Turtles in Omaha

Excerpts from Turtles in Omaha by Michael J. Mauboussin, Chief Investment Strategist, Legg Mason Capital Management. Click here to view more.

The Mindset of Great Investors
The difference in [investment] return had nothing to do with knowledge and everything
to do with emotional and psychological factors. We had all been taught the same thing,
but my return . . . was three times that of the others. Over the years, I kept finding
evidence that emotional and psychological strength are the most important ingredients
in successful trading.
-- Curtis M. Faith, "Way of the Turtle"

Nassim Taleb’s latest book, The Black Swan, is a treatise on the improbable events Buffett has in mind. 5 The term black swan comes from philosopher Karl Popper’s criticism of induction: We get closer to truth if we focus on falsification instead of verification. Seeing lots of white swans
(verification) does not allow for the statement “all swans are white,” but seeing one black swan
(falsification) does disprove the statement. This is relevant in investing because investment
strategies based on the reoccurrence of white swans can be toppled by one black swan event.
Taleb suggests all black swans have three attributes: they are outliers, they have an extreme
impact, and people seek to explain them after the fact.

Three High Hurdles
Here are three psychologically-difficult barriers great traders and investors must overcome: loss aversion, frequency versus magnitude, and the role of randomness. How individuals cope with these barriers provides good insight into their investing temperament.
Loss aversion. In what is now a well-documented and well-known phenomenon, humans suffer roughly twice as much from losses as they receive pleasure from comparable gains. An important consequence is investors will turn down positive expected-value financial propositions, especially when their recent results have been poor.
Faith provides a powerful example of this point. Following the expiration of the confidentiality
agreement he signed, Faith explained the turtle system to a friend. Noting that cocoa presented a great trading opportunity in 1998 through early 1999, he inquired how his friend was doing in
cocoa. The friend replied he stopped trading cocoa because he had lost money and thought the
trade was “too risky.”
Then Faith explains the circumstances. Following the system would have generated 28 total
trades (average size $10,000 – $15,000) from April 1998 through February 1999, producing a
total profit of nearly $56,000. But of the 28 trades, 24 were unprofitable (average loss of about
$930) while 4 were profitable (average gain of roughly $20,000). Even more difficult, the first 17 trades in a row lost money.
Given this profit pattern, it is not difficult to see why a trader would abandon the commodity and perceive it as overly risky. But Faith’s point is crucial: Recency bias and loss aversion often cause you to give up right before the trade becomes profitable. Sticking with positive expectation financial propositions is essential to maximizing profits over time.
Frequency versus magnitude. This concept is really an extension of loss aversion. Most of us frame the success or failure of a financial proposition in terms of the price. For instance, if you buy a stock at $30, any price above that level is mentally successful; any price below it is
mentally unsuccessful.
What investors often fail to consider is that change in wealth is not a function of how often you’re right, it’s a function of how much money you make when you’re right versus how much you lose when you’re wrong. You need to consider both frequency and magnitude to understand investment results.
Faith illustrates this point by sharing 20 years of results for a trading system. Over that time span, the system generated about 5,600 trades, or around 250 a year. Of those trades, a shade over two-thirds lost money, making the success ratio less than one-third. But the winning trades
earned 2.2 times the losing trades on average, netting a substantial overall profit.
As with loss aversion, operating according to the frequency-and-magnitude maxim is easier said
than done. Faith notes, “Some of the Turtles had a hard time with this concept; they felt the need to be right and to predict markets.”
The expected-value mindset has served many well-known investors well. One example is George Soros. Former colleague Scott Bessent said in a recent interview, “George has a terrible batting
average—it’s below 50 percent and possibly even below 30 percent—but when he wins it’s a
grand slam. He’s like Babe Ruth in that respect.”

Role of randomness. Most people agree stock prices move more dramatically than business
values move. In the stock market, like most probabilistic systems, there is a great deal of noise in the system. However, most investors fail to recognize the degree to which randomness affects
short-term results. And, as bad, many investors have emotional reactions to short-term
randomness that undermine the quality of their decision making.
This is Faith’s comment; the idea applies to nearly everyone involved with markets: 16
Most traders do not understand the degree to which completely random chance can
affect their trading results. The typical investor understands this even less than the typical
trader does. Even very experienced investors such as those who operate and make
decisions for pension funds and hedge funds generally do not understand the extent of
this effect.
Here’s the point: A trader, or investor, can put on a positive expectation bet (correct process) and still have poor results (outcome) for some period of time due solely to randomness. But many investors attribute bad outcomes to bad processes, which leads to substantial error. As insidious is attributing good outcomes to a good process. A thoughtful investor must carefully consider process and recognize long-term outcomes will follow.
Here are some data to substantiate the point. The first is a study by The Brandes Institute called “Death, Taxes, and Short-Term Underperformance.” 18 The researchers screened for largecapitalization, actively-managed funds that had a 10-year track record through 2006. This yielded 591 funds. They then ranked the funds by decile based on annualized gains.
The top-decile group had returns in excess of 10.9 percent, and all of them delivered better
returns than the S&P 500 index. The researchers posed two questions: Did these funds have
periods of relative underperformance? If so, by how much?
The answer to the first question is a resounding yes. In fact, all 59 of the funds in the top decile
underperformed for at least one year. In its worst one-year period, the average top-decile fund
underperformed the index by 1,950 basis points, with a range of negative 650 to 4,410 basis
points.
Over a three-year period, the average underperformance was still 810 basis points, with a range
of positive 250 to negative 2,240 basis points. The one- and three-year numbers of these good long-term funds clearly show the limitations of relying on short-term results to decipher the
ultimate outcomes.
Unfortunately, the randomness in short-term results exerts a cost. Most institutional investors,
including pension funds, endowments, and foundations, rely on short-term investment results to
judge the managers they hire. Despite this, they would be better off with a robust way to assess
process. The focus on outcomes, combined with the limited appreciation for randomness, leads
to bad decisions.
In a recent academic paper, researchers tracked the decisions of 3,500 plan sponsors over a
decade. 19 What they found is not surprising. Plan sponsors hire managers after they have
enjoyed three years of excess returns. After they are hired, the managers generate excess
returns “indistinguishable from zero.”
Further, plan sponsors often fire managers after a period of underperformance, but the managers often go on to generate excess returns after they’ve been fired. Said differently, plan sponsors would have been better off on average keeping the manager they fired. And this analysis leaves aside costs.
While very understandable, this performance chasing shows many plan sponsors are fooled by
randomness. Evidence is voluminous that individual investors, too, chase performance to the
detriment of their long-term results.
Faith adamantly argues for a focus on process:
Good investors invest in people, not historical performance. They know how to identify
traits that will lead to excellent performance in the future, and they know the traits that are
indicative of average trading ability. This is the best way to overcome random effects.
This mindset fits comfortably with Buffett’s point about assessing chief investment officer
candidates based on “how they swing at the ball.”