Thursday, August 25, 2011

The fair value assessment implied from the recent Warren Buffett BAC deal

Warren Buffett recently entered a deal to buy $5 Billon BAC 6% cumulative perpetual preferred stock, plus free warrants to buy $5 Billion BAC common stocks at $7.14 within 10 years.

From this deal, we can ROUGHLY estimate Mr. Buffett's assessment of BAC's fair value.

First, the preferred stock is clearly a bad deal. 6% is not a high return, plus it is "perpetual". So it has high interest rate risk since interest rate could go a lot higher later. So the warrants must provide good compensation to make the overall deal good.

There are many ways to compute this. An easy way would be considering the return on the combination of preferred stock and common stock.

Assumptions:
1. Assume 10% return is a fair return in this case for both preferred stock and common stock (banks will be much safer after the new BASEL III capital ratio requirment, so 10% return is good enough. My requirement on returns are usually higher than the market. Banks used to have a P/E of 10 in the pre-2008 market, but I think only now a P/E of 10 is justified.).
2. Assume BAC will not buy back the preferred stock forever (at least for the next couple of decades)
3. Buffett will not excise his warrant until the end of 10 years (he has no reason to do so in general, after all, warrant is capital-free investment, like long term call options. This is common sense, we don't need Black-Scholes formula to tell us about this).
4. Assume Buffett will require a margin of safety of 33%. (33% is really just my own preference, but this figure was also mentioned as an example in Ben Graham's book)

First, in the next 10 years, preferred stock will be losing 4% each year. At the end of 10 years, it will be losing 0.04 * 1.1^9 + 0.04 * 1.1^8 + ... + 0.04 * 1.1 + 0.04 = 0.04 * (1 - 1.1^10)/(1-1.1) = 0.636 of $5B.

Now we add this amount to the effective cost of the eventual purchase of common stock at the end of 10 years, which gives an effective price of 7.14 * (1 + 0.636) = 11.68 per common stock. At the end of 10 years, Buffett effectively spent $8.18B to buy the common stock. This amount has to earn enough annual return to compensate the 6% preferred stock's low return. It has to earn an extra 2.444% return on top of normal 10% return (5 * 0.04 / 8.18 = 2.44). So that means the fair value at that time would be 11.68 * 12.444 / 10 = 14.53 per share.

Now discount this price back to present using 10% discount rate, it gives 14.53 / 1.1^10 = 5.61 per share.

Buffett would always require a margin of safety, if this margin of safety is 33%. We get the real fair value of 5.61 / 0.6666 = $8.41 per share.

BTW, there is a wild card here: since preferred stock no longer counts in the capital ratio under BASEL III, and the dividend is not tax deductible, banks have a tendency to buy back preferred stock and issue bond instead. So it still has a possibility to buy back the preferred stock even if it has a low interest rate. When they do it, they have to pay the 5% premium to Buffett. If Buffett did his calculations with this in mind, it would actually push down the fair value (we get a lower fair value than $8.41).

Since there are many moving parts, this has to be just a rough estimate. The actual figure in Buffett's mind could be wildly off from this number since a small change on the assumptions would yield a big difference (for example, if we require 11% return or 9% return instead). That said, this analysis gives some insight to the fair value, or at least a wide range of it.

My own analysis shows that bank stocks are depressed mainly because they need to accumulate capital to meet the 9.5% tangible common capital ratio under the new BASEL III rule. Although they have no urgency to do so until 2019 (thus no need to issue new stocks now), it will eventually reduce returns to shareholders and reduce ROE along with future growth. Regulations are not fully defined yet, people are worried about the impact of regulations, such as price fixing (like the new debt card fee limit) or reduced efficiency (higher operational cost). On the litigation side, there is some risk, but I don't think it will be the major impact.

Among the four biggest banks (JPM, BAC, WFC, C), BAC has the lowest capital ratio and largest risky asset which requires even more capital. It would take many years to reach the required capital ratio, so we have to discount its share price based on that fact. However, the fear on regulation and litigation may be still overblown.

On the positive side, higher capital ratio will make large banks much safer and in turn reduce the required cost of capital. Also, it would increase the barrier of entry and reduce the competition in some degree.