# Mohanram G-Score Overview Part 1

While listening to the The Acquirers Podcast this past week, I discovered Partha Mohanram’s G-Score analysis. Inspired by the Piotroski F-Score, Mohanram set out to create a similar stock fundamental algorithm to separate out the winners from the losers in low book value to price stocks that he indicates are usually growth stocks, but others might just call them glamour stocks.

The G-Score has eight factors that can be grouped into three main growth stock signals. Those factors are as follows:

- G1 is defined to equal 1 if a firm’s return on assets (ROA) is greater than the industry median and 0 otherwise
- G2 is defined to equal 1 if a firm’s cash flow ROA exceeds the industry median and 0 otherwise
- G3 is defined to equal 1 if a firm’s cash flow from operations exceeds net income and 0 otherwise
- G4 is defined to equal 1 if a firm’s earnings variability is less than the contemporaneous industry median and 0 otherwise
- G5 is defined to equal 1 if a firm’s sale growth variability is less than the contemporaneous industry median and

0 otherwise - G6 is defined to equal 1 if a firm’s R&D deflated by beginning assets is greater than the contemporaneous industry median R&D intensity
- G7 is defined to equal 1 if a firm’s capital expenditure deflated by beginning assets is greater than the contemporaneous industry median capital expenditure intensity
- G8 is defined to equal 1 if a firm’s advertising expenditures (I used selling, general & administrative expense for the backtest as a close proxy) deflated by beginning assets is greater than the contemporaneous industry median capital expenditure intensity

The first three signals (G1, G2, and G3) are based on earnings and cash flow profitability and are similar to some of the F-Score factors. The next two signals (G4 and G5) measure variability in earnings and sales. The idea here is that the market does simple extrapolation for forecasting earnings and sales, so forecasts of highly variable companies are likely to disappoint. Research has also shown that companies with more stable earnings tend to have better earnings performance in the future. The final three signals (G6, G7, and G8) are related to accounting issues that undervalue investments in R&D, advertising, and capital expenditures. These items are usually treated conservatively by accounting practices and result in potentially lower than real book value.

Mohanram used these 8 factors on the 20 percent of stocks with the highest price-to-book value to develop long-sort portfolio made up of long positions on high P/B stocks with G-Scores of 6 and above and short positions on the remaining stocks with G-Scores of 1 or 0. He found that firms with higher G-Score earn higher returns than firms with low G-Score. Mohanram also found that G-Score was even better at picking out losers among “growth” stocks. The “high” G-Score stocks returned an average of 17.4% per year during the study period of 1979 to 1999 and the combined long-short portfolio returned an average of 21.4% annually. You can review all the details of the research in Mohanram paper, “Separating Winners from Losers Among Low Book-to-Market Stocks Using Financial Statement Analysis.”

I’ve started my own backtest of the G-SCORE using Portfolio123. I’ll post my results shortly. Be sure to subscribe by email or by RSS feed in order to be notified when the next article is published.

Continue reading **Mohanram G-Score Backtest Part 2** of this series by clicking on the link.

Thanks very much for this thread and series. I’m also a portfolio123 subscriber and was trying to build out the G-score for testing.

Would you be willing to share the formula for it on p123 or via email?

Hi Mark,

I’d be more than happy to share the G-Score formula with you. I’ll send it via email shortly.