A bank’s cash flows tend to be highly volatile and related to macroeconomic factors. This makes forecasting cash flows extremely challenging and prone to mistake. Thankfully, there is an easier way. For most businesses, the balance sheet is largely affected by management assumptions and historical events. For example, the decision between LIFO and FIFO inventory valuations can have a large impact on a business with a large inventory balance in an inflationary environment (See my post on Tesoro’s massive LIFO reserve here). The consequence of this is that, for non-banks, shareholders equity is a somewhat arbitrary measurement that is difficult to compare across firms of different ages, sizes and business strategies. For banks, this is not the case.
Banks use Mark-to-Market accounting, which carries most assets and liabilities at fair market value, rather than historical cost. In this manner, unrealized gains and losses are actually recognized (either via the income statement directly or through other comprehensive income on the balance sheet). This translates into Shareholders Equity on the balance sheet that is more reflective of the net difference between the actual market value of assets and liabilities.
Since the book value of equity is more reliable than in other businesses and the statement of cash flows is highly volatile and less accurate as a metric of assessing management competence (given the greater impact of macro rather than microeconomic factors), most analysts rely on shareholders equity as a starting point for valuing banks. This method is known as the Excess Return Model and it arrives at the value of equity as the sum of the current equity capital and the present value of expected excess returns to equity.
Finding the current equity capital is as easy as looking at the balance sheet. Finding the present value of excess returns is more challenging. Here is the equation:
Excess Return = (Projected Return on Equity – Cost of Equity) * (Beginning Equity Capital)
Projecting a bank’s future return on equity can be challenging. A logical starting point is to look at a long history of the bank’s actual returns on equity, and then making adjustments for the future. This approach grounds the analysis in real returns that have been attained in the past (rather than committing the classic business school mistake of starting from zero and building up to returns by layering assumptions upon assumptions, resulting in projections that are orders of magnitude off of the best or worst performance ever achieved by the company, or even any company in the industry!), and then makes allowance for projected changes in the operating environment. This is the stage where the investor takes into account the bank’s strengths and weaknesses relative to its competitors, as well as expected changes to the macroeconomic environment. I encourage conservatism here, and note that this is a highly imprecise procedure that many people fail to embrace by instead relying on a series of highly imprecise assumptions.
Coming up with a firm’s cost of equity usually means (to the MBA crowd at least) the Capital Asset Pricing Model, which incorporates a firm’s beta and the equity risk premium. Again, don’t let the market fool you into using an abnormally low cost of equity just because the CAPM spits one out. I tend to push the cost of equity up (admittedly, this is somewhat arbitrary) to reflect what I think a minimum acceptable return would be, and I adjust for risk by focusing on firm’s fundamentals rather than the wholly irrelevant volatility of its stock price.
If a bank is earning extremely high excess returns now, it is important to do a multi-period valuation whereby these returns decline to a long-term sustainable level over time. Once the firm reaches its long-term sustainable operating level, you calculate a terminal value that incorporates this long-run moderate growth. The objective is to arrive at expected excess returns for each year in the future, either through a period of higher than normal growth with a terminal value, or modeling normal growth beginning now (for a relatively established bank).
There are other methods of valuing a bank (See Aswath Damodaran’s chapter on valuing financial service firms here). One can value the dividend stream using payout ratios (even if dividends are not currently being paid, but are expected to be forthcoming in the future), or asset liquidation values (using weighted average interest rates on the loan portfolio relative to market rates), and a plethora of relative valuation models. It appears that the excess return model is the most common, and I think the most theoretically justifiable given its focus on absolute value and grounding in what has actually been achieved and current equity value.
Now that we know the general framework for valuing a bank, stay tuned as we will next dig into a fairly simple bank, the Bank of Internet (BOFI).