The standard deviation of batting averages has gone
down, says Mauboussin. If Ted Williams was a
4-standard-deviation event in 1941, then he would have hit .380
in 2011. Awesome, but still less.
In investing the paradox of skill means that many managers
are producing similar results. In other words, alpha is going
down. Mauboussin refers to the writings of Charley Ellis:
He said that before the 1970s we institutions competed
against mom and pop. You can get the money from them all day;
my positive alpha is an individuals negative alpha.
If Ellis was starting to see the paradox of skill play out five
decades ago, I was convinced that alpha might barely be kicking
Mauboussin says that from 1967 to 2012 the standard
deviation of excess returns for U.S. large-cap mutual funds has
declined from about 16 percent to less than 8 percent (see
chart). He explains that investors should go after areas of the
market in which they can be the most skillful (remember all
those institutions playing against individuals in the middle of
the last century). If standard deviation declines
meaning many managers investment results are about the
same the game is over.
There has been a steady decline in standard deviation
in excess returns exactly as Stephen Jay Gould predicted for
baseball, Mauboussin says. People will grind toward
peak performance, and the difference between the best and the
average is narrowing, so the results get really
The elusiveness of alpha cuts to the core of capitalism
itself. Good ideas whether the iPhones touch
screen or convertible bond arbitrage get copied.
Its creative destruction, says AQRs
Asness, adding that good ideas actually remain good ideas for
far longer in finance than in other industries. In part,
thats because it is harder to tell whether the good idea
is the result of a brilliant insight or just luck.
There was a lot less creative destruction of good investment
strategies when individual investors still ruled the markets.
In the past 50 years, Ellis says, volume on the New York Stock
Exchange has increased more than 2,000 times, with exchanges
around the world experiencing similar growth. At the same time,
the balance of trading in the U.S. has gone from 90 percent
individuals to about 90 percent institutions, according to
Ellis. And not just any institutions: professional mutual fund
companies, hedge fund firms and others earning high fees that
could then be plowed into technology, research and expensive
analysts to follow companies and trends. The U.S. mutual fund
industry alone went from $3 trillion in assets in 1995 to $15
trillion at the end of 2012. Foreign investors also have
entered markets as cross-border capital flows have been
encouraged and investors diversified out of their home
countries. These investors have been able to arbitrage away
inefficiencies that existed among markets that previously were
entirely separate. Stock picking itself has become
professionalized. In 1987, 15,500 people held the designation
of chartered financial analyst from the CFA Institute; both
Ellis and Duncan point me to the fact that there are now
110,000 CFA holders worldwide.
Alpha has fallen victim to technology. In the 1990s the
growth of personal computers, telecommunications networks and
the Internet drove stock market returns as investors flocked to
companies in those sectors. But the still-nascent technologies
had a much more permanent impact on the way investors
discovered information about securities. The Internet made
once-arcane sources of data available to everyone. The rate at
which valuable information could be found, disseminated and
acted upon rose exponentially. An analyst could no longer look
smart simply by watching Wal-Mart Stores parking lots for
clues about trends in discount retail sales.
In 1996 the stage was being set for high frequency trading,
another dangerous thief of alpha. That year regulators enacted
order-handling rules designed to introduce fairness after the
Nasdaq Stock Market price-fixing scandal. The rules created
electronic communications networks (ECNs) that could publish
stock prices alongside exchanges. In 1999, Regulation of
Exchanges and Alternative Trading Systems went live, allowing
ECNs to act as markets without having to register as exchanges.
Two years later high frequency traders got their shot when the
exchanges began quoting prices in decimals rather than
fractions. Bid-offer spreads shrank, and competition increased
dramatically, as lightning-speed traders moved in and out of
transactions, picking off tenths of pennies.
The more integrated fewer trade restrictions
and capital controls, supportive technology, to name a few
the greater the efficiency of the local market,
says State Streets Duncan. High frequency traders have
forced money managers to implement expensive trading operations
to protect the alpha in their good investment ideas from being
picked off by more-nimble computer algorithms.
In 2000, Yales Swensen published Pioneering
Portfolio Management, which laid out his thesis for a
portfolio that looked entirely different from the general
template 60 percent equities, 40 percent bonds
traditionally used by institutional investors. Swensen extended
diversification to include private equity, hedge funds, real
estate and commodities.
The time was ripe for Swensens method. The dot-com
bubble had burst, and investors were feeling the effects of
real-time information on alpha in the public markets.
Regulation Fair Disclosure was enacted in 2000, prohibiting
public companies from releasing information to select groups
like Wall Street analysts. By then Harvard, Princeton
University and Stanford University were using Swensens
model; soon other big investors were copying it. Allocations to
alternatives, as well as for investments in emerging and
frontier markets, grew. According to State Street and
Tufts Fletcher School, pension funds in Organization for
Economic Cooperation and Development countries raised their
allocations to alternatives from 6 percent to 19 percent of
total assets under management between 2000 and 2012.
Investors using the endowment model didnt fare well in
the credit crisis. Correlations went to 1, meaning every asset
class dropped at the same time. Many endowments that had jumped
on the Swensen bandwagon had to issue bonds to generate cash
even as the public markets were reeling. But the troubles with
the endowment model had little to do with the validity of
Swensens thinking on portfolio diversification.
Its an idea thousands of years old, says
André Perold, co-founder and CIO of HighVista
Strategies, a Boston-based firm that manages $3.6 billion in
endowment assets for institutions and private investors.
Chinese merchants famously divided cargo among several
vessels when traveling dangerous rivers.
But now there are more vessels traveling the same dangerous
river. During the five years since the financial crisis, the
vast majority of endowments have lagged the performance of a
simple, traditional 60-40 portfolio. Perold, a professor
emeritus of finance and banking at Harvard Business School,
says diversifying into foreign equities, real estate and
commodities didnt reduce volatility; it actually
The problems had everything to do with the fact that
everyone had been copying Swensens investment moves in
hopes of achieving a similar outcome. But everybody dancing the
same dance was merely a symptom of something larger going on in
the investment world. Alpha was dying, and investors were hot
for an edge. Ellis cheekily alleges that most of the copycats
jumped right to the end of Swensens 2000 book the
how to section skipping the first ten
chapters on ways to protect principal and build an approach
suitable to their institutions.
Perold takes me through his thinking on
correlations-as-1-risk. Alpha can exist in a highly correlated
world, but active investors need to hedge out the macro risks
to reduce the correlations. If you werent able to
hedge your macro risk, it made alpha generation hard, he
explains. Most active investors arent that good at
that. So when hedge funds make a bet in a world where
correlation is high, they are just taking on more risk than
they otherwise would. Their results then are much more random.
Perold argues that alpha isnt really dead, but if
investors cant identify managers with the skill to find
it, it might as well be.
Looking back at the numbers, market return beta, not
alpha has been declining in certain asset classes for
some time. According to State Street and the Fletcher School,
the mean return for both U.S. and non-U.S. equities was more
than 11 percent between 1973 and 2010. For government and
nongovernment global bonds, the mean return was 8 percent. But
the story is different for the period between 1998 and 2012:
Average annual returns for the S&P 500 and World Equity
indexes, proxies for U.S. and global stocks, were 2.6 percent
and 2 percent, respectively.
Investing in a low-return environment is extremely
difficult. Returns for both stocks and bonds are expected to be
in the single digits, at best, for the next several years, if
not for the rest of the decade. Even a great alpha generator
will deliver less if the pie is smaller. A bull market
covers a lot of sins, says AQRs Asness, whose
firms view is that equity markets are expensive and bond
markets are wildly expensive, leading to paltry
returns going forward: Whether its alpha or not,
investors will feel like its tougher.