Large U.S. technology companies, the so-called FAANG,1 dominated the U.S. stock market in the last few years and had a significant impact on many investment strategies. These companies have been underrepresented in most factor-based strategies due to their unattractive factor characteristics. Have factor investors suffered from not investing in these stocks?
To answer this question, we examined the MSCI Diversified Multiple-Factor Index (DMF), a broad multi-factor index. The DMF Index represents an investment strategy that selects stocks based on four characteristics, also known as factor exposures: attractive valuations (value factor); smaller company size (size factor); high profitability, low leverage, stable earnings (quality factor); and positive recent trend in price (momentum factor). During its live 3-year history, the index has performed well in both global and international (ex-U.S.) markets. Despite underweighting leading U.S. tech stocks, the U.S. version of the index has slightly outperformed during this period.
We backtested the DMF strategy extensively prior to launching the DMF indexes in January 2015.2 Unsurprisingly, the backtests, covering the period December 1998 to December 2014, showed positive risk-adjusted returns relative to the market (see exhibit below). After all, we knew from previous research that systematically targeting stocks with attractive valuations, high profitability, small company size and positive price trend outperformed historically over long time periods.
Simulated performance of MSCI Diversified Multiple-Factor Indexes
Backtested results covering the period from December 1998 to December 2014.Active return is the return of the strategy relative to the market. Active risk is the risk of the strategy relative to the market. The multi-factor strategy is represented by the respective country or regional MSCI DMF Index. The market is represented by the respective country or regional MSCI Standard Index. Information ratio is the ratio of active return over risk, a commonly used measure of risk-adjusted return.
But how has the strategy performed since its launch in January 2015? Has it continued to deliver positive risk-adjusted performance? The next exhibit illustrates that the international, global and emerging market DMF indexes have performed in line with their backtests. However, in the U.S., the strategy has outperformed the market by only a small margin since it was launched three years ago. What was behind this lackluster U.S. performance? Did targeting stocks with attractive valuations, high profitability, small company size and positive price trend (the so-called target factor exposures) stop working in the U.S.? Or did something else hinder performance?
Live performance of MSCI Diversified Multiple-Factor Indexes
Index performance covering the period from December 2014 to December 2017. Active return is the return of the strategy relative to the market. Active risk is the risk of the strategy relative to the market. The multi-factor strategy is represented by the respective country or regional MSCI DMF Index. The market is represented by the respective country or regional MSCI Standard Index. Information ratio is the ratio of active return over risk, a commonly used measure of risk-adjusted return.
To answer this question, we turn to another tool of the quant trade, called performance attribution. Essentially, this tool answers the key question that active fund managers and their clients are asking: Did my bets pay off? In fact, this tool is able to distinguish between the contributions from different types of investment bets, such as bets on countries, currencies, sectors, themes (or factors) and pure security selection. The exhibit below is quite revealing. It shows that factor bets worked well historically and continued to work in the most recent period. What changed in the last three years is that stock selection, reflecting the influence of FAANG stocks, hurt performance, especially in the U.S.
Stock selection had negative impact on performance in the last three years, especially in the US
Performance attribution using the Barra Global Equity Model for Long-Term Investors (GEMLT). Analysis covers the period from December 1998 to December 2014 (left panel) and from December 2014 to December 2017 (right panel). All returns are gross annual total returns in USD. The multi-factor strategy is represented by the respective country or regional MSCI DMF Index. The market is represented by the respective country or regional MSCI Standard Index.
Unlike classic stock-picking strategies, where picking winners is paramount, in typical multi-factor strategies, security selection is often viewed as incidental. Given two stocks with similar valuations, similar capitalization, similar profitability and similar trend in price, a multi-factor strategy — unlike an active fund manager — cannot make a judgment as to which one to pick! Sometimes you get lucky, and stocks provide excess return. At other times, like the last three years, this incidental contribution to performance is negative.3
In the last three years, stocks that were avoided by multi-factor strategies due to their unattractive common characteristics (factor exposures) performed extremely well. The exhibit below shows that the FAANG stocks were not selected or underweighted due to their negative factor exposures. However, these stocks performed well over this period, and as a result impaired the performance of the strategy.
Characteristics of large technology companies and impact on performance
Analysis covered the period from December 2014 to December 2017. The left panel shows average stock level exposures. Active weight is the average weight of each stock in the DMF index relative to the market. Active return is the annual return of each stock relative to the market. The multi-factor strategy is represented by the respective country or regional MSCI DMF Index. The market is represented by the respective country or regional MSCI Standard Index.
Most asset managers advise clients to evaluate performance over long time periods of three years or more. However, the stock market sell-off and associated spike in market volatility in the early part of February 2018 was a significant stress test for all investment strategies. The final exhibit shows that the DMF strategies fell along with the market, but outperformed in relative terms by a small margin.
DMF outperformed during the stock market sell-off in early February 2018 by a small margin
Fundamental active managers seek to invest in high-quality companies with attractive valuations. Multi-factor strategies select stocks based on predetermined criteria that are similar to those used by fundamental active managers. The MSCI Diversified Multiple-Factor Indexes reflect a strategy that selects securities with attractive valuations, low company size, high profitability and positive recent trend in price. Despite avoiding large tech stocks that produced strong returns, these strategies outperformed the market over the last three years.
The author thanks Padmakar Kulkarni for his contributions to this blog post.
1 FAANG, a popular acronym coined by the press, stands for Facebook, Apple, Amazon, Netflix and Google.
2 This blog contains hypothetical, backtested or simulated performance results. There are frequently material differences between backtested or simulated performance results and actual results subsequently achieved by any investment strategy. The analysis and observations are limited solely to the period of the relevant historical data, backtest or simulation. Past performance — whether actual, back tested or simulated — is no indication or guarantee of future performance. None of the information or analysis herein is intended to constitute investment advice or a recommendation to make (or refrain from making) any kind of investment decision or asset allocation and should not be relied on as such.
3 In the DMF strategy, stocks that are equally attractive in terms of factor exposures may be discriminated through their risk characteristics. This particular strategy also includes a number of constraints to mitigate stock-specific risk, for example a constraint on active stock weights to be within 2% of the benchmark weight. This explains why Apple was held in the DMF strategy during this time with average underweight close to 2%.