosophy may follow the discipline of limiting its investments to stocks that trade at low Price-to-Earnings (P/E) ratios, similar to the factor example we just looked at.  If the general market environment over several years is dominated by investment flows into popular stocks with strong price momentum, those stocks with low P/Es—typically an indication of unpopularity—are on balance going to under-perform momentum funds and the market, as was the case in the late 1990s.  It may well be that a value fund manager made mistakes in this period that weighed on returns, but we also must recognize that the market environment made it highly unlikely that that fund was going to produce a return close to that of the market and momentum strategies.  The fact is, however, many value managers were fired and value funds shuttered by early 2000.  Note well that the dominant cause of the fund’s under-performance—sticking to the fund’s advertised discipline of buying inexpensive stocks—was the seed of the fund’s future strong returns.

Pulling all this together, take a look at the graph below, which looks at all the mutual funds (active and passive) investing in U.S. stocks over the 10 years from 2005 to 2015.  It separates funds into those that delivered above average returns over the first five years, from 2005 to 2010, and those with below average returns over this period.  We then look at the funds’ returns over the subsequent five years, from 2010 to 2015.

Constancy Investors analysis: source data is Bloomberg, FSRC mutual fund screening functionality.

The study shows that a significantly greater percentage of the “below average” funds over the first five years delivered above average returns over the next five years than did the “above average” funds.  This is true whether we look at all US stock funds as a whole or group them by investment style (blend, growth, value).  In each case, around 60% of funds that lagged in the first five years delivered above average returns in the subsequent five years.  In each case, less than 50% of the “winning” funds from the first five years did so in the next five.  In summary, recent performance data may say something about manager skill, but they say more about recent market dynamics. 

Untimely Fund Switching Has Been Institutionalized

Inopportune timing in buying and selling investment funds is not just a “rookie mistake.”  Both individual investors and professionals serving pension plans and other institutions show shortfalls in returns relative to the funds in which they invest.  As crazy at it seems, the self-defeating behavioral tendencies that cause the shortfalls—“picking winners” and sharing blame, the twin pillars of performance chasing—have actually been institutionalized.

Individual investors find information about investment funds in financial publications in print and on-line, in fund advertising, at their on-line brokers and on websites dedicated to rating investment funds.  A staple of all these sources is the current performance of funds over the last 1, 3, and 5 years.  Fund rating services offer ratings on a 1-5 scale, with such ratings determined largely on a formula that weights 3-year, 5-year and 10-year returns.  In the institutional world, professionals and their consultants access more sophisticated subscription-based databases, but these database providers note that 3-year and 5-year performance are among the most searched criteria.  All of this serves to bait investors into picking winners.

522486519-squareThe “blame game” may actually be exacerbated in institutional settings.  Individual investors today often have a financial planner or advisor that helps them put together an overall plan, allocate dollars among different asset classes and select investment managers and funds in these asset classes.  Institutions like pension funds nearly always have such advisors or consultants. If the ultimate investor, i.e. whose money it actually is, becomes unhappy or just nervous with how things are going, the easiest thing an advisor can do is suggest a re-shuffle in investment funds or managers. The ones that have provided lower returns of late are the obvious ones to put out into the dog house.

So, how should we assess investment funds for purchase or sale?

Style Boxes: Still Very Imperfect

In the realm of performance, first, it is much better to look at investment funds versus their closest peers, as opposed to general market indices like the S&P 500 and Russell 1000.  Ideally, you would examine funds setting out to do exactly the same thing, and look for the one that does it best.  Unfortunately, this can be a challenge, especially for individual investors.  You do have free access to limited information, including ratings, on sites that do group funds into asset classes and further into “style boxes” that look something like this.

Like any construct that tries to categorize things in general buckets, the “style box” approach is very imperfect.  For example, the criteria used by the raters to group value investors often use a single measure of value, perhaps Price to Book Value.  A value manager that emphasizes a different measure, like Price to Earnings, or that considers a number of measures will often select very different companies.  Similarly, the raters often use a low revenue or earning growth criteria as an attribute of companies that should be in a value portfolio.  Some value investors do simply consider low valuation multiples outright, but others think about the valuation multiple relative to other financial attributes, including profitability or growth rates—high profitability or growth rate is viewed favorably, while still requiring a low valuation.  Again, the companies selected for investment can be quite different.

In the graph below I again show all the funds from the study we presented just above, but now grouped according to the style boxes into which they’ve been placed.  (The source data, which is from Bloomberg, includes two additional buckets for blend and value called “broad market,” which are funds that invest in stocks across the entire market capitalization spectrum.)  The rest of the mechanics of the analysis remains the same.

Constancy Investors analysis: source data is Bloomberg, FSRC mutual fund screening functionality.

When we group funds by style boxes, we see that selecting funds based on recent returns performance is basically a coin toss.  In six of the 11 categories, a higher percentage of the funds with below average returns in the first five years went on to post above average returns in the next five years compared to the “winning” funds from the first five years.  In five categories, the opposite was true.

In short, style box groupings necessarily use a broad brush, which results in lumping together funds that are not truly peers.  As a result, variations in recent performance of funds in the same bucket may again say as much or more about the recent market environment than about the relative skill of the managers.

Institutional investors have the resources to pay to access databases, as mentioned earlier, on investment funds and managers that have very detailed information on how the funds approach investing and on their return figures.  While these databases often provide similar style boxes, the more granular information provides the opportunity to make finer distinctions about the comparability of funds and managers.  Even so, this is not easy, and studies show sophisticated investors’ returns on balance also suffer from inopportune switching, though not to the same degree.

A Sound Basis for Selecting Funds

In evaluating performance, it helps to look at long stretches of time.  This helps to judge consistency.  Not in the sense that the fund always or regularly outperforms its peers or the market, but rather that there is a discipline being adhered to rigorously.  Unfortunately, such data can be hard to find.  Moreover, funds come and go.  And even funds that have operated for decades may have undergone change, including in personnel.

Given all the challenges in simply looking at performance (return numbers), the best way to select an investment fund is through good old-fashioned judgment.  First, you should seek out funds that commit to using sound, proven investment principles and techniques, a value discipline above all.  In the tab at left, A Sensible Approach, I elaborate on what I believe are the most effective of these principles and techniques.  Second, you should seek out funds whose words and actions demonstrate that they place your interests ahead of their own, including reasonable fees.  When you’ve found funds that make sense, stick with them.  When you select funds with shared beliefs in what works in investing, you can think of that as a partnership or a marriage.

Under the Invest tab, I provide details on an investment approach that I believe is particularly compelling, as it synthesizes the best of active and passive investing and puts clients first.

(1)The data is drawn from Table 2 of an excellent paper titled “Timing Poorly: A Guide to Generating Poor Returns While Investing in Successful Strategies.” Jason Hsu, Brett W. Myers and Ryan Whitby.  May 2015.  The source data for Table 2 in their comes from Research Affiliates using data from the CRSP Mutual Funds Database and Morningstar Direct.