Robots Suck at Value Investing

Value investing, a style of stock picking that dates back to Benjamin Graham and David Dodd’s 1934 tome, ‘Security Analysis,’ can’t be mimicked by a simple algorithm.

That’s one of the findings of a revised study entitled ‘Facts About Formulaic Value Investing’ that was published this month by the CFA Institute. The research found that portfolio managers need to use a comprehensive set of metrics to calculate a company’s ‘intrinsic value,’ as laid out by Graham and Dodd. They then need to add human judgment to the mix to find undervalued stocks that will outperform so-called glamour or growth stocks over a market cycle, a pattern of returns that academics have called the value premium.

“If you go back and read Graham and Dodd, they advise against investing simply based on ratios such as market price-to-book value. They only use these measures as a starting point,” Richard Sloan, a professor of accounting and international business at the Haas School of Business at the University of California, Berkeley, and one of the authors of the study, said in an interview.

The revised research was published this month by the CFA InstituteFinancial Analysts Journalamid the rising popularity of value style, factor-based investments that use a set of rules to buy and sell stocks. Sloan says factor-based investments aren’t delivering outperformance, in part because they’ve been designed based on back testing of past information.

“Those back tests emphasize time periods when they worked well, but they haven’t worked well since,” he said in an interview. The word ‘value’ is increasingly attached to quantitative investment funds, but investors aren’t getting what they think, according to Sloan.

The researchers looked at commonly used measures in value investing that don’t necessarily lead to stocks that outperform. For example, they found the book-to-market ratio generally identifies stocks with inflated book values that are written down over time, and that the forward earnings-to-price ratio often points to companies that brokerage analysts are bullish on because of future earnings expectations. They also analyzed the trailing earnings-to-price ratio, finding that it systematically identified securities with temporarily high earnings, setting the stage for losses.

“They all have the same problem. In recent years, there is no evidence of outperformance,” said Sloan, adding that there’s a “labeling issue” when it comes to value-style, factor-based investing today.

While Graham and Dodd identified underpriced securities using fundamental analysis, he said “formulaic factor-based funds appeal to that, make reference to it, but they don’t deliver that.”

U-Wen Kok, chief investment officer of the developed markets team at RS Investments, and a co-author of the study, points to fee compression in the asset management industry and advanced technology as drivers behind the rise of factor investing. RS, which is owned by Victory Capital Management, offers factor-based funds, as well as those managed by humans.

“This paper emphasizes the importance of not just going with a quant screen or simple model,”Kok saidin an interview. “You need human insight.”