Individual investors often look toward mutual funds as opposed to US treasury notes as a way to have greater return, but still less risk that picking stocks on their own. The idea of using mutual funds to increase personal wealth in a safe manner is not a new one. However, some mutual investor behavior indicates that this commonplace theory is not true
In Dumb Money: Mutual Fund Flows and the Cross-section of Stock Returns, Andrea Frazzini and Owen A. Lamont look at investor behavior in picking mutual funds and whether individual personal sentiment, or opinions on the market, affect stock prices indirectly through mutual fund holdings. Wanting the highest return, or “return-chasing”, investors often move their money from funds with recent low returns to funds with recent high returns or funds that have current popular themes, such as technology. Frazzini and Lamont speculate that the reallocation of money among mutual funds by retail investors actually reduces personal wealth in the long run, an effect they call “Dumb Money.” Dumb money also relates to the corporate issuance of stock.
To calculate the investor sentiment, Frazzini et. al use a model based on flows. Flows are the amount of ownership that is due to the transfer of money to or from another fund or stock. Flows do not include new money, but the reallocation of money that describes investors’ changing sentiment. The flow for mutual funds is the difference between the actual value of the fund and the value of the benchmark fund. The actual value is taken from real world data. The benchmark fund is a hypothetical situation where the same percentage of inflow money would go to the same funds. The percentage of the total money each fund had would remain the same. The difference in those two scenarios signifies the sentiment an investor has for a fund. Negative difference shows an investor took money out of the fund while a positive difference shows an investor put more money than he would have in a proportional world. The benchmark fund is used to show a situation where investors use no personal feelings when picking a fund because they weight each fund the same.
Using mutual fund flows to find the flows of stocks, the authors create a model of actual mutual fund ownership of a certain stock vs. hypothetical ownership of a stock by a fund. Similar to the allocation in the previous benchmark fund, the stock allocation is based on a proportional system, where inflows into the fund are distributed in the same percentages as before the inflow. This hypothetical strategy is used because the portfolio managers (PMs) choose stocks independently of inflows and outflows. Again, the difference between the actual holding of stock by mutual funds and the benchmark holding of stock shows where real PMs over or underweight certain stocks due to the flows into the funds. Frazzini and Lamont recognize that PMs often have some skill at picking stocks for the funds, an effect called “smart money.” However, individuals’ investing behavior forces PMs to pick stocks even if they don’t consider them valuable because of the need to invest the new inflow of money.
Drawing on 13 years of data on mutual fund asset values and mutual fund holdings, Franzzini and Lamont demonstrate that high sentiment predicts low future returns at long horizons. The authors found the returns by looking at what returns an average investor would receive with dumb money strategy vs. always using the same percentages. Dumb money is present in large and small funds and over different periods of holding. While smart money shows weakly at a short horizon (3 months), investors fail to capture the high returns. Investors end up overloading in high momentum stocks and are deficient in low momentum stocks. Usually high momentum stocks are growth, and low momentum stocks are value, noting a correlation between dumb money and the value effect.
Taking the analysis one step further, the authors observe the ability of firms to take advantage of dumb money. Firms tend to increase shares either through seasoned equity offering, a merger or another method, when the previous year’s flow for the firm’s stock was high. Unlike individual investors, firms are able to play the market, selling shares when investors are overly investing in the companies stocks via mutual funds and buying back shares when the firm’s stocks are out of favor. The authors don’t delve into the any theories behind the dumb money, investors, and smart money, firms, but leave the readers with a large bank of knowledge on the correlation.
Once I finally understood this paper, I thought it was pretty interesting. However, I had two large problems with the paper, one on the writing style, one on the model. First, this paper was very difficult to read. It requires a great deal of knowledge to understand what the authors lay out. They fail to give definitions of the theories they constantly refer. The paper builds off a plethora of other working papers, which, while described briefly sometimes, have not been read by the reader and thus the reader does not know what the authors are referring. In addition, the wording is arduous to get through. The authors never seem to say exactly what they want, so they constantly say similar things in different ways, none of which make it clearer, only more confusing. The in-depth explanation of the findings is lacking.
Second, the idea of finding investor “sentiment” indirectly via flows is hurt greatly by the contrafactual world or what I called the benchmark. For dealing with inflows or outflows, the proportional method is always used. A proportional investment strategy isn’t the only strategy that doesn’t involve sentiment. For PMs, investing is based on a lot of fundamental analysis of companies. With the ever changing markets, PMs change their decisions on stocks. When new money comes in, a PM doesn’t buy more shares of each company he already owns, but reevaluates the companies, seeing which ones are worth it, which ones aren’t and which one’s aren’t even in the fund yet. In addition, most mutual funds have a cash balance, which the inflow could also go into.