Research Review: Quality Investing – Robert Novy-Marx, 2013
posted on November 17th, 2015 by Bellmont Research Team
“Buying high quality assets without paying premium prices is just as much value investing as buying average quality assets at discount prices. Strategies that exploit the quality dimension of value can be profitable on their own, and accounting for both dimensions of value yields dramatic improvements over traditional value strategies.” So says Robert Novy-Marx – the Berkely trained financial economist and fellow at the National Bureau of Economic Research. Novy-Marx is effectively the first high profile economist to prove in theory, what seasoned investors have long known in practice – that quality matters!
This assertion of the importance of quality should really come as no surprise to avid disciples of Warren Buffett, who famously developed from a die-hard value investor, chasing Benjamin Graham style ‘cigar butts’ (cheap stocks that are acknowledged to only have one puff left in them – but that one puff is all profit!), to an enthusiastic proponent of quality investing, who now repeatedly exclaims that it is “far better to buy a wonderful business at a fair price, than to buy a fair business at a wonderful price.” Yet, for decades academics had not acknowledged the importance of quality in their asset pricing models, with many struggling to find a direct link between traditional quality metrics such as Return on Equity (ROE), or Net Profitability and subsequent stock returns.
Yet Novy-Marx’s study shows startling results, with a dollar invested in the market (NYSE) in 1973 growing to over $111 at the end of 2013, whilst the same dollar invested in stocks that met the dual criteria of quality and value, grew to $595 over the same period. So what were the unique insights that Novy-Marx brought to the table, that led to such a wholesale re-assessment of the importance of quality?
In order to thoroughly investigate the potential of quality from an investment standpoint, Novy-Marx looked at seven of the best known and most widely used indicators of quality. These include measures proposed by Benjamin Graham (Warren Buffett’s teacher and mentor), Jeremy Grantham (of GMO) and Joel Greenblatt (The Little Book that Beats the Market), as well as Sloan’s (1996) measure of earnings quality, Piotroski’s (2000) F-Score measure of financial strength, a low volatility measure, and Novy-Marx’s own gross profitability measure (full details of each are in the paper). For the period 1963 to December 2013, Novy-Marx then examined both a long-only portfolio, which features the top 30% of NYSE stocks on each measure, and a long-short portfolio, that also adds a short position for the bottom 30% of the universe. Both portfolios are rebalanced annually, and interestingly rather than comparing the long-only portfolios’ performance against the market average, Novy-Marx chose to measure against a much tougher benchmark – a traditional value portfolio based on book to market (probably the most well researched market anomaly, that has been shown to significantly outperform the market averages over time).
For the long only portfolios, none of the quality measures managed to keep pace with the classic book to market value portfolio, with Novy-Marx’s own gross profitability portfolio performing the best, with a 7.1% annualised return, but still trailing the value portfolio by a substantial 1.6%. In the long-short test, similar results were observed, with 4 of the seven measures delivering positive returns, and gross-profitability performing the best, but none challenging the traditional value portfolio.
Navy-Marx then looked at combining quality and value, firstly by looking at the combined returns from side by side specialist portfolios of each. Whilst combining the two approaches in this way significantly reduced volatility relative to the pure value portfolio, it also reduced returns – leading to an improvement in the Information Ratio (tracking error relative to the market), but doing so by reducing risk not increasing returns, leading to an overall deterioration in the Sharpe Ratio (risk-adjusted returns).
It was here though that Novy-Marx achieved his breakthrough, choosing next to combine quality and value in a combined measure, by ranking stocks on the basis of both quality and value, then selecting the stocks with the highest combined ranking – rather than through two specialist portfolios side by side. In his earlier tests, the pure quality portfolios were dominated by very expensive holdings, making them effectively ‘short’ value. And similarly, a traditional value portfolio tends to be filled with cheap, low quality stocks, making it effectively ’short’ quality. As a result, an investor that put half of their money in each strategy ended up only tilting half way towards each of the factors. By selecting stocks on their combined rank on the other hand, an investor was able to achieve an average tilt to each factor of 71% – effectively increasing their exposure to both the quality and value factors by around 40%!
And what does that mean in the real world? It means that the combined quality and value portfolio not only trounces the broader market over the testing period, but also outperforms the classic value portfolio by an impressive 1.71% pa. And importantly, from a practical perspective, it achieves this by drastically improving the investability of a traditional value strategy. Whilst the long term returns of traditional value strategies are widely acknowledged, their tendency to significantly underperform the broader market from time to time – sometimes for many years at a time, makes them famously difficult to invest in. With a maximum 43% drawdown relative to the market over the testing period, an investor in a traditional value portfolio would have required significant intestinal fortitude to stick with the strategy (mind you – he would have been well rewarded if he did!). The combined quality / value portfolio on the other hand suffered a maximum drawdown of only 18.9% relative to the market, whilst delivering even better long term returns. Whilst traditional value strategies outperformed the market in 55.7% of one year periods, and 67.5% of 5 year periods, the combined quality / value portfolio outperformed in 72% of one year periods and an impressive 81% of 5 year periods.
Novy-Marx’s breakthroughs effectively came from two interesting and unique observations. Firstly, rather than traditional measures of quality commonly used by other academics that examined metrics based on the net profit of a firm, Novy-Marx focused on gross profitability (scaled by assets). Whilst initially counterintuitive, he observed that gross profitability is in fact a better predictor of future stock returns, possibly because there are so many forms of economic investment (e.g. R&D, advertising, sales commissions and staff training) that are treated as expenses, and hence reduce net profits, while increasing future profitability. Secondly, by constructing a combined quality / value screen rather than examining each metric individually, he was able to construct a portfolio of cheap, profitable firms, rather than firms that are just cheap or profitable. Doing so significantly improves returns, and reduces risk compared with examining each metric separately.
In his own words:
“Quality investing exploits another dimension of value. Value strategies endeavor to acquire productive capacity cheaply. Traditional value strategies do this by buying assets at bargain prices; quality strategies do this by buying uncommonly productive assets. Strategies based on either of value’s dimensions generate significant abnormal returns, but the real benefits of value investing accrue to investors that pay attention to both price and quality. Attention to quality, especially measured by gross profitability, helps traditional value investors distinguish bargain stocks (i.e., those that are undervalued) from value traps (i.e., those that are cheap for good reasons)… Cheap, profitable firms tend to outperform firms that are just cheap or just profitable… Several practical considerations make joint quality and value strategies look even more attractive. The signal in gross profitability is extremely persistent—even more persistent than that in valuations—and works well in the large cap universe. Joint quality and value strategies thus have low turnover, and can be implemented using liquid stocks with the capacity to absorb large trades.”
I couldn’t have said it better myself.