The Interpretation of Financial Statements

Benjamin Graham and Spencer B. Meredith · 1937

Thesis

A company's balance sheet and income account, read correctly and over several years, tell an investor most of what can be known about its financial strength and earning power, and an investor who knows a company's present position and past record is better equipped than one who relies on its story or its stock price.

The main ideas

Financial statements are the investor's main evidence. Read them before forming a view of the future.
Working capital and the current ratio measure whether a company can meet its near-term obligations.
Book value and net current asset value show what the shareholder owns, but both can be overstated.
Intangibles are worth what they earn, so the income account, not the balance sheet, reveals their value.
Earning power is an average over years, not a single year's result.
The safety of bonds and preferred stock depends on how many times earnings cover their fixed charges.

Statements as evidence

Benjamin Graham and Spencer Meredith, who taught security analysis together at the New York Stock Exchange Institute, wrote this short primer in 1937 for investors who wanted to read company reports without an accountant's training. Its premise, stated in the preface, is that if you have precise information about a company's present financial position and its past earnings record, you are better equipped to gauge its future possibilities.

The book is organized as a tour of the statements. It takes the balance sheet item by item, from property and depreciation through intangible assets, current assets, current liabilities, reserves and capital, then turns to the income account, earning power and the safety of fixed charges. It closes with a method for analyzing statements by ratios and a glossary of financial terms.

The mechanism is that statements record what has actually happened to a company's money, while prices and forecasts record what people expect. Expectations change quickly; the record changes slowly and is harder to fake.

Implication: before judging a stock by its price or its prospects, read its statements for several years and form a view of what the business has actually done.

Working capital and the current ratio

Graham and Meredith treat working capital, current assets minus current liabilities, as of major importance in determining the financial strength of an industrial enterprise. Current liabilities are debts arising from ordinary operations and payable within a year, together with any other debts maturing within a year.

The current ratio, current assets divided by current liabilities, is their main test. A ratio of two to one is the conventional standard for industrial companies, but they note that the satisfactory ratio varies with the line of business: the more liquid the current assets, the less margin is needed above current liabilities. A stricter test asks whether quick assets, current assets other than inventory, exceed all current liabilities; failing it matters less if inventories are readily salable.

The mechanism is that a company fails not when it is unprofitable but when it cannot pay what it owes as it falls due. Working capital is the buffer between operations and default.

Implication: check a company's current position before anything else. A thin or shrinking margin of current assets over current liabilities is a warning, however good the earnings look.

Book value and its distortions

Book value, or equity, is what the balance sheet says belongs to the shareholders. The book devotes several chapters to calculating it, and to its pitfalls. Property can be inflated: the authors cite U.S. Steel, whose property account was written up by more than $600 million, giving its common stock a largely fictitious book value. Reserves can mislead in either direction. Graham and Meredith distinguish reserves that offset assets, reserves that are really surplus, and reserves that are real liabilities, and each must be treated differently to arrive at a true figure.

Net current asset value, current assets less all liabilities, is a stricter measure. The authors describe the net current asset value of an industrial security as likely to constitute a rough measure of its liquidating value, what shareholders might receive if the business were wound up.

The mechanism is that balance sheet figures depend on accounting choices made by management. The reader must adjust them to see what the assets are really worth.

Implication: never take book value at face value. Ask what the assets would fetch, and treat net current assets as the most conservative floor.

Intangibles and earning power

Goodwill, patents and trademarks may have a very large value, the authors write, but it is the income account and not the balance sheet that offers the clue to it. What counts is the earning power of the intangibles, not the figure at which they are carried.

Earning power itself is defined as the earnings that can reasonably be expected over a period of time in the future. Under reasonably normal conditions, the authors argue, an average over a period of years may be a better index of earning power than the current figure alone, and they warn readers to be particularly careful not to exaggerate the significance of a single year's earnings. They also tell readers to examine the surplus account alongside the income account, because companies sometimes charge real business losses to surplus or reserves, bypassing the income account and flattering reported earnings.

The mechanism is that a single year can be distorted by cycles, one-off events and accounting choices, while an average over several years reveals what the business typically earns.

Implication: value a company on its average earnings over several years, adjusted for losses charged elsewhere, not on its latest year.

The safety of fixed charges

For bonds and preferred stocks, the question is not growth but safety: whether the company will always be able to pay what it owes. The book's chapter on the safety of fixed charges and preferred dividends measures this by how many times earnings cover interest and preferred dividends, and sets minimum average coverage standards that differ for public utilities, railroads and industrial companies, reflecting how stable each kind of business is. For a holding company's bonds, the preferred dividends of its subsidiaries count as fixed charges ahead of them.

The mechanism is that coverage measures the cushion between earnings and obligations. The more variable a company's earnings, the larger the cushion it needs.

Implication: before buying a company's bonds or preferred stock, check that its average earnings cover the fixed charges several times over, with a larger margin the less stable the business.

Limits

The examples belong to another era. The reviewer at The Rational Walk called the book an excellent primer but noted that its examples come from 1930s industrials, railroads and utilities, before modern accounting standards and before most of today's industries existed.
Book value has lost much of its relevance. Baruch Lev and Feng Gu argued in The End of Accounting (2016) that reported earnings and book value explain a far smaller share of market value than they did in the 1950s, largely because accounting ignores intangible investment. In his 2018 letter, Buffett said the annual change in Berkshire's book value had lost the relevance it once had.

The Intelligent Investor — Graham's guide to investing, which relies on reading statements this way.
The Essays of Warren Buffett — owner earnings and intrinsic value, Buffett's refinements of Graham's measures.