A New Magic Formula? How is the old one working?

Joel Greenblatt was made famous by his so-called “Magic Formula” for stock picking, which he describes in his book, “The Little Book That Beats The Market”. Greenblatt claims in the book that by purchasing the market’s top 30 stocks that conform to the formula with a high earnings yield and a high return on capital and at the same time shorting the market’s most inefficient and expensive stocks, he beat the S&P 500 96% of the time and achieved an average 17 year annual return 30.8%.

Joel Greenblatt
Image source: YouTube Video Screenshot

 

Ever since the book was first published, Wall Street has questioned whether or not these returns are possible. A new research paper by Douglas W. Blackburn and Nusret Cakici at Fordham University titled “The Magic Formula: Value, Profitability and the Cross-Section of Global Stock Returns” asks this question and arrives at an interesting result.

Is This The New Magic Formula?

The paper tests Greenblatt’s Magic Formula for its ability to generate positive risk-adjusted abnormal returns and to explain the cross-section of returns across global developed stock markets. The authors construct a portfolio that is long the quintile portfolio of profitable, value stocks and is short of the quintile portfolio of unprofitable, growth stocks (profitability is defined as the return on capital EBIT-to-tangible capital employed and value is defined using the earnings yield of EBIT-to-enterprise value. This portfolio is replicated for the four global regions of North America, Europe, Japan, and Asia using the stock return data representing 23 developed markets over the period January 1991 through December 2016.

This study, as the authors write, yields both good and bad news.

Firstly, the backtest of equity data shows that the Magic Formula, as described by Greenblatt in 2006, is “not very magical”.

 

New Magic Formula“Return differentials from portfolios sorted using the Magic Formula methodology are insignificant as are the risk-adjusted alphas from the Fama-French-Carhart model for North America, Japan, and Asia. Return differential are only significant in Europe.”

The good news is, that by making just one small adjustment, the Greenblatt Magic Formula can be resurrected. By simply replacing EBIT, with gross profits, returns from the strategy become “much more magical.” When EBIT-to-tangible capital employed becomes gross profit-to-tangible capital employed and EBIT-to-enterprise value becomes gross profit-to-enterprise value, profitable value stocks significantly outperform unprofitable growth stocks with return differentials of 0.88%, 0.65%, 0.66% and 0.62% from North America, Europe, Japan, and Asia respectively. Further, as well as yielding significant risk-adjusted abnormal returns, these new magic portfolios have significantly negative market betas.

For example, during 2008 when regional value market portfolios declined between 26.52% and 50.2% depending on the region, the long/short, adjusted magic portfolio gained 21.11% to 64.52% depending on the area. Moreover, a portfolio comprised of 50% the value -weighted market portfolio and 50% the adjusted magic formula long/short portfolio leads to substantially larger Sharpe ratios due to diversification brought about by the observed negative correlation. For North America, Europe, Japan, and Asia the Sharpe ratios for the 50/50 portfolios are 0.97, 1.02, 0.43 and 0.62 respectively, compared to the market Sharpe ratios of 0.50, 0.39, 0.10 and 0.34 respectively. These results show that by “tilting the value -weighted portfolio toward inexpensive, profitable stocks away from expensive, unprofitable stocks one can actually beat global markets.”

New Magic Formula Source: Blackburn, Douglas W. and Cakici, Nusret, The Magic Formula: Value, Profitability, and the Cross Section of Global Stock Returns (April 14, 2017). Available at SSRN: https://ssrn.com/abstract=2956448