Value investors will sometimes look at book value when considering stocks that may be good investments. In his annual letter, Warren Buffett of Berkshire Hathaway (NYSE:BRK.A) always begins with a discussion of Berkshire's book value (for reference, Berkshire currently trades around 1.4x price-to-book value). In one of my previous articles, I also looked at book value to identify five potential stocks that appear cheap because they trade below 1.0x price-to-book. However, book value sometimes is not a good indicator when determining the appropriate value of a stock. Specifically, when companies like IBM engage in consistent stock repurchases over the years, investors should ignore book value because it undervalues the true value of the company.
Stock Buybacks and Treasury Stock
When Buffett invested in IBM, he probably did not look at book value because IBM currently trades at 12.9x price-to-book. That is a rich multiple, especially for a value investor like Buffett (P/B multiples of 1-2x or less would typically indicate potential bargains). However, because IBM has consistently engaged in stock buybacks in the past 10 years (totaling over $100 billion), looking at book value will actually undervalue the true value of the company. Let's look at the accounting for stock buybacks to see why that is.
When a company uses cash to buy back stock, they will incur a deduction of cash on the asset side and a deduction of shareholder's equity or book value (through treasury stock) on the equity side. Here is a simple illustrative example (remember that the Assets = Liabilities + Equity formula has to balance for each accounting entry).
As you can see, the book value gets lowered by the amount of the stock buybacks. However, just because the book value is lowered by the stock buybacks, it does not mean the company's annual net income or annual free cash flow is affected (which is what really should be looked at to determine the appropriate market value of a company). If you look at IBM's 2012A P/E multiple, it's around 14.6x P/E, which is much more reasonable. Also if you adjust book value (by adding back IBM's treasury stock, then you are looking at a 1.7x multiple (again much more reasonable).
Stock Buybacks versus Dividends
Remember that share buyback versus dividend payout is a more efficient way to return cash back to shareholders (Buffett talked about this in his latest annual letter). For example, if we look at the illustrative example below, if a company uses cash to buy back shares instead of paying out dividends, they return 5.7% versus 5.3%, assuming the same tax rate for dividends and capital gains and assuming the P/E multiple is the same 15x under both scenarios.
(click to enlarge)
Note that this is also before considering that the gains on stock buybacks can be deferred until you actually sell the stock.
What Drives IBM's EPS Growth
EPS, or Earnings Per Share, can be driven either by 1) net income or earnings increasing (numerator), or 2) share count decreasing (denominator) due to stock buybacks. If you look at IBM's past 10 year EPS growth (2003-2012), you can see that EPS growth has been driven on average approximately 65% by net income growth and 35% by share buybacks.
Net income growth has been driven primarily by IBM shifting more and more of their sales and profits to Software, which has the highest gross margin of +85%.
From 2003 to 2012, IBM increased its software business contribution to pre-tax income for 35% in 2003 to 45% by 2012, nearly half the company's total profits.
Net share buybacks have averaged $8 billion per year in the past 10 years (I have also netted out positive share increases due to stock compensations from table below), which has resulted in lowering diluted shares outstanding by 4-5% each year (which results in EPS increasing by 4-5% per year).
At first glance, IBM may look like an expensive investment if you look only at their price-to-book multiple of 12.9x. However, remember that IBM has significantly bought back shares in the past 10 years and a more appropriate multiple to look at is price-to-adjust book value (which adds back treasury stock), which is just 1.7x. Furthermore, IBM trades around 15x FCF multiple, which is a fair multiple to pay for a company with 50% of profits coming from software (recurring revenue) and a shareholder friendly management team (consistent buybacks and dividends). I don't own any IBM currently, but would be very interested in establishing a position on any future pullback (maybe if/when it gets back to more of a 12-13x FCF multiple, or $180 per share). Good luck as always.