WHAT WE HAVE LEARNED, KNOW
AND HOLD DEAR
Q2 2011
.
Traumatic historical events have a way of jolting us into questioning deeply held beliefs. The All Saints’
Day earthquake of 1755, which killed tens of thousands of Lisbon’s poorest residents, inspired European
philosophers such as Voltaire to reject the idea of an omniscient, omnipotent, and benevolent deity.
More recently we have seen how wars, natural disasters, or terrorist attacks can lead to sharp
turnarounds in public opinion and political preferences.
It is only natural that the ’07‑09 financial crisis has prompted a wave of intellectual ferment in the
investment industry. The newsletters and pitchbooks that cross our desks these days read like obituaries
of onceâ€popular investment concepts; active management, strategic asset allocation, diversification, and
mean reversion have all been pronounced dead. Meanwhile, new ideas and products—some of them
new only in their packaging—have sprouted up to replace the old.
We welcome the introspection that is taking place in our industry. An important part of our process
when we evaluate a money manager or design a portfolio for a new client has always been to lay out
our ideas unambiguously in writing. It is this discipline that forces us to commit to a point of view and
creates accountability. In that spirit, we would like to discuss some of the core principles that we follow
as investors.
1. Strategic Asset Allocation, the starting point for a longâ€term portfolio, is insufficient
The premise behind Strategic Asset Allocation (SAA) is that investors are compensated for being longâ€
term holders of risky asset classes. In the short term, the risk premium for owning a single type of risky
asset can be volatile—that is, in any given year the stock market may underperform cash, sometimes by
a lot. However, SAA holds that investors can reduce volatility by combining asset classes that respond
differently to economic conditions, such as stocks, bonds, and cash. For most longâ€term investors it thus
makes sense to begin with a plan that specifies target weights for each asset class and a disciplined
process for rebalancing one’s holdings as asset prices fluctuate.
Before 2008, SAA was almost universally accepted among large, professionally run pools of capital as the
best way to build a portfolio. Since then it has come under attack by critics who argue that
diversification failed to protect investors from losses during the financial crisis. As evidence, they point
to the dismal performance during late 2008 of many of the diversifiers that investors added to their
portfolios in prior years: private equity, real estate, infrastructure funds, commodity futures, and hedge
funds to name a few. These asset classes had appeared to be uncorrelated to public equity markets
before the crisis, but moved in lockstep with stocks at precisely the time when investors hoped they
would diverge.
As we wrote in this space over two years ago, our own perspective on SAA echoes Winston Churchill,
who famously quipped that “democracy is the worst form of government, except for all the others that
have been tried from time to time.” As with democracy, poor implementation can lead to bad results—it
should be no surprise that the illiquid, equityâ€like asset classes that many investors added to their
portfolios at the height of the Great Moderation1 performed poorly in 2008 and 2009. Defensive
positions such as high quality fixed income performed reasonably well during the crisis. Unfortunately
1
The Great Moderation refers to the reduction in the volatility of the business cycle that lasted for much of the 1980s and
1990s.
. many investors had cut their exposure in the midâ€2000s, enticed by the prospect of higher returns
promised by more exotic investments. In most instances, it was a poorly designed asset allocation
backed by inadequate risk management and reinforced by greed (which later turned to fear) that was
the main culprit for investors bad experience.
The obvious alternative to SAA is Tactical Asset Allocation (TAA), which relies on an advisor’s ability to
predict shorterâ€term movements in asset prices, e.g. bonds will outperform stocks next week. We have
no great objection to this approach when used in moderation. We have occasionally expressed tactical
views in our clients’ portfolios. At times this has been implemented organically such as when our equity
managers sold nearly all of their financial stocks by the end of 2006. Other times, we have introduced a
new asset class when the benefits were clearly beneficial to the overall program as we did in 2009 with
Master Limited Partnerships or in early 2010 with emerging market local debt.
Where we disagree with TAA is on the weight that shortâ€term shifts should carry within a longâ€term
portfolio. Few investors, if any, have the ability to predict asset class movements consistently. The
spotty track record of even the smartest and most celebrated global macro managers is a testament to
this. For the rest of us to think we can do better is hubris. TAA decisions should, therefore, be modest in
size and limited to periods of extreme market dislocation.
2. Markets are not perfectly efficient—but they are tough to beat
To paraphrase decades of academic research in a few sentences, proponents of market efficiency argue
that security prices at any point in time fully reflect all relevant public information. If markets always get
prices right, then active managers have no inefficiency to exploit and it follows that one should hold an
indexed, marketâ€weighted portfolio rather than paying to try to outperform the market. It also follows
that asset price bubbles should not exist, or at least can’t be identified before they burst.
The apostles of efficient markets often argue that all of the academic research is on their side. For
example, for years we have heard the claim that one’s asset allocation policy explains over 90% of a
portfolio’s return, the implication being that active management has a negligible impact on
performance. The source of this myth is “Determinants of Portfolio Performance,” a seminal 1986 paper
by Brinson, Hood, and Beebower which regressed market returns against a number of funds and found
that almost 95% of the funds’ returns over time could be explained by market movements2. Yet this is
not the same as saying that asset allocation can explain 95% of the difference in performance between
one investment program and another. Roger Ibbotson, one of the founding fathers of asset allocation
research, has pointed out this distinction in a series of articles and performed his own studies to answer
this very different question. His results suggest that asset allocation policy is important, but is
responsible for only around half of an investment program’s return. The combination of the money
managers’ active bets, including security selection and timing decisions, accounts for the other half.
It may seem odd that the efficient markets theorists still have any credibility after a decade that saw two
asset bubbles inflate and implode, with devastating consequences for investors and the broader
economy. Yet their ideas continue to exert a powerful influence on the investment management
industry. Proof of this is the growing popularity of indexed, market capitalizationâ€weighted investment
2
Funds in this context refer to pension funds not individual money managers.
. products. With many active portfolio managers having failed to distinguish themselves in the 2008/2009
crisis, investors have begun to ditch active management altogether and implement portfolios with
cheap, indexâ€based Exchange Traded Funds (ETFs).
Since the 1970’s, conventional wisdom assumed axiomatically that market cap weighted indices were
the most efficient means to gaining market exposure. Yet there is a problem with these indices. Because
a security’s weight in the index is driven by its price, a market capâ€weighted index can be distorted by
the greed (2007) and fear (2008) that tend to dominate most people’s investment decisions. For
example, at the height of the tech bubble in 1999, technology stocks accounted for 30% of the S&P 500.
Five years later their share of the index had declined by half. In 2006, financials were over 22% of the
index—by the first quarter of 2009 this number had fallen to 11%.
Most recently, as investors seeking fixed income exposure flock to issuanceâ€weighted index products,
they face the risk of buying funds that are laden with the debt of some of the world’s most precariously
leveraged companies and governments. For example, at the moment 25% of the iShares International
Government bond ETF is in Japanese debt, while an additional 22% is in the bonds of Portugal, Italy,
Ireland, Greece and Spain—the notorious PIIGS.
This major flaw associated with market capitalization weighted passive equity funds and issuanceâ€
weighted passive bond funds is one example of the imperfections associated with the popular myth that
markets are efficient. Implementation of academically supported theories often leads to suboptimal
outcomes when they are not critically examined. These examples illustrate the ways in which a popular,
academically validated approach like index investing can lead investors astray if it is not overlaid with a
strong helping of common sense.
3. A skilled manager with the right incentives can earn his fees by producing high riskâ€adjusted
returns
There is a vast body of academic literature on the topic of mutual fund performance over the past 50
years, and most of it points to the same conclusion: the average mutual fund manager is no better at
picking stocks than a monkey throwing darts at an issue of Barron’s. Here is how the empirical research
is summarized by Mark Rubinstein3: “The continuing performance of these funds…should not simply be
put on one side of the ledger and given equal weight as any market anomaly on the other side. In
fact…the behavioralists have nothing in their arsenal to match it; it is a nuclear bomb against their puny
rifles.” In other words, the evidence of poor active manager performance is insurmountable despite the
handful of studies that suggest the capital markets are not entirely efficient.
So why do we insist that it is possible to find skilled active managers who can beat a passive market
index over the long run? After all, mutual fund managers are professional investors—the soâ€called
“smart money.” That the vast majority have failed to beat the market should convince anyone to give up
on active management, right?
But what if mutual funds are exactly the opposite of smart money? This is what Owen Lamont and
Andrea Frazzini posit in a paper entitled “Dumb Money: Mutual Fund Flows and the Crossâ€Section of
Stock Returns4.” The authors do not question mutual fund managers’ talent. Instead, they point out that
3
Rubenstein is a professor in the Master of Finance program at the University of California, Berkeley who was instrumental
in creating portfolio insurance—a financial innovation that many blame for causing the 1987 stock market crash.
4
Owen Lamont and Andrea Frazzini, “Dumb Money: Mutual Fund Flows and the Crossâ€Section of Stock Returns,: Journal of
Financial Economics, 88 (2008): pp.299â€322
. the average mutual fund customer is a retail investor who selects funds based on recent performance,
and this performanceâ€chasing, in turn, influences the behavior of money managers for the negative. In
practice, it appears that individual investors drive the behavior of mutual fund managers, making them
much less smart than we might suppose. This dynamic is commonly referred to as investing by looking in
the rearâ€view mirror. Yet while no reasonable individual would drive their car in such a manner,
investors invariably chose this method when it comes to investing.
We saw this phenomenon at the market peak in the dot.com era. After a meteoric rise in stock prices
for technology companies, mutual fund investors sought the best performing money managers that
gave them maximum exposure to this industry. Even if the portfolio manager was convinced that his
portfolio was overvalued, shifting into another sector (say energy) or selling his high flyers (Cisco, Dell,
Pets.com) would have slowed the fund’s asset inflows and his own remuneration. The portfolio manager
knew that investors were buying his fund because of those names. In fact, Frazzini and Lamont estimate
that by the end of 1999, 18% of Cisco shares were owned by mutual funds and 3% of these shares were
attributable to recent fund inflows. More recently, we saw that investment banks sold securities backed
by risky mortgages not because they thought these were good investments (in fact it is being revealed
that they knew they were bad investments) but because clients demanded them.
In addition to the rearview mirror approach, we believe that there are a host of other institutional and
behavioral factors that create market inefficiencies. A talented money manager can exploit these
situations to his or her advantage:
-
Institutional constraints on ownership of certain securities (for example, the requirement that
many institutions immediately sell securities that they acquire as part of a spinout);
-
Overreliance on the opinions of ratings agencies that are illâ€equipped and/or too conflicted to
provide high quality analysis;
-
Boom/bust cycles, in which investors become irrationally exuberant when prices are rising and
overly pessimistic when prices are falling. Excessive use of leverage causes market prices to
overshoot on the upside (investors borrow to pile into the “hot” position or asset class) and also
on the downside (investors carrying too much leverage are forced to liquidate positions in a
firesale).
All of these factors create opportunity for skilled active managers. However, skill alone is not a sufficient
condition for an active manager to outperform the market consistently, organizational factors and
incentives are also important. Unfortunately, most investment managers do not operate in an
environment with the right incentive structure for producing consistent outperformance. As a study by
Cohen, Polk, and Silli entitled “Best Ideas”5 found, there are signs that the top holdings of the average
mutual fund manager do consistently outperform the market, even though their overall portfolios do
not. This suggests that the problem with most mutual funds is excessive diversification—they own a
large number of positions in which the manager may have little or no conviction. Most managers feel
pressured to diversify, because a highly concentrated portfolio may perform very differently from the
market or index and scare away existing and potential investors.
Likewise, the conditions for continued success by even the most talented investors frequently change.
5
Randy Cohen, Christopher Polk, and Bernhard Silli, “Best Ideas” MIT, LSE, Goldman Sachs, April 20, 2010
. Successful money managers evolve, sometimes to the detriment of their clients’ wealth. They become
victims of their own success as they seek to empire build (add new funds), dilute their own contribution
to the investment process (add too many people), or they simply choose to enjoy the fruits of their labor
(spend more time travelling and playing golf). In any of these cases, they may be taking their eye off the
ball as the skill they previously exercised is diverted to other activities. Understanding a money
manager’s motivations and conditions that might lead to good performance are nearly as important as a
manager’s talents.
It turns out that Rubinstein’s “nuclear bomb” of evidence for market efficiency is not that hard to
explain considering the behavior of performanceâ€chasing investors and assetâ€gathering money
managers. Just as one does not look at statistical NBA averages to know that basketball players such
Michael Jordan, Magic Johnson, or Dirk Nowitzki exist, we do not require an academic study to confirm
the existence of talented and focused active money managers who produce strong riskâ€adjusted
performance for a considerable period of time. Just as professional sports teams engage talent scouts to
discover the next Lionel Messi, Lew Alcindor, or Derek Jeter, our challenge is to find the small subset of
individuals who are committed to producing great performance rather than pandering to a backwardâ€
looking and short termâ€oriented customer base.
4. Risk is more than just a number; risk management must be contextual, qualitative and forwardâ€
looking
Nearly three years after the market crash, even casual observers recognize that there were gross failures
of risk management throughout the financial industry, particularly in the big banks. One of the banks’
favorite risk management tools was Value at Risk, or VaR, a statistical measure which sought to quantify
the maximum amount a portfolio could lose on a given day assuming markets behaved in line with
recent history. Today, it is widely accepted that such a reductive and backwardâ€looking approach to risk
contributed heavily to the crisis. We are all much more aware of concepts such as fat tails, black swans,
and the nonâ€normality of market return distributions.
Old habits die hard, and our industry continues to resist Einstein’s dictum that “not everything that
counts can be counted, and not everything that can be counted counts.” The widespread use of an
investment’s volatility as a measure of its risk is one example. Massive, permanent drawdowns are our
concern; we believe that it is more important for investors to focus on the factors that could lead to a
permanent loss of capital rather than on the likelihood that an investment’s price may fluctuate widely.
This typically requires us to plan for scenarios that have no historical precedent. Unfortunately,
investors spend too much time trying to protect portfolios against a repeat of past events and not
enough thinking critically about what may happen in the future. We see signs of this in the way that
portfolios have been revamped in response to the events of 2008/2009, with large amounts of capital
flowing to Commodity Trading Advisors (CTAs), “Black Swan” funds, and principalâ€protected structured
notes, all of which performed well in the crisis.
While we don’t deny the wisdom of market thinkers like Nassim Taleb, who popularized the phrase
“Black Swan” and saw the 2008 crisis coming, we do question the timing of those investors who have
. only embraced these strategies since 2009. As with any other investment, the key factor is price. Are
there reasons to believe that portfolio insurance is priced cheaply these days? Given the heavy flows of
capital into soâ€called “Black Swan” and “tail risk protection” funds, we suspect that investors are
dramatically overpaying for them. Compounding their error, many of these investors, content with their
“insurance” have confidently ratcheted up risk elsewhere in the portfolio as they exhibit a classic “have
one’s cake and eat it too” mentality. In short, having purchased expensive fire insurance, investors are
letting their kids play with matches. Such behavior rarely ends well.
A recent piece by Financial Analysts Journal editor Bob Litterman points out the irony of investment
banks selling insurance against equity tail risk to institutional investors. “One might expect,” he argues,
“that investment banks, with high leverage and quarterly earnings reports to worry about, would be the
natural buyers of such insurance and long horizon investors the natural sellers.” Yet both institutions
and individuals (in the form of structured notes) have been enthusiastic buyers of tail risk protection
from banks over the past two years. Because volatility has plummeted over that period, this has been a
very lucrative business for the banks. As Warren Buffett once remarked, “If you've been playing poker
for half an hour and you still don't know who the patsy is, you're the patsy."
We use recent behavior only by way of example. We are suggesting that risk management is
considerably more difficult than looking at past events especially those that occurred most recently.
There is a reason that the investment banks are selling tail risk protection products these days. Much as
the price for hurricane insurance skyrockets after a hurricane hits, buyers are paying exorbitant prices to
insure against another 2008/2009â€type event ignoring the likelihood that the next major event will
unfold quite differently. What if the next shock to the financial system stems from the US Treasury
market not the stock market? It would be the equivalent of buying flood insurance after a recent flood
only to find that an earthquake or fire is the real danger. Yet we find that investors do not consider the
risks of a fire or earthquake until it is too late. Likewise, no amount of modeling, VaR, or credit rating, all
of which rely on recent historical data will tell you what the future will hold. For this, one requires
imagination and good judgment.
5. There is no substitute for judgment
A common thread running through many failed investment approaches is the idea that one can apply a
mechanical, rulesâ€based technique to making decisions. Some of the major financial catastrophes in
history occurred during periods when large numbers of investors simultaneously adopted similar risk
management formulas: portfolio insurance in the midâ€1980s or the use of stopâ€loss orders in early 2010
are two examples of this phenomenon. While the evidence in favor of these strategies was statistically
sound at the time they were developed, in each case they did not account for the new risk created by
virtue of the strategy’s early success and eventual broad popularity. The Flash Crash is a classic example
of risk management theory gone wrong. On May 6, 2010 at 2:45pm EST, the Dow Jones Industrial
Average fell 998.5 points in a matter of minutes. Thousands of investors who employed stopâ€loss orders
sold their stock positions at or near the market lows before the market regained nearly all of its losses.
What was meant to be a risk mitigation instrument failed miserably as it became a loss realization tool.
A century ago, Economist editor Hartley Withers noted the weakness of mechanical decision making
. when he sought to explain why the Panic of 1907 ravaged the US banking system6, where tight
regulation had caused the average bank to maintain a capitalâ€toâ€assets ratio of 17â€18%, while passing by
the lightly regulated City of London, where the ratio was less than 10%. “Good banking is produced”
Withers wrote, “not by good laws but by good bankers. Just as the most carefully planned constitution
will inevitably break down if the men at the helm of government are incompetent or dishonest, so no
skillfully devised banking system will make banking good, unless the banking is conducted by straight
and able managers.”
We could not agree more with Mr. Withers. Laws, rules, regulation, and mechanical processes will not
prevent the next market disaster. One important lesson that we take from financial booms and busts is
that markets are heavily influenced by our own actions as market participants. George Soros’ writings on
reflexivity capture this idea well. Soros argues that markets are inherently unstable because of their
dynamic nature. As he explains “Reality helps shape the participants’ thinking and the participants’
thinking helps shape reality in an unending process.” Unfortunately, this means that we can’t rely on
timeâ€tested rules to protect us from future risks, because the nature of risk is constantly changing as
investors react to events.
While it may be uncomfortable to acknowledge, investing is about judgment, not about models and
rules. One of our more popular catch phrases is that “there are no bad bonds, just bad bond prices.” It is
judgment that is most needed in deciphering whether a bond, asset class, or money manager is worth
the investment. No investment or money manager is born inherently good or bad. Assessing judgment is
not easy, but it is a responsibility that can’t be ignored. Toward that effort, one cannot shorten the time
spent interviewing money managers and memorializing each interaction in writing. We believe that the
only way to assess their judgment is by understanding the rationale behind many different decisions
over a variety of market environments. To paraphrase Hartley Withers, good investing is produced not
by good risk models but by good investors.
“The only things we have to fall back on are intrinsic value, company survival and our own staying power
as investors. These things mean we have to make judgments about what the future is likely to look like.
That requirement, in turn, means nothing can be approached with complete safety or certainty.”â€
Howard Marks
As in many of our national policy debates, we often find ourselves arguing against the extremes when
discussing the path to building a successful investment program: strategic asset allocation versus market
timing, active versus passive management, quantitative modeling versus qualitative analysis. What is
almost invariably lost in the discussion is the single most critical input: common sense judgment. It is
this judgment that provides us the freedom to employ elements of Strategic Asset Allocation while
recognizing and mitigating its flaws. It is our judgment that supports the occasional use of Tactical Asset
Allocation. It is our judgment, arising from our knowledge and experience, that gives us the conviction to
appropriately and prudently deploy our money managers even when performance is diverging sharply
from a market index. Employing conception and judgment, we attempt to draw informed conclusions
with regard to the riskiness of markets, strategies, and managers—sometimes in direct contradiction to
backwardâ€looking statistics.
6
With Basel III and the Frankâ€Dodd Act requiring US banks to hold more capital, we find comments by our regulators and
lawmakers suggesting that this will prevent any repeat of 2009 to be shortâ€sighted at best and dangerous at worst.
. “…the most important but the most difficultâ€toâ€identify ability in business management (or investment)
is the ability to judge other men’s ability to judge. Economists dislike the notion of judgment not only
because they have no way of verifying that it is not actually luck but also because it limits economics. If
good judgment is a driving force of capitalism, then economics as modern professors wish to practice it
can’t explain capitalism. Economics want to be physics.” – Andrew Redleaf in his book, Panic: The
Betrayal of Capitalism by Wall Street and Washington
Much to the consternation of economists and academics everywhere, judgment can neither be defined
nor measured. By extension, investing is art, not science and great investing begins and ends with sound
judgment. We recognize that our philosophy can be difficult to accept as it can tilt against conventional
wisdom, simple heuristics, Nobel Prize winning academic studies, or the math and science that might
suggest a different path. Investing cannot be modeled or tightly controlled like a physics experiment.
Investing is hard work and never ending. While we pledge to continue our hard work, we believe that it
should be on the basis of our judgment that we ourselves should be judged.
.