Friday, May 4, 2018

Unusually Fine Advanced PBV Analysis

When valuing financial service firms, the Price to Book Value (PBV) ratio is possibly the most commonly used valuation method by analysts because it's safe, simple, has a strong historical basis and requires no IQ at all.

Back testing the PBVs to stock price has a good correlation. PBV needs however be looked at in the context of return on equity (ROE). Historical analysis has shown that ROE has a strong impact on banks’ value creation in the long run. Bottom line: Firms that have high price-book value ratios should also have high returns on equity. Firms that have low price-book value ratios should have low returns on equity.

While emphasizing this correlation between PBV and ROE, one should not ignore other fundamentals such as political instability, global regulation changes, IT attacks, manager scandals etc. which can all be categorized under "risk". For any given ROE, riskier firms should have lower PBVs. Similarly firms with much greater potential for growth and less risk should have higher PBVs.

I have a world-class solution, which I proudly found elsewhere and marshaled all my calculus skills and computational resources to put it to the test. Off the top of my head I listed some major banks together with their current PBVs and ROEs (TTM). As a proxy for the risk I used Beta. I could have worked with Standard Deviation in stock prices for the last 52 weeks; but Beta was computation-friendly. Took Betas from FT; PBVs and ROEs both from Morningstar. Both services are pro-bono. Regressing the PBVs against the ROE and Beta yields the following:

PBV Ratio = 1.100 + 0.033 * ROE - 0.235 * Beta

R-squared = 24%

I then used this regression to calculate the predicted PBVs for all the banks in the sample and finally compared to their current PBVs to check whether they are under- or over-valued: