How to Analyze Banks (part 5)
A deep dive on how to calculate the intrinsic value of a bank.
20 minute read
This is part five of the series How to Analyze Banks. In this article, we explore how to estimate the intrinsic value of a bank.
Overview
This article is based on Aswath Damodaran’s book The Dark Side of Valuation [1] . In his view, it is difficult to value financial companies because of four reasons. The first is that it is hard to define what debt is for a financial company, a bank, for example, has financial obligations from depositors. This means that notions like cost of capital and enterprise value, that apply to non financial companies, are not useful for valuing financial companies. Second, is that they are heavily regulated, and these regulations can impact how profitable a bank can be. The third is that, as we saw in other articles, accounting rules can vary across jurisdictions. We will confront these problems in the next paragraphs. The final factor is that defining reinvestment for a bank may be very difficult, as cash flows cannot be computed.
In the next paragraphs, we will learn how to value financial firms using equity valuation models, adapting a discounted cash flow model to best value these types of businesses. We also look at how relative valuation works with financial firms and what multiples we can use to estimate the value of these businesses.
Government Regulation
Generally, bank regulations fall into three categories. First, banks are required to maintain regulatory capital ratios, calculated based on the bank’s equity vs total assets. Second, banks are constrained in terms of where they can invest their money. For instance, until a decade ago, the Glass-Steagall Act in the United States restricted commercial banks from investment banking activities. Third, the entrants of new businesses and mergers between firms are regulated by the government. This all means that to value banks we have to be aware of the regulatory structure that governs them.
Accounting Rules
Accounting rules to measure earnings and book value are different for financial companies, for two reasons. The first is that the assets of financial service firms tend to be financial instruments (bonds, securitized obligations) that are price to market. This has implications in how we view return on equity, while the return on equity for a non-financial service firm can be considered a measure of return earned on equity invested originally in assets, the same cannot be said about return on equity at financial service firms, where the book equity measures not what was originally invested in assets but an updated market value.
The second is that the nature of operations for a financial service firm involves credit cycles with cyclical profitability. Given that banks can choose how much to provision against losses, a conservative bank might appear as less profitable during good times. A conservative bank will set aside more for loan losses, given a loan portfolio, than a more aggressive bank, this will lead to the former reporting lower profits during good times.
Debt and Equity
For financial firms, debt is a raw material, not capital. Debt is to a bank what steel is to a manufacturing company, something that is used to generate profits. Consequently, capital at financial service firms seems to be narrowly defined as including only equity capital.
So, is money taken from depositors debt? It is hard to define what debt is for a bank. Especially on interest-bearing checking accounts, there is little distinction between a deposit and debt issued by the bank.
Reinvestment Needs
For non-financial companies, growth is a function of how much a company reinvests. We consider two items in reinvestment – net capital expenditures and working capital. Unfortunately, measuring either of these items at a financial service firm can be problematic.
Consider net capital expenditures first. Unlike manufacturing firms that invest in plant, equipment and other fixed assets, financial service firms invest primarily in intangible assets such as brand name and human capital. Consequently, their investments for future growth often are categorized as operating expenses in accounting statements.
With working capital, we run into a different problem. Most of a bank’s liabilities are short term, while the minority of assets are short term. Changes in this number can be both large and volatile and may have no relationship to reinvestment for future growth.
As a result of this difficulty in measuring reinvestment, we run into the problem that we cannot estimate cash flows for future periods because we do not know how much a company is reinvesting for future growth.
Common Pitfalls
Cashflows
Taking the mental model of going inverse, in this section I analyze the common pitfalls in valuing banks. For non-financial companies, it is good to estimate their intrinsic value by forecasting cash flows after taxes and reinvestment, but before debt payments, and discounting these cash flows back at an adequate discount rate. This, however, does not work with banks because, as noted earlier, we cannot identify reinvestment needs.
There are some analysts who value banks by discounting their earnings back to the present. They make the argument that banks have little or no net capital expenditure needs and that working capital needs– inventory, accounts receivable etc. – are non-existent. The problem, though, is that banks need to maintain regulatory capital ratios. To see why, consider a bank that does pay out 100% of its earnings as dividends. If this firm issues no new equity, its book equity will stay frozen at current levels forever. If this bank continues to grow its loan portfolio, it will end up with capital ratios that are lower than the regulatory minimum sooner or later. That is why reinvestment has to include investments in regulatory capital, acquisitions and other such investments that banks need to make to continue to grow.
Many analysts accept the reality that estimating cash flows for financial service firms is not feasible and fall back on the only observable cash flow – dividends. While this makes sense, these analysts are implicitly assuming that the dividends that are paid out by a bank or insurance company are sustainable and reasonable. However, that does not always have to be true.
For example, Wells Fargo paid out dividends per share of $1.30 in 2008, reflecting growth of about 4% a year from 2001 to 2008. If we allow for a cost of equity for banks of approximately 9% and assume that dividends will continue to grow at 4% a year forever, we can derive the value of equity per share from a stable growth dividend discount model: (1.30 * 1.04) / (0.09 - 0.04) = $27.0. Since the stock was trading at $15.75 per share at the time of this analysis, this indicates a significantly undervalued stock. However, there are reasons to be skeptical, mainly related to the fact that current dividends might not be sustainable. The growth in dividends between 2001 to 2008 reflected the fact that Wells Fargo was going through a boom period, with net income increasing from $3.4 billion in 2001 to $8.1 billion in 2007.
Book Value
Since the majority of a bank’s assets are valued at their fair market value, rather than original cost, some analysts can argue that it makes sense to value financial firms based on their book value. This can be true, but there the process has its costs.
Assets may be marked to market, but that does remove the need to assess their value independently. The market can be wrong, and the price of some assets may not reflect their true intrinsic value. For instance, in 2008, the values of mortgage-backed securities were overstated. Also important is that the loan book portafolios can vary between banks. As such, their risk levels are not the same, and the values of the loans can deviate from what is the current market value, as some borrowers that are likely to default in the future can, at a moment in time right after making the loan, appear as a performing loan.
Discounted Cash Flow Valuation Models
We know that for financial companies, we cannot estimate net capital expenditures or non-cash working capital, and thus cannot estimate the free cash flow to equity. Based on this, we have three choices. The first is to use dividends as cash flows to equity and assume that firms over time pay out their free cash flows to equity as dividends. The second is to adjust free cashflow to equity to account for the types of reinvestment a bank needs to make, like maintaining a regulatory capital ratio. The third is to keep the focus on excess returns, rather than on earnings, dividends and growth rates, and to value these excess returns. Excess returns are earnings that a company generates above the cost of capital.
Dividend Discount Model
In the basic dividend discount model, the value of a stock is the present value of the expected dividends on that stock. While many analysts view the model as old fashioned, it retains a strong following among analysts who value financial service companies, because of the difficulties we face in estimating cash flows. In this modal, the estimate value of a business is calculated by the sum of the present values of the dividends over the growth period and the present value of the terminal value.
Looking at the inputs into the dividend discount model, there are three sets of inputs that determine the value of equity. The first is the cost of equity that we use to discount cash flows, with the possibility that the cost may vary across time, at least for some firms. The second is the proportion of the earnings that we assume will be paid out in dividends; this is the dividend payout ratio and higher payout ratios will translate into more dividends for any given level of earnings. The third is the expected growth rate in dividends over time, which will be a function of the earnings growth rate and the accompanying payout ratio.
Personally, I do not like the dividend discounted cash flow model because I think of buying a stock as buying a business, and it does not matter if the business pays out their earnings, I am still an owner of the business and thus own a percentage of them.
Cashflow to Equity Models
This second choice focuses on estimating the value of a company through a discounted cash flow model that incorporates the specific reinvestment needs of a financial company. With banks, the investment is in regulatory capital; this is the capital as defined by the regulatory authorities, which, in turn, determines the limits on future growth.
Free cashflow to Equity = Net Income – Reinvestment in Regulatory Capital
The reinvestment can be estimated from two parameters. The first is the book equity capital ratio, which refers to a bank’s equity compared to total assets. This regulatory capital ratio will depend on government regulation. It is important to note that conservative banks may choose to maintain a higher capital ratio than required by regulatory authorities. For instance, a bank that has a 5% equity capital ratio can make $100 in loans for every $5 in equity capital. When this bank reports net income of $15 million and pays out only $5 million, it is increasing its equity capital by $10 million. This, in turn, will allow it to make $200 million in additional loans and presumably increase its growth rate in future periods. The second parameter is the profitability of a bank’s lending activity, defined in terms of net income. Staying with the bank example, we have to specify how much net income the bank will generate with the additional loans; a 0.5% yield ratio will translate into additional net income of $1 million on the additional loans (based on the increase of $200 million in loans).
Personally, I like this model, it is intuitive and easy to understand, and the free cash flow to equity calculated is similar to how we think of owner earnings. Where owner earnings are the earnings left to shareholders after all necessary reinvestments have been made. It's how much cash you're going to get between now and judgment day, discounted and comparing it to other investments.
Excess Returns Model
The third model for estimating the intrinsic value of a bank is based on excess returns. In such a model, the value of a bank is the sum of capital currently invested in the firm, equivalent to the current book value, plus the present value of excess returns that the company is expected to generate.
Value of Equity = Equity Capital invested currently + Present Value of Expected Excess Returns to Equity investors
We start with book value because it is what equity share holders would get if the business was liquidated today. We then focus on excess returns because they actually add value to shareholders. A firm that invests its equity and earns just the fair-market rate of return should see the market value of its equity remain the same as its book value. A firm that earns a below market return on its equity investments should be priced lower than its equity value. Thus, if the bank is able to generate excess returns, return above the cost of equity or discount rate, then its book value should be priced higher. This is because its present value of future cash flows will be higher because of excess returns. Take for example, a bank that generates a return of 10% over $100 next year, the equity next year would be $110, but the present value today would be $100, given a cost of capital of 10%. On the other hand, if the bank is able to generate a return of 15% and has a cost of capital of 10%, then the present value would be $104, higher than the book value of $100.
Inputs in Model
There are two inputs needed to value equity in the excess return model. The first is the equity value of the company. The second and more difficult is the expected excess returns to equity investors in future periods.
While the book value of equity is an accounting measure and is affected by accounting decisions, it should be a much more reliable measure of equity invested in a financial service firm than in a manufacturing firm for two reasons. The first is that the assets of a financial service firm are often financial assets that are marked up to market. The second is that depreciation, which can be a big factor in determining book value for manufacturing firms, is often negligible at financial service firms.
The excess returns, defined in equity terms, can be stated in terms of the return on equity and the cost of equity.
Excess Equity return = (Return on equity – Cost of equity) (Equity capital invested)
When analyzing a financial service firm, we can obtain the return on equity from the current and past periods, but the return on equity that is required is the expected future return. This requires an analysis of the firm’s strengths and weaknesses as well as the competition faced by the firm.
Personally, I like this valuation model because it focuses on excess returns. If the business earns poor returns on equity, then the calculated intrinsic value will be lower than the firm’s book value. For instance, if a bank has an equity capital of $100 and earns 3% on equity and the appropriate discount rate is 10%, then in year 1 the capital would be $103, however, the present value would be $93.6, less than the equity book value of the bank. On the other hand, if it earns superior returns, then it will be worth a premium over book value. As Buffet puts it “book value is not key when valuing banks, earnings are” [2]. Problems of this valuation include its dependence on calculating a cost of capital and that it is not very intuitive.
Relative Valuation
Finally, the multiples that we can use to compare bank valuations in the same markets are price earnings ratios and price to book value ratios.
References:
[1] https://docs.google.com/document/d/1nTGpujiXcpKfmGW2pvmGulxHVHUoFIT3l0fOTqgCkaQ/edit
[2] https://www.cnbc.com/video/2013/03/04/earnings-are-key-to-valuing-banks-buffett.htm