Under Control


Recently, I talked to the past president of one of my old companies. We were lucky and the company ( Andover Controls ) made it. He bought me an steak and told me his sad tale. He tried his hand in venture capital (VC) at the "Angel" level, and seemed not to be doing as well as his peers. Most VC guys hover around 20 percent compounded ROI ( return on investment ) in real dollars since the time of Caesar. Most VC's consider the business of investing in small companies as low risk over time.

George was trying to cut his losses by concentrating on the lowering of risk across the breadth of the portfolio. And in doing so was decreasing the value of the entire portfolio. What was that so? And why was he dead wrong in trying to reduce risk?

Imagine, if you will, a 20-animal horse race. Each horse has a one-in-20 chance of winning. And the payoff is $100 for each dollar bet. Clearly we should bet on every horse - our winnings are guaranteed. It would be folly to bet on only one, and if it was losing, to bet even more on the same horse. Yet in many companies, we insist that the losers be supported at the expense of opportunity with the winners.

Economists have observed that tax revenues are dependent on the tax rate. So far, so good. If the rate is zero, for example, the state revenue is zero. Conversely, if the rate is 100 percent, the state revenue is zero again. Somewhere between is the optimum tax rate for maximizing state revenue. About 30 percent seems to be the optimum across many societies. If you increase the rate to increase revenue, and the rate is already high, the state net revenue will go down. Fascinating.

In an investment or R&D portfolio, the examination of the extremes leads to the same types of considerations. Having 20 investments of minimal risk ( such as government bonds ) yields a return of about 3 percent, discounting inflation. Close enuff to zero for me. In fact, if others ( competition ) are doing the same, you only stay even.

In the same way, infinite risk also yields an R&D benefit of near zero. What is the optimum.? About 80 percent failure rate will optimize portfolio return. In other words, you should do deals that have a likelihood of failure greater than half. George has the problem that he tried to reduce risk, rather than increase risk. Risk management is not the unilateral reduction of risk, but the optimization of the risk-to-reward ratio.

In the Darwinian universe, mutations guarantee the continuation of life. Most mutations don't work. But the ones that do are fortuitously adapted to an environment in flux. We cannot predict the outcome of a single coin toss, but we can predict the outcome of a single coin toss, but we can predict the outcome of a large number of tosses. Having a feel for the "Law of Large Numbers" is essential for managing portfolios. Many managers are too conservative and manage too closely. Many managers are too conservative and manage too closely.

Imagine that you have $20 million to invest. Where do you put it. In a portfolio of 20 companies that have a 50 percent chance of being worth $2 million for each $1million invested, and a 50 percent chance of being worth $1million?. Or in a portfolio of 95 percent chance of zero return and a five percent chance of returning one billion dollars. At the end of the 10 years, you have about $35 million in the first case for your $20 million investment. And in the second case you have a 64 percent chance of becoming a billionaire. Think Bill Gates. The greater the risk, the greater the reward.

R&D investments planning should be centered around failure. Increase risk and the likelihood of success increases across the board. But not for the individual project. We cannot predict individual stars or bums, but we can predict the group behavior.

It is reasonable to establish risk as a parameter of management. But too often the culture of companies penalizes failure for each endeavor, rather than the total success. Success in poker is knowing when to fold, not when to bet. Establishing a failure rate and know that it's the way to maximize return. Without unpredictability, you lose control and lower benefits. Too much individual risk never happens in real companies. Encouraging failure as a policy seems strange, but it is the way to succeed with minimum risk. Mature companies cannot grow in market share without knowing the "Law of Large Numbers."

As appeared in Manufacturing Systems Magazine April 1995 Page 8

Manufacturing Systems Magazine Articles
References - Table of Contents

Send mail to rmi.info@barn.org for more information.
Please send mail to webmaster@barn.org regarding web site structure.
Copyright © 1996-2002 R.Morley Inc. All Rights Reserved

R. Morley Incorporated
614 Nashua Street, Suite 56
Milford, NH 03055-4992 USA
Tel: 603-878-4365 FAX: 603-878-4385