Benjamin Graham's fourth lecture at the New York Institute of Finance tackled the complex question of analyzing future earning power and its role in equity valuation models.
Graham began by differentiating precise terminology between "past earnings power" as average historical earnings over a defined period and "future earnings power" as expected earnings over the upcoming five years. He explained that while past earnings are known quantities, the future remains uncertain. However, past results still provide a rough predictive guide given the general continuity of business operations. This underlying persistence means that while value depends primarily on the future, historical data remains insightful for analysis.
Understanding the role of past earnings in bond evaluation
Graham noted how for high-grade, fixed-income securities, adequate past earnings coverage alone often suffices to determine safety without heavily weighing untested predictions. If historical profits comfortably cover interest obligations with a large cushion, the bond can be considered sound without relying extensively on earnings forecasts. However, for equities, valuation requires more intricate modeling to supplement historical figures.
The art of estimating equity value
Graham delineated the standard technique for equity valuation: carefully estimate future earnings based on sales volume, prices, costs and taxes. Then multiply this earnings forecast by an appropriate capitalization rate (price-earnings ratio) to derive valuation for guiding buy and sell decisions.
He shunned arbitrary rules of thumb in selecting the price-earnings multiple. Graham condemned naive approaches like penalizing a disliked company with an ultra-low multiplier or rewarding a favored stock with sky-high valuations. The earnings multiple should remain grounded in rational analysis, albeit with room for subjective judgment within reasonable limits.
The unpredictability of earnings: A thought experiment
To illustrate the difficulties of earnings prediction, even with perfect foreknowledge, Graham presents a thought experiment. The year is 1939 and an investor possesses a crystal ball foretelling 1939 to 1946 sales growth by industry. Given this prescience, Graham asks students to forecast likely stock price performance. The results starkly contradict expectations.
Despite aircraft manufacturers' sales exploding by 3,100%, their share prices dropped 74%. Amusement stocks rose 242% against 240% sales growth. This experiment demonstrated the hazards of overconfidence when analyzing equities. As Graham concluded, "Even if we knew what was going to happen to a company’s business and earnings, we might not be able to accurately predict what would happen to its market price."
A practical example: The Childs Company valuation
Graham provided a detailed example of security analysis by walking through the SEC’s 1946 valuation of Childs Company for a complex reorganization. The core challenge was estimating normalized future earnings to derive a fair valuation.
The SEC started with a bottom-up analysis of each revenue and expense component. This included a three-page assessment of management, along with detailed breakdowns of sales, costs, profits, depreciation, rent, overhead and more.
After extensive modeling, the SEC forecasted “normal” sales of $18 million based on the 53 restaurants’ 1945 results rather than latest figures. A 6% profit margin based on industry data yielded $1.1 million in pre-tax income. The SEC discounted the tax rate from 38% to 35%, predicting Congress would maintain relatively high corporate taxes despite reform pressure.
This resulted in normalized earnings of $715,000. Capitalizing at 12.5 times based on restaurant industry comps gave a $9 million valuation. The SEC further tweaked this figure by adding $1.2 million in tax benefits and subtracting $1.8 million for rehabilitation costs to derive $8.4 million earnings value. Including $5.1 million of excess working capital and real estate took the total valuation to $13.5 million. Deducting $3.2 million of debt gave $10.3 million in equity value.
With $7.6 million owed to preferred shareholders, $2.7 million in value remained for the common stock, implying a 25% to 30% allocation to common in the reorganization.
The takeaway from the case study
While noting some critiques, Graham affirmed the diligence and logic behind the SEC’s systematic approach. Beginning with historical data, the SEC incorporated conservative growth estimates based on past trends and industry benchmarks. The 12.5 earnings multiple fell within a reasonable range with an 8% capitalization rate. Graham observed that other acceptable valuations could have been derived using straightforward rules of thumb given the inherent uncertainty in predicting the future.
The SEC’s elaborate methodology required many judgment calls on sales, margins, taxes and the appropriate earnings multiple. Graham stressed that while detailed modeling appears highly precise, estimates of the future remain guesses. Security analysts should acknowledge this uncertainty rather than claiming false precision. Leaving ample margins of safety and contingency cushions helps compensate for imprecise forecasts.
Conclusion
Ultimately, Graham demonstrated that skilled security analysis entails both art and science. The Childs valuation illustrated best practices of the craft – a rigorous process grounded in historical performance, conservative growth estimates and peer benchmarks. But material uncertainties remain. Diligence around earnings projections improves the odds, but perfection lies beyond human reach. In Graham’s world, margin of safety and humility reign supreme.
