Because mutual fund flows can impact securities prices, the relationship between investment flows and the performance of factors and mutual funds is of great interest. For example, do investors naively look only at raw returns when making asset allocation decisions, or do they adjust returns for risk, and exposure to various factors, using an asset pricing model?
Latest Research On Flows
Recent research has found that mutual fund investors are largely ignorant about systematic risks when allocating capital among mutual funds.
For example, Itzhak Ben-David, Jiacui Li, Andrea Rossi and Yang Song, authors of the November 2018 study “What Do Mutual Fund Investors Really Care About?”, sought to determine whether investors use prominent asset pricing models—such as the capital asset pricing model (CAPM) and the three- and four-factor versions of Fama-French models—to allocate capital, or whether Morningstar’s star ratings (which do not account for systematic exposure to explanatory factors) explain mutual fund flows better than risk-adjusted returns.
Following is a summary of their findings:
- Ratings are the main determinant of capital allocation across mutual funds, followed by past returns.
- Morningstar ratings predict the direction of flows up to 68% of the time versus 60% for the CAPM, and even lower (between 58% and 60%) for other common asset pricing models.
- The spread between flows to top and bottom funds is best explained by Morningstar ratings. For example, when using Morningstar, 67% of top-ranked funds receive positive flows, while only 16% of bottom-ranked funds receive positive flows, generating a difference of 51 percentage points—significantly higher than all other measures, which generate differences in the 16 to 23 percentage points range.
- At the aggregate level, in every single year, funds rated highest by Morningstar received more money than the funds ranked highest according to any asset pricing model.
- There is no evidence that investors discount fund returns related to market risk exposure or to the other risk factors.
- Recent past returns, but not volatility (once Morningstar ratings are accounted for), explain capital allocation beyond Morningstar.
- Fund flows are weaker for high volatility funds only because Morningstar ratings penalize funds for high volatility.
The authors concluded: “In summary, we find no evidence that investors use the CAPM, or any other of the commonly-used factor models, to allocate capital to mutual funds. Rather, they naively rely on external rankings as a way to chase past winners.” They added that their results “support the proposition that mutual fund investors do not care about risk or do not understand risk.”
This naive behavior is why individual investors are referred to as “noise traders.” Despite trading on noise—rather than fundamentals—the fund flows that result from the behavior of retail fund investors can impact securities prices and returns to factors that explain performance.
Flows & Factors
Shiyang Huang, Yang Song and Hong Xiang, authors of the January 2019 study “Fragile Factor Premia,” sought to determine what, if any, impact mutual fund flows had on returns of the well-known Fama-French (FF) size and value factors.
To determine the impact, they estimated mutual fund flow-induced trading (FIT) for each stock quarter from 1980 to 2017. They used FIT, rather than all trading of mutual funds, because FIT only captures those trades driven by the demand shifts from mutual fund investors, which are largely ignorant about fundamentals.
Following is a summary of their findings:
- Aggregate flow-driven demand shifts across the size spectrum and across the book-to-market ratio spectrum significantly affect returns of the size (SMB) and value (HML) factors, respectively.
- The returns of six FF size and book-to-market portfolios are largely determined by the uninformed mutual fund flow-induced trades.
- Within each of the six FF portfolios, stocks with higher FIT have higher return performance. For example, stocks with a top-quartile FIT, on average, outperform the bottom-quartile-FIT stocks in the same portfolio by about 1% per month, although they have similar firm size and book-to-market ratios.
- Across the FF size and book-to-market portfolios, growth stocks with a positive FIT significantly outperform value stocks with a negative FIT, controlling for firm size. Controlling for book-to-market ratio, large cap stocks with a positive FIT significantly outperform the negative-FIT small cap stocks. As a test of robustness, they confirmed this reversal in premiums across 11 other CAPM anomalies.
The findings led the authors to conclude that “The well-known size (SMB) premium is due to the component of small-cap-inflow stocks minus large-cap-outflow stocks, while the value (HML) premium is due to the component of value-inflow stocks minus growth-outflow stocks. The other components of the size and value factors actually have significantly negative average premia.”
They also found that this flow-induced performance is more pronounced in the more recent sample period, consistent with the rise in the size of the mutual fund industry over time.
Flows Influence Time Variation Of Returns To Factors
Huang, Song and Xiang proceeded to determine how aggregate flow movements influence time variation of the size and value returns. Based on what we have seen, we should expect that SMB (HML) returns should be high in the periods when there are more flow-driven trades into small-cap (value) stocks relative to large-cap (growth) stocks and vice versa.
That is exactly what they found: “Aggregate flow-driven demand shifts across the small-cap and large-cap portfolios and demand shifts across the value and growth portfolios are statistically and economically significant drivers of the size and value returns, respectively.”
For example, they found that “A one-standard-deviation change of the aggregate FIT across the size spectrum is positively associated with a 1.65% change in quarterly SMB returns (6.61% on an annual basis).”
Importantly, the authors also found that “The flow-induced effects on factor returns significantly revert over longer horizons.” In other words, while inducing momentum in factor returns, the flows are just noise—while they positively affect contemporaneous factor returns, they don’t cancel out long-term premiums as factors experience strong reversals over longer horizons.
For example, they found that “A one-standard-deviation increase in the difference of flow-induced trades into value stocks and flow-induced trades into growth stocks over the prior five years, on average, predicts a 4.19% decrease in the HML returns over the next year.”
FIT Across Anomalies
In an April 2019 study, “Flow-Induced Trades and Asset Pricing Factors,” Huang, Song and Xiang expanded their work to include 50 well-known factors (anomalies to the CAPM). Their findings were consistent: “Our results indicate that these factors are heavily exposed to flow-driven ‘noise trader’ risk, which we further show is significantly priced.”
They added that the flow-driven effects on factor return dynamics can partially explain factor momentum (as well as the underperformance of large-sized mutual funds relative to small funds). Summarizing, the authors noted: “Our results indicate that these asset pricing factors are heavily exposed to non-fundamental risk that is due to mutual funds’ flow-driven demand shifts.”
Mutual fund investors are largely ignorant about systematic risks when allocating capital among actively managed equity mutual funds, causing them to trade based on noise, not fundamentals. Their naive performance-chasing behavior induces short-term momentum in factors but does not impact long-term premiums.
Note that the finding of momentum in factors is consistent with the findings of Tarun Gupta and Bryan Kelly in their December 2018 paper “Factor Momentum Everywhere.” They built and analyzed a large collection of 65 characteristic-based factors that are widely studied in the academic literature, including a variety of valuation ratios (e.g., earnings/price, book/market); factor exposures (e.g., betting against beta); size, investment and profitability metrics (e.g., market equity, sales growth, return on equity); idiosyncratic risk measures (e.g., stock volatility and skewness); and liquidity measures (e.g., Amihud illiquidity, share volume and bid-ask spread).
Following is a summary of their findings:
- Individual factors exhibit robust time series momentum, being positive for 59 of the 65 factors, and significantly positive in 49 cases.
- Robust momentum behavior among the common factors is responsible for a large fraction of the covariation among stocks.
- A portfolio strategy that buys the recent top-performing factors and sells poor-performing factors achieves significant investment performance above and beyond traditional stock momentum.
- On a stand-alone basis, factor momentum outperforms stock momentum, industry momentum, value and other commonly studied investment factors in terms of Sharpe ratio.
- While factor momentum and stock momentum are correlated, they are also complementary—factor momentum earns an economically large and statistically significant alpha after controlling for stock momentum and expenses.
- Demonstrating pervasiveness, factor momentum is a global phenomenon—it manifests equally strongly outside the U.S.—in a large global (ex. U.S.) sample, and Europe and Pacific region subsamples.
Larry Swedroe is the director of research for The BAM Alliance, a community of more than 130 independent registered investment advisors throughout the country.