Risk tolerance, smart teams, and results

Imagine this: you just accepted a position as a member of the investment committee of a global corporation. Other members include the CFO, the Treasurer, and other senior executives. Your job is to oversee the management of the company’s cash and marketable securities portfolio. You’re armed with your CFA knowledge (you passed Level 3 last year) and your usual boundless curiosity.

Fortunately, the company is a member of peer forums where you can exchange ideas with other large corporations. You talk to investment managers at high-tech, industrial, and pharma companies. You learn about their investment objectives, investment policies, asset allocation, systems and operational infrastructure, headcount, processes, you name it. You ask your peers about everything that you can think of.

Company X is so sophisticated

You observe that there is a wide spectrum of styles. For instance, Company X and Company Y have similar portfolio size, yet Company X has consistently achieved 5x the return of Company Y. “Well”, you think, “Company X has a very large investments team (10+ professionals), and Company Y has a team of one. That has to be what drives the difference. Company X is way more sophisticated, they hire seasoned pros and run all types of analytics all of the time.”

But… Company Z always beats Company X

You investigate more and find Company Z. It’s only two guys running the portfolio and their returns have exceeded those of Company X by a wide margin each year for the last five years.

How is this possible? Only two people vs. a battalion at Company X…

With the same portfolio size, they both manage money in-house and also leverage a suite of external asset managers to run specialized strategies. In terms of analytics, both companies use modern portfolio theory tools, statistical analysis, stress testing, etc. One difference is that, given the limited size of the team, Company Z needs to take an 80/20 approach. “We have to be selective in our search of signal vs. noise”, they say.

You talk to every single team member of both companies. Are the folks at Company Z smarter than the ones at Company X? You don’t think so. If anything, Company X has made a few hires from Wall Street (which one would think are the smart ones).

So what is it then?

Risk tolerance is the master factor

It turns out that the investment committees of the three companies had set out the following eligible investments:

• Company Y: money market funds, commercial paper, T-bills
• Company X: same as Company Y, plus treasuries, agencies, corporate bonds (investment grade), and mortgage-backed securities
• Company Z: same as Company X, plus emerging market debt, and high-yield bonds

The difference in risk tolerance is clear. No matter how many resources we assign to Company Y, they cannot beat Companies X and Z.

Risk tolerance is the master factor. You have to be aware of it, especially if you are in a position to set policies.

The interplay of risk tolerance, smart teams, and results

If you dial risk tolerance way to the left (i.e., you’re scared of your own shadow), you don’t really need skilled managers. Also a missed opportunity is that the people you put in charge will not learn much about the profession.

As risk tolerance increases, a more sophisticated operation is needed and you will be better served with smarter people than with robotic operators. Your investments team will also enjoy their job more because their sandbox is larger.

In terms of headcount, more is not necessarily better. There is a right size and, in many (or most) treasury areas, a small team of “A players” should be all you need. As with the example of Companies X and Z, more bodies doesn’t correlate with more value or better results.

Closing thoughts: be aware of risk tolerance signals

There is a mirage that mixes headcount, sophistication, and results → “the larger the team, the more sophisticated it is, the better the results.”

It misses the master factor: risk tolerance.

Don’t blame your team for low returns or low sophistication without first examining the risk tolerance signals you’re broadcasting.

Btw, stating the obvious: the application of this goes beyond the realm of investment management. See the footnote.

Footnote

When I was drafting this post, Jason Fried from 37signals happened to be typing something very similar: see it here.