When people talk about the book value for a company, what do they mean? It’s a term that usually comes up when discussing when trying to figure out the true worth of the company. It’s particularly used in the context of pinning a fair value on insurance companies and bank stocks.
The reason why I rarely use it in my own writings is my because it is a technique that measures “the accounting value” of a company rather than the intrinsic value of the company. Book value attempts to measure the liquidated value of a company if you add up all the assets and subtract all the liabilities.
That is a very interesting thing to know. If a stock sells for $25, and you believe that the firm could be chopped up into parts and then sold so that you would receive $50 per share in cash at the end of a liquidation, that gives you a nice margin of safety.
But with book value, the devil is in the details. When calculating the value of assets and liabilities, the calculation is measured according to historical cost. If the management of a company has a buy-and-hold orientation, then the relationship between book value and the current earning capacity of the business gets completely skewed.
At a firm like Berkshire Hathaway, where purchased assets are held on the balance sheet for decades, the distortion created by using book value can become enormous. See’s Candies has a book value in the $20 million range because that is the historical cost of the firm. Well, it’s now paying out somewhere between $50 million and $100 million per year in cash that flows straight to Omaha. Depending on the economic conditions of the time, the earnings capacity of the business would value it somewhere between $800 million and $1.4 billion on the private market.
Now, of course, Warren Buffett’s management style is unusual–the typical firm has far more regular asset churning, and therefore, you don’t see as frequent of distortions between long-term growth and historical cost.
Even recognizing this, there are still issues using book value to gauge the earnings capacity of the firm: (1) it assumes that the management paid an accurate price to acquire a business–if it paid too much, the historical cost of acquiring it and therefore the value of the asset on the balance sheet may be overstated, and conversely, management teams that get good deals and pay less than the asset is worth will have understated book values; (2) the long-term liabilities on the balance sheet of the acquired firm are subject to changes as interest rates ebb and flow; if you buy a company for $200 million that has a balance sheet with $300 million debt at 8% interest, the earnings capacity of the firm will increase substantially if you can refinance at 4.5% and then lower the amount of funds being used to pay long-term interest. None of this affects book value.
Some of these concerns are alleviated by retained earnings which is a way for book value to re-establish some type of connection to the value of a company. When Warren Buffett purchased Precision Castparts, he used a lot of Berkshire’s cash to do so. Presume that some of that cash came from a See’s Candies cash payment to Berkshire Hathaway that made its way to Omaha as retained earnings. In that case, the hypothetical $75 cash payment from See’s Candies to Berkshire Hathaway did become a part of Berkshire’s book value as it was used for the $37 billion acquisition of Precision Castparts that increased the book value of the stock.
So, in an indirect way, even as the See’s Candies book value is measured according to historical cost, the growth of the business is somewhat reflected in Berkshire’s book value as it was used to purchase Precision Castparts that raised Berkshire’s book value substantially.
With this in mind, you should be aware of the limitations associated with book value. If Exxon buys an oil well for $5 billion, and it immediately generates $500 million in annual profits but ten years down the road happens to generate $1 billion in annual profits, you will see no increase in book value if Exxon takes those profits and ships them off to shareholders in the form of dividends. The asset has grown, and is showering shareholders with cash, yet the book value won’t increase when the growing earnings are paid out to shareholders as dividends.
So why does every financial calculator provide the book value for a firm if there is sharp limits to its use in evaluating earnings capacity, and even worse, can distort the picture? (1) Because it is easy to calculate, (2) it provided historical importance back in the day of railroads, bank stocks, and natural resources companies when the purchase of physical assets rather than the individual talents of employees drove value, (3) there is an assumption that a business constantly purchases assets as it becomes successful and these larger purchases will reflect close to reality at the time and dwarf the book value distortions created by the book value of old assets, (4) there is an assumption that management pays fair value when purchasing an asset, (5) and it takes a ballpark swing of the liquidation value of the firm and thus provides a floor for the stock’s value.
Personally, the only time I ever use book value as a meaningful metric when studying a company is when I see banks selling at substantial discounts to book value. Even then, it must be cross-referenced with the anticipated default rate of the existing loans.
If a bank has a large loan portfolio of E&P commodity stocks during an oil bust, the $1 billion book value that comes from the loan portfolio may not mean much if there are going to be substantially higher default rates than were envisioned at the time the bank purchased the loan portfolio.
But if there hasn’t been a deterioration in the credit quality or expected default rates of a loan portfolio, and a bank is selling at a discount to book value, then I would actually use book value as a meaningful element of my analysis. That is because the loan portfolios usually aren’t that old, are straightforward to value, are not subject to technological obsolescence in a way that the inventory of IBM might be, and bear a relationship between fair market value and historical cost. Other than that, the distortions associated with book value are too great for me to actually make decisions based on what a book value per share figure is trying to tell me.