Benjamin Graham's Risk-Arbitrage Equation

Graham's Risk-Arbitrage Equation is from an Appendix to the 1951 edition of his earlier book, Security Analysis.

Introduction

"I helped Ben with the third edition of Security Analysis, published in 1951. In the appendix is an article on special situations that first appeared in The Analysts Journal in 1946. In the article, he had worked out an algebraic formula for risk-reward results that could be applied today, 37 years later."

Walter J. Schloss, Benjamin Graham and Security Analysis: A Reminiscence (1976) [PDF].

The above Reminiscence is referenced from The Rediscovered Benjamin Graham: Selected Writings of the Wall Street Legend (1999) [PDF], which puts the date of writing as 1976. However, that does not seem to add up to the time period given by Schloss: 1946 + 37 years.

A copy of the article on the Columbia Business School site puts the year at 1976 as well [PDF].

However, since the first edition of Security Analysis was published in 1934 and the second edition in 1940, It's possible that the formula was developed before it was published in The Analysts Journal in 1946.

The Equation

Indicated annual return = GC – L(100% - C) / YP

Where:
G be the expected gain in points in the event of success;
L be the expected loss in points in the event of failure;
C be the expected chance of success, expressed as a percentage;
Y be the expected time of holding, in years;
P be the current price of the security.

Special Situations

Warren Buffett wrote the preface to Graham's later work — The Intelligent Investor — in which he calls it "by far the best book about investing ever written".

Graham does not mention this equation in The Intelligent Investor. Instead, he includes all such operations under the category of special situations; and writes:

"[special situations] include intersecurity arbitrages, payouts or workouts in liquidations, protected hedges of certain kinds.... The exploitation of special situations is a technical branch of investment which requires a somewhat unusual mentality and equipment. Probably only a small percentage of our enterprising investors are likely to engage in it".

Comments

Walter Schloss mentioned this equation in his 2008 Ben Graham Center for Value Investing speech. I was just wondering what's the basis of determining the expected loss in points in the event of failure or "L"? Also, what the basis of "C"?

Thank you for your comment, Will!

Graham does not specify any criteria for calculating L and C, but does give the following example with the equation.

We may take as a current example the Metropolitan West Side Elevated 5s selling at 23. It is proposed to sell the property to the City of Chicago on terms expected to yield in cash about 35 for the bonds. For illustrative purposes only (and without responsibility) let us assume (a) that if the plan fails the bonds will be worth 16; (b) that the chances of success are two out of three—i.e., 67% (c) that the holding period will average one year.

Then by the formula:
Indicated annual return = (12 x 67) - (7 x 33) / (1 x 23) = 24.7%

Note that the formula allows for the chance and the amount of possible loss. If only possible gain were considered, the indicated annual return would be 34.5%.

Please note again that this equation is from an Appendix to the 1951 edition of Graham's earlier book, Security Analysis. Graham makes no mention of this formula in his later work, The Intelligent Investor.