Go to the Prior Tip** ***Real
Options
*Go to the Next Tip PowerPoint: an Inferior Solution ... for
me

Return to MaxValue Home Page

Risk and decision-making continue to inspire new books. This one is a particularly easy read. Mathematical details are minimal. There are many stories about decision failures; proper problem framing would have steered decision makers toward better choices.

This book prescribes a method to adjust values for attitude toward risk. The
authors are particularly concerned about Regret (which they capitalize). The value
of a risk situation is:

Value = U - lR

where U is the expected value Upside (positive outcomes from the benchmark
reference), R is the expected value Regret (negative outcomes from the benchmark),
and l (Greek letter lambda) is the ** risk aversion
constant**. l values > 1 represent risk
aversion.

is another book featuring
a "prescriptive" form of decision analysis. (A better one, I think, is Seeing TomorrowSmart Choices, reviewed in an earlier Tip.) These
methods are helpful. However, in my humble opinion, the method of adjusting for risk
aversion is inferior to using an exponential to
express risk policy. Another book that I recommend is utility functionAgainst
the Gods. |

The best part of ** Seeing Tomorrow** is the stories.
Most of the transactions are drawn from banking, insurance, and mergers and
acquisitions. Decision failures happened because people failed to recognize or give
adequate weighting to possible downside outcomes. In several cases, losing the deal
was often accompanied by more Regret than was appreciated beforehand.

Dembo and Freeman start early with a good definition:

"Risk is a measure of the potential changes in value that will be experienced in a portfolio as a result of differences in the environment between now and some future point in time." |

There are four key components in how the authors approach risk management:

- Time Horizon
- Scenarios (what events might unfold, and what would these do to the future value of investments?)
- Risk Measure (they advocate using Regret, the expected value or probability-weighted average of unfavorable outcomes at the time horizon)
- Benchmarks (reference for comparing our results)

The approach so far is excellent reading. Their quantitative method might appear workable at first glance. Upon deeper inspection, you might have the same problems as I found.

The book twice uses an example where a person is faced with an 0.80
chance of gaining $1000 and 0.20 chance of losing $3000. The authors suggest that
the selling the upside, 0.20 chance of $1000, might be worth $300 to a decision-maker.
Simultaneously, this person might be willing to buy insurance for the 0.20 chance
of -$3000 downside for -$100. Thus the value is U-R
= $300-$100=$200. [l
was not used in the equation this time.] [Coincidentally, this is the same as the expected
value, 0.8($1000)+0.2(-$3000)=$200.] It believe that the fatal flaw in "the authors ground-breaking risk rules" (from the dustjacket) is that a rational (read )
person should be willing to pay conservative that $600 to insure against
(or eliminate) a 0.20 chance of losing $3000.more |

I recommend reading this book for a broad discussion of risk and decision-making with colorful examples. Please ignore the calculations. While I won't go so far as to call it "junk analysis," I believe that most people can find the mainstream decision analysis calculation methods to be more logical yet accessible.

—John Schuyler, September 1999

Copyright © 1999 by John R. Schuyler. All rights reserved. Permission to copy with reproduction of this notice.