What Really Drives Bank Stocks

What's the secret behind the best bank stocks?

Is it profitability?

Asset growth?

Revenue maximization?

It turns out, it's none of the above.

The variable that correlates most closely with high returns is the amount a bank provisions for future loan losses. That is, how well it manages credit risk.

You can see this in the figure below, which plots the provisions for loan losses against the compound annual returns at the nation's 18 largest banks from the beginning of 1995 until the end of 2012.

It doesn't take a statistician to see that there's a strong correlation between these two variables. While correlation is not causation, it's hard not to conclude from this relationship that lower loan losses are one of the primary ways (if not the most important way) to boost shareholder returns.

With a coefficient of determination of 0.7318 (this measures the "fit" of the data to the trend line on a scale from zero to one), the relationship between these two metrics was more closely aligned than that between CAGR and all of the more commonly assumed links, such as revenue as a percent of assets, asset growth, and even return on assets.

Source: S&P's Capital IQ. *All data covers the period 1995-2012. **Loan loss provisions and operating expenses are both expressed as a percentage of pre-provision net revenue. ***Revenue relates to pre-provision net revenue as a percentage of average assets. 

The takeaway for investors is both straightforward and implementable: If you invest in bank stocks, the first thing you should examine is a bank's history of loan losses.

Doing so over the time period examined above would have led you to the likes of New York Community Bancorp (NYCB), M&T Bank (MTB 2.93%), and Wells Fargo (WFC 2.74%). These three banks proved themselves to be superior risk managers, with loan loss provisions ranking them at the top of their class, and, as a result, exceptional investments.

For the 18 years included in the analysis, these banks notched compound annual returns of 18%, 14%, and 13%, respectively. The average, meanwhile, was less than 8%. And many of their less prudent competitors like Citigroup, Regions Financial, and Huntington Bancshares, turned in single-digit figures that were closer to zero.

Does growth matter? Sure -- though, at least in the banking space, it matters far less than most analysts assume. Should banks try to maximize revenue? Yes -- though, again, within reason.

Finally, though, should they pursue either to the detriment of credit risk? Absolutely not. That's a recipe for disaster, and it's one you're now armed to identify and avoid.