Before continuing exploring the possibilities of our new tool, the Texas Ratio, I think it is important to have some historical perspective on the risks of bank failure. Banks are quite fragile leveraged institutions and bursts of financial failures are essential to consider for normal times bank investing. I'd much rather invest in banks at the end of crisis when the risks can be quantified and valuations are cheaper, but there is a case to be made for invest in them in any case. (click to enlarge)
The extension of risky loans with overpriced collateral, thin covenants, to lenders with insufficient cash flow, and even in some cases becoming complicit or target of frauds, can lead to over leveraging a particular sector of the economy and reaching unsustainable bubble prices. In the old times, the boring banking times of the 50s and 60s, those four condition were called the 4Cs of lending with fraud avoidance called character. Even in the old times, there were periods when these standards were breached and problems ensued.
But as we can see from our most recent history, something different is going on. There are different views on the reasons, but I would like to emphasize the rise of credit scoring that was a much less costly and standardized way of lending than checking the character of a borrower. That in turn opened the possibility for the rise of securitization, that is not as new as people think: the 1980s collateralized junk bond obligations or 1990s car loans securitization being just two examples.
In theory, securitization helped banks to reduce risky assets in their balance sheets. But as we have learned it is not so simple: there is less control of potential massive scale frauds, and asset bubble deflation will nevertheless impact more traditional loans so that banks can sometimes end up holding junk like in the S&L 1980s crisis. It all started with the best of intentions but we have to live now with the consequences.
A pile of junk is still junk no matter how you stack it – Margin of Safety, Seth Klarman