Fact, Fiction, and Value Investing

Clifford Asness et al., The Journal of Portfolio Management, vol. 42, no. 1 (Fall 2015): 34–52

Many investors have different opinions about the merits of value investing, despite it having been a relevant investment strategy for almost a century. The authors separate value investing “fact” from “fiction,” making the case that the strategy is a cornerstone of good investing, particularly when combined with other factors.

Fiction: Value investing is an idiosyncratic skill that can only be implemented with a concentrated portfolio. An idiosyncratic, concentrated strategy and a diversified, systematic strategy can both deliver the value premium. In the authors’ view, the latter is available to all investors and generates more persistent returns with less downside risk at a lower fee.

Fact: Value investing is applicable to more than just equities. Value investing’s goal is to identify cheap assets, not cheap equities. Using various measures, the value premium has been identified in bonds, equity index futures, commodities, currencies, and global equities. The research even points to a positive correlation of value strategies across asset classes.

Fact: Value is best measured by a composite of metrics. While book value–based metrics are a leading way to measure value, an average of multiple measures (i.e., book value, earnings, and cash flow) delivers a more stable portfolio, with a higher Sharpe ratio. The reason is simple—an average of measures reduces the error associated with each individual measure.

Fact: By itself, value is surprisingly weak among large-capitalization stocks. A review of data across four sample periods reveals no strong stand-alone value premium among large-cap stocks. However, value can still be beneficial for large-cap stocks—combining it with momentum produces similar market-adjusted returns as the same strategy for small-cap stocks.

Fact: Profitability can be used to improve value investing. Adding profitability, or other measures such as momentum, can enhance value strategies. These measures are highly negatively correlated with value and can add significant diversification benefits.

Fiction: Value is the result of a risk premium, not a behavioral anomaly. Whether the value premium exists because of a risk- or behavioral-based explanation is up for debate. The authors believe both theories hold some truth. In either case, a century of evidence points to the existence of a value premium in the future. Even if the value premium disappeared, the diversification benefits of combining it with other factors, such as profitability or momentum, would still make it a valuable investment tool.

Clairvoyant Value and the Growth–Value Cycle

Robert D. Arnott, Feifei Li, and Katrina F. Sherrerd, The Journal of Portfolio Management, vol 35, no. 4 (Summer 2009): 142–157

Investors often seek to distinguish companies with higher growth potential and have been willing to pay a premium for their stocks. However, the premium paid can frequently be too much when compared to subsequent returns. The authors suggest investors use this fact to their advantage—when the growth stock premium is high, value stocks may be poised to outperform.

To determine how growth and value stocks perform, the authors act as a “clairvoyant” investor, using hindsight to determine what premium growth stocks should have been valued at over time. Based on their data, the authors determine that growth stocks were on average only worth 80% of the premium investors paid, when measured by ten-year subsequent returns. When measured by 20-year subsequent returns, growth stocks look even more overpriced.

The authors develop a clairvoyant value portfolio to understand the degree to which performance is better than market capitalization–weighted and size-weighted returns over the same period. While the authors unsurprisingly find the clairvoyant portfolio performs the best, they discover that the market-cap portfolio performs the worst. Market cap–weighted portfolios weight stocks based on the total market value of each company’s outstanding shares, leading to higher weights for stocks priced above their fair value. None of these portfolios perform well during bubbles such as the early 1970s Nifty Fifty bubble and the 1999 tech bubble.

In a perfect world, the growth stock premium would be linked to the market’s ability to accurately discern future performance, meaning investors would be compensated for the premium by subsequent outperformance. However, the authors use historical data to conclude that the premium paid for growth is largely a factor of investors’ confidence in their ability to forecast long-term growth, not their actual ability to forecast. Investors can use the market’s overconfidence to their advantage. When the growth stock premium is low, growth stocks are more likely to outperform, while value stocks are more likely to outperform when the premium is high.

Growth/Value, Market Cap, and Momentum

Jun Wang et al., The Journal of Investing, vol. 23, no 1 (Spring 2014): 33–42

Investing styles play an important role in the performance of diversified portfolios. The authors combine momentum with various valuation-oriented and market-capitalization strategies, arguing that their analysis of performance shows momentum complements the other strategies in various market conditions.

The authors create momentum portfolios utilizing valuation-oriented (growth, blend, and value) and market-cap (large cap, mid cap, and small cap) style indexes. The portfolios buy the best-performing style and sell the worst-performing style, aiming to capitalize on trends. The authors use time horizons of one, three, six, nine, and 12 months to select the winner and loser and to hold the investments, making a total of 25 momentum portfolios.

The results from their analysis of monthly returns suggest that each of the 25 momentum portfolios is generally less risky than the individual style indexes alone or the corresponding winner or loser alone. Of the time periods, the returns are highest over a three, six, and nine month horizon. The authors also note that the returns linked to momentum are mostly due to the long positions, which tended to have smaller standard deviations.

Based on their results, the authors argue investors can be profitable when rotating their portfolios based on recent style performance. Although transaction costs are a headwind, the same style index can persist as either a winner or a loser for several consecutive time periods, resulting in transaction cost savings. The correlation between the momentum portfolios and the S&P 500 Index is low, if not negative, indicating that momentum may make a good hedging tool during economic contractions. The authors also find their momentum portfolios are not impacted by seasonality, including when investors sell losing investments at year-end for tax purposes.