Tuesday, March 29, 2011

Breach of Trust: Bank Valuation after the banking crisis

Until the banking crisis of 2008, investors had made a Faustian bargain, when it came to valuing and investing in banks. Banks were opaque in their public disclosures and investors often had little information on either the risk of the securities held or the default probabilities of loan portfolios. However, investors were willing to accept this opacity and view banks as "safe" investments for two reasons:
  1. Banks were regulated in their risk taking: In effect, we were assuming that bank regulators would bring enough scrutiny to the process to prevent banks from taking "rash" risks. (We also assumed that the regulatory authorities had access to far more information that we did and would act accordingly.)
  2. Assets (and equity capital) were marked to market: The notion of marking to market was adopted much more quickly in financial service firms than at other sectors. Our distrust of accounting notwithstanding, we assumed that the book values for banks actually were good reflections of market value.

How did this faith in the regulatory overlay get reflected in valuation/investing?
  • In intrinsic valuation, banks remained the last holdout for the use of the dividend discount model. Unlike other companies, where our distrust in managers paying out what they could afford to had led us to move on to free cash flows, we retained the faith that bank managers, constrained by the need to meet regulatory capital constraints on one hand and "dividend seeking" investors on the other, would pay out what they could afford to in dividends. (In effect, banks that paid too much in dividends would be punished by the regulators and those that paid too little in dividends would be punished by investors.) 
  • In relative valuation, the book value of equity in a bank was given more weight than in other sectors, because it was marked to market and subject to regulatory capital rules. Thus, price to book ratios (with returns on equity as companion variables) were widely used in analysis: a bank with a low price to book ratio and a high return on equity was viewed as a bargain. Worse still, risk averse investors were asked to buy the highest dividend yield banks and assured that these yields were secure.
So, what's changed? First, our faith in both bankers and regulators has been shaken, perhaps to a point of no return. We can no longer assume that having regulatory rules on risk taking will result in sensible risk taking at individual banks. There can be, as there are in other sectors, very risky banks, risky banks, safe banks and very safe banks, as a consequence. Second, the erratic and often ill-thought out dividend policies adopted by banks since the crisis indicates that bank managers, at many banks, use dividends as a blunt weapon. How else can you explain banks with precarious capital ratios that continue to pay and increase dividends, while raising fresh capital in preferred stock at the same time? In fact, it is a sign of the times that the Fed  stepped in to stop a major money center bank from paying dividends, as it did with Bank of America a couple of weeks ago.

So, what do we do now? In intrinsic valuation, we have two choices.
1. One is to use a modified version of the dividend discount model, where we estimate future dividends based upon expected growth and the return on equity that we foresee for a bank, rather than the actual dividends in the last period. Thus, if a bank is expected to grow at 8% and has a return on equity of 10%, it an afford to pay out only 20% of its earnings as dividends:
Payout ratio = 1 - Expected growth rate/ Return on equity
Thus, we can bring in both the quality of a bank's investments and expected changes in regulatory capital rules into the valuation. Increases in regulatory capital requirements will reduce the return on equity and by extension, the capacity to pay dividends.
2. The other and more complicated route requires knowledge of regulatory capital requirements and involves the following steps. You first estimate the growth in the asset base of the bank (growth in loans, for instance). You then follow up by estimating how much regulatory capital will be required to sustain the asset base - that will depend upon the risk in the asset base and the regulatory capital ratio that the bank wants to maintain. (Note that this ratio will not necessarily be at the regulatory minimum since conservative banks will maintain a buffer.) Changes in regulatory capital from period to period than take on the role that capital expenditures do in a more conventional firm and can be used to compute free cash flows to equity:
FCFE for a bank = Net Income - Change in Regulatory capital required for future growth
These FCFE are potential dividends and can be discounted to arrive at fair value. In fact the cost of equity for a bank can then be tied to its regulatory capital buffer: banks that build in a bigger buffer will be safer and have a lower cost of equity whereas banks that are more aggressive in both their asset holdings and regulatory capital policies will have higher costs of equity.

In relative valuation, I think that the use of price to book ratios, in conjunction with return on equity, still makes sense, but risk now has to be treated as a third dimension. The risk itself can be measured using a variety of measures: regulatory capital ratios (higher ratios are safer), losses on bad loans (higher is riskier) or holdings of toxic securities (higher is riskier). A bargain bank will then be one that trades at a low price to book ratio, has a high return on equity and is well capitalized. I expand on both notions in this paper that I wrote a couple of years ago on valuing banks (which subsequently became a chapter in one of my books):
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1798578

I think that there are broader policy implications.
  1. More transparency in financial statements: Since banks have broken their side of the bargain with investors, we need to respond by removing the opacity from the financial statements of banks. Banks should be forced to provide far more detail about the riskiness of their security holdings and the default risk in the loans that they make. Much more information needs to be provided about regulatory capital requirements and the policies that banks adopt on regulatory capital should be more transparent.
  2. Regulatory capital has to be common equity: Banks that are under capitalized should be required to issue common stock, and face up to their fears of dilution. We need to scrap the notion that preferred stock (a tax-inefficient mismash) or convoluted hybrids (such as these) will be treated as equity, since it exposes us to game playing and worse.

I am not ready to give up on investing in banks. In fact, I am sure that some banks are great bargains and the payoff to finding these, in this time of greater uncertainty, is higher than ever before. But I will be more careful in my assessments of banks and not take numbers for given, just because they have been rubber stamped by regulators and appraised by accountants. That is more a promise to myself than to you!

16 comments:

Mike said...

Well, what are the chances of such policy changes when the Wall Street insiders become the policymakers or close pals of policymakers?

Remember, there is a revolving door between Wall street and Washington. May be it is not apparent to the academic world, Is it?

Gaurav Mehta said...

It would be great if you could take a valuation of a bank such as Wells Fargo, Goldman and a Risky bank like Lloyd TSB and Citi in your Valuation case studies through the 2nd more complicated method rather than the normal DDM method.

Aswath Damodaran said...

Mike,
I think you are making my point. As investors, we cannot wait to be protected by regulators or the government. We have to assume that regulated firms do some of the same stupid things that unregulated firms do and thus assess them with the same diligence.
Gaurav
I do have a valuation of Deutsche Bank using the more complicated method...

robert said...

Please address how suspending mark-to-market rules on bank assets affects this discussion. Also, did the suspension apply to all assets or only MBS? Thanks

Aswath Damodaran said...

The suspension of mark to market rules is only some securities, not all. From a valuation standpoint, the effect is minor, partly because we are cash flow focused,rather than book value focused.
The big danger is that the suspension of the rules may also allow truly troubled banks to stay undetected for a lot longer. It thus increases the onus on investors to make their own risk assessments.

Gaurav Mehta said...

ok... will have a look at the current year case studies... Thanks

Krishnan said...

Prof
With Basel III kicking in, and both US & European banks short of capital, is the banking sector a wise place to be? So banks would be forced to hold more capital, they may do so through preference stock or a subordinated debt. Historically the insurance companies have been buyers of these type of hybrid securities, rules are getting strict for them as well, so who would subscribe to these sort of instruments?
Any thoughts?
Thanks
Krish

PPS said...

Dear Sir,

On a slightly different note, we find alongwith PE ratio EV/Ebitda is used extensively for comarison of valuation of companies (measure of cheapness/expensiveness). However, EV takes in to account book value of debt and not market value.

With companies having different leverage and paying different interest rates doesn't it misguide more than guide regarding comparative valuations? PE ratio would mean the dollars one is willing to pay for 1 dollar of earnings. Wouldn't taking Market value of Equity and Book value of debt for calculation of EV confuse the concept of EV.

My guess was that Book Value of debt is used in equity research reports because it can be easily calculated from the Balance Sheet while calculation of market value would be cumbersome. Do yo think it would be interesting to find some ratio/metric which is a bit closer to Market value of debt and can be calculated using PL statement and BS of a company only?

PPS said...
This comment has been removed by the author.
Aswath Damodaran said...

While EV multiples are almost useless for banks, they are used extensively for other companies and I do think that market value of debt is a more reasonable value than book value. I guess that the defense that analysts would offer is that since the market value of debt is generally lower than book value, the use of book value is conservative. I don't buy that... Conservative is not necessarily a plus in assessing value. It is a virtue when making your investment decision, where you compare price to value.

Ritesh said...

Dear sir,

you wrote that the conservative banks, who hold higher capital than required should be at premium (because of lower risk involved) than the banks which are aggressive.
But, at the same time these conservative banks will be loosing on the other front as their WACC (Normally) will be on higher side because of high weight of equity.



i think we should consider this point while valuing a Bank.

Ehsan Rafizadeh said...

Dear Professor,

This is a really interesting post; I’m wondering how this effects target valuation in a merger and acquisition situation as the "Deal Value to Book Value" is the key comparable in financial sector and an important determinant of premium price paid? Have investment banks adjusted their valuation methods?
Thanks.

Ed said...

Thanks a lot for this interesting post. Could you please point me towards your valuation of Deutsche Bank using the more complicated method? I could not find it on your website. thank you.

Santiago Camilo said...

Prof Damodaran,

Given the bad performance of european banks lately, and the low price-to-book ratios, I have been drawing a few conclusions:
1. Banks' management clearly feel, ans show in their books, that their assets and equity have a higher price than what the market considers. From that point of view, if a Bank is trading at 0.5 price-to-book, and the Bank's CEO has accurately valued his BS, if he decides to liquidate, would he be doubling shareholders investment?

krishna said...

Two banks in Nepal are going to merge and i am currently valuing a bank. let say it A and Bank B. According to our central bank the bank must increase their paid up capital to 2 billion by next year Bank A already has 2 billion and bank B doest not have . Bank B is 700 million short of 2 billion. Bank B cannot raise capital by issuing right share because or market not positive. so they are merging.

i am using excess return model to value. i forecasted balance sheet and income statement of tw banks and came up with a certain value.

Now how do i compensate for that less 700 million.

Radhika said...

Professor

This is not strictly related to the blog post, but if possible could you shed some light on using the DDM model for companies that pay interim dividends?