A Systematic Framework For Selecting Financial Stocks

I’ve thought for a long time about better ways to select stocks in the financial sector. Many screens explicitly or indirectly exclude financial stocks. It’s become a somewhat popular chorus in recent years that financials are too risky or too opaque to invest in. However I do not take this view, I’m of the opinion that no asset class or sector is inherently unattractive. There is plenty of money to be made in analyzing financial stocks judging by the portfolios of some of the more renown value funds such as Arlington Value and Tweedy Browne. 

I’ve settled for customizing Piotroski’s Fundamental F-Score as a way of quickly and efficiently sorting promising financial companies from the ugly ones. Recall that the original F-Score is a 9-point scoring system used to screen promising stocks. For each of the 9 measurements that improved over the last year, a company would score a 1; otherwise it would get 0. With scores ranging from 0 to 9, companies with low scores were unattractive and companies with higher scores were attractive. The F-Score alone was found to produce good results historically. However, financial companies are explicitly screened out. Indeed, a lot of the metrics would not apply to financial sector stocks anyways. Therefore I removed from consideration the following tests that I felt were irrelevant to the sector.

- Positive Free Cash Flow

- Current ratio improves over the previous year

- Gross margin improves over the previous year

- Total asset turnover improves over the previous year 

I kept the remaining 5 tests and added one more of my own to capture improvement in overall leverage (from having thrown away the current ratio test), therefore making it a 6-point scoring system, with the following tests:

1) Net income is positive, score 1, otherwise 0
2) Return On Assets latest year is greater than ROA previous year, score 1, otherwise score 0
3) Operating cash flow is greater than Net income, score 1, otherwise 0
4) Ratio of long-term debt to total assets latest year is equal or less than Ratio of long-term debt to total assets previous year, score 1, otherwise 0
5) Shares outstanding latest year is equal or less than Shares outstanding previous year, score 1, otherwise 0
6) (New) Financial leverage (total assets divided by total equity) latest year is equal or les than Financial leverage previous year, score 1, otherwise 0

Using Portfolio123’s screen building tool, I backtested groupings of this modified Piotroski score against an equal-weighted portfolio of all the Financial sector stocks in the total market S&P 1500 index. Portfolios are formed every 4 weeks and held for 1 year from June 1999 to June 2015. 

 

 

While the annually rebalanced portfolios of Financials stocks averaged an annual return of 6.75%, portfolios of stocks with a modified Piotroski score of 5 or 6 (out of 6) returned an average of 8.64%, portfolios of stocks with scores of 3 or 4 averaged 6.54%, while portfolios of stocks that scored 2 or less only returned 4.20%. This suggests that there is a good way of systematically choosing stocks in the financial sector. There might be other systematic methods, and some other criteria should likely be applied, namely the nature of the business, low valuation in relation to earnings or book value, etc. But this is a good, simple starting point for researching financial stocks and is the first criteria I use for selecting and recommending companies in that sector.