Thoughts on Hiring

April 1, 2014

Our trust is in people rather than process.  A “hire well, manage little” code suits
both them and me.  – Warren Buffett

We are frequently asked when we anticipate growing our research team; the implication being that the size of our team is a disadvantage and that adding people would improve investment results.  Our general attitude towards hiring a full-time analyst can best be described as extreme reluctance based on the following:

In 1984, Warren Buffett delivered a classic speech attacking the efficient market hypothesis at Columbia Business School called The Superinvestors of Graham-and-Doddsville. He describes a hypothetical national coin flipping contest where every citizen wagers a dollar on a daily coin flip with losers eliminated. After twenty rounds there would be 215 people that had successfully called twenty flips in a row and would have each accumulated over $1 million. He suggests that by this point the group would be writing books, appearing on television and hosting seminars explaining their coin flipping techniques. At which time a business school professor would point out that if the exercise were repeated with orangutans the outcome would be the same – “215 egotistical orangutans with 20 straight winning flips.”

However, Buffett argues that there is an important difference between the human and simian coin flippers. If 40 of the 215 winners after 20 days came from a particular zoo in Omaha, observers would be curious to learn from the zookeeper about the diet and training his orangutans receive. Buffett concludes, “if you found any really extraordinary concentrations of success, you might want to see if you could identify concentrations of unusual characteristicsthat might be causal factors.”

Buffett then proposes that origin does not have to be defined by geography. He proposes an intellectual origin and describes the intellectual village of Graham-and-Doddsville as students of acclaimed Columbia professors Ben Graham and David Dodd. He then proceeds to show the outstanding investment records of seven residents of Graham-and-Doddsville, each of whom, over a period of at least a decade, beat his benchmark by at least 700 basis points:

Investor

Duration (years)

Return

Benchmark

Advantage

Warren Buffett

13

23.8%

7.4%

16.4%

Bill Guerin

19

23.6%

7.8%

15.8%

Stan Perlmeter

18

19.0%

7.0%

12.0%

Tweedy, Browne

15

16.0%

7.0%

9.0%

Charles Munger

14

13.7%

5.0%

8.7%

Walter Schloss

28

16.1%

8.4%

7.7%

Sequoia Fund

14

17.2%

10.0%

7.2%

We think Buffett overlooked a second unusual characteristic of the superinvestors – the size of their investment teams. All but two worked alone. The exceptions, Sequoia and Tweedy Browne, each operated a single fund run by a small team.

We know of very few outstanding investment results that have been produced by large teams and “small team” seems to be a common characteristic of Buffett’s Superinvestors.

Fred Brooks, an IBM software engineer, published The Mythical Man-Month: Essays on Software Engineering in 1982. In it he describes the myth that time and labour are interchangeable. Managers often assume that the time to complete a job that takes one person two months to complete can be halved by adding a second person. However, the reality depends on the divisibility of the task.

Brooks points out that some tasks cannot be divided such as, for example, child-bearing. Some tasks, like picking grapes, are perfectly divisible because they require no communication between workers. But most tasks require at least some communication and the burden of communication must be added to the total amount of work required.

The burden of communication has two components: training and intercommunication. Each worker must be trained in a project’s goals, strategy and plans. The added cost of training varies linearly with the number of workers. But intercommunication is much more costly because it increases exponentially with the size of the team. The time spent communicating can quickly overwhelm any benefit of adding labour.

This problem is acute in investment management. In our observation, outstanding investors rarely possess outstanding interpersonal skills. The recent fallout between Bill Gross and Mohamed El-Erian at Pimco represents an unusually public demonstration of this dynamic. When you hire an investment manager, you want him managing money, not people.

A second issue with investment team size is the natural bias for action. Investment analysts are almost universally measured based on how many of “their” ideas are selected by the portfolio manager. This means that time is usually misallocated; too much time is spent looking for new ideas and not enough time is spent monitoring existing investments. It also leads, almost inevitably, to poor quality portfolio decisions. Good investment ideas are scarce and we reject dozens of ideas for every one we act upon. Most humans are not equipped to handle that much rejection, especially when their compensation is at stake. Consequently, in the spirit of office harmony, the portfolio manager feels obligated to occasionally act on a sub-standard idea to the detriment of the portfolio.

With all that said, if we can attract investment talent to our firm that is better than us, then they will undoubtedly increase our productivity and enhance returns. No individual has a monopoly on making great investment decisions and there are examples of great investment track records achieved by groups (i.e. Brookfield Asset Management). However we are realistic about the cost-benefit equation of expanding our research team.

Like great investment ideas, investment talent is rare. The challenge for Ewing Morris is, and will always be, hiring well so that we can manage little.  We are constantly meeting with candidates and when we meet the right person we will invite them to pull up a chair at the boardroom table.

« Back To Blog