The second lecture of Benjamin Graham, the father of value investing, focused on using comparative balance sheets to analyze the accuracy of reported earnings numbers. The reported earnings that companies publish can sometimes be misleading, especially during abnormal periods like wartime. By studying changes in the equity section of the balance sheet over time, you can calculate the true earnings and compare that to the reported earnings. This balance sheet approach provides a way to normalize the earnings and better evaluate a company's earning power.
Significance of discrepancies between reported and balance sheet earnings
To illustrate the concept, Graham provides an example comparing two companies - Transue Williams and Buda. Both companies sold stock in early 1945 at the same high price of $33.50 per share. But studying their financials shows their reported earnings per share over 10 years differed significantly from the earnings calculated using the balance sheet method.
Specifically, Transue Williams had higher reported earnings of $14.73 per share over 10 years. However, the balance sheet calculation showed it actually earned only $12.32 per share over that period. This means it lost $2.41 per share that did not show up in the reported earnings.
Case Study: Transue Williams and Buda
This discrepancy happened because Transue Williams set up reserves for things like renegotiation of wartime contracts. It charged part of these reserves to earnings each year, reducing reported profits. But then the reserves proved inadequate - it had to charge even more than the reserves to cover the actual costs when they came due.
In contrast, Buda set up very high-contingency reserves that turned out to exceed any real liabilities. So the reported earnings of $24.57 per share over 10 years understated its balance sheet earnings of $29.74 per share. Buda's reserves impacted the reported earnings in the opposite direction compared to Transue Williams.
Understanding the nature and treatment of reserves
The reason for these differences comes down to the nature and treatment of the reserves. For Transue Williams, the reserves represented real liabilities that were actually underestimated. The reserves were necessary to cover future renegotiation and other charges. But for Buda, the large contingency reserves apparently never materialized into actual liabilities. The reserves just sat on Buda's balance sheet rather than hitting the income statement.
To summarize, Transue Williams' reported earnings were overstated because its reserves ended up too low. Buda's earnings were understated because it set aside excessive reserves that never got used. The balance sheet analysis adjusts for these reserve accounting impacts to show the real earnings.
Curtiss-Wright case study
An dditional example he provided was "war baby" aircraft makers like Curtiss-Wright (CW, Financial) that show even larger discrepancies between reported and balance sheet earnings. Curtiss-Wright's reported earnings per share over 10 years were $12.28, while balance sheet earnings were $18.53, 50% higher.
In this case, Curtiss-Wright set up massive reserves during the war years that ended up far exceeding the actual liabilities or costs when peacetime resumed. So the reported earnings were understated thanks to excessive contingency reserves. The balance sheet approach reveals the higher earnings after adjusting for these overestimated reserves.
Applying balance sheet analysis on railroad companies
Graham recommended applying a similar balance sheet analysis for railroad companies. Focus on earnings before taxes, depreciation and amortization, as these non-cash charges can distort the earnings picture.
Case Study: Denver Railroad
He provided an example of the Denver Railroad's reported loss of $7 million in 1945. But earnings before taxes and depreciation were actually $27.7 million in 1945, higher than 1944's $23.2 million. So Denver's operations were stronger in 1945 if you exclude the tax and depreciation charges.
The reported loss in 1945 happened because Denver took much higher depreciation and tax charges related to prior years, not the current year's operations. For instance, $7.4 million of the 1945 tax payment covered potential deficiencies for previous years. And $5.3 million of depreciation related to prior periods was also charged in 1945.
Adjusting for these anomalies shows Denver actually earned a $1.8 million profit in 1945, not the reported $7 million loss. The excessive 1945 depreciation and tax charges were semi-manipulative in the sense that they did not reflect the actual current year performance. So the balance sheet earnings before deductions provide a better picture of Denver's earning power in 1945 versus 1944.
Conclusion: The importance of a comparative balance sheet approach
In summary, the comparative balance sheet approach reveals a more accurate view of true earnings by adjusting for reserves, depreciation and other unusual or abnormal charges. Just focusing on the reported earnings numbers can result in a misleading representation, especially for wartime companies. But analyzing the equity changes on the balance sheet provides a way to normalize earnings and evaluate earning power consistently over time.
