“Are Investors Rational?” Elton, Gruber & Busse (2004)

by vonNudge

This paper looks at investor allocation decisions between index funds – mutual funds which are designed to track the returns of an index such as the S&P 500 or the FTSE 100. This is interesting, as the funds should, other than tax management, costs, and manager skill, be perfect substitutes. There is little reason to stay invested in a fund which will perform strictly worse than another (other than minimising unrealised capital gains taxes).This paper asks two questions: can the returns of these funds be predicted, and do investors make good decisions based on this information?

Perhaps unsurprisingly costs are the biggest driver of returns. These funds are designed to match the returns of the index minus costs. If one fund costs 0.06% a year, while another 1.35%, it stands to reason that the first fund should outperform the second. The authors find almost an exact one-for-one relationship between costs and returns: increasing fees by 1% reduces returns by 0.999%. Investors can outperform simply by choosing a low-cost fund. (Or one with high past returns in this sample, which correlates highly with low fees. Although buying funds with high past returns is generally a poor strategy when choosing between index funds and actively managed funds.)

The authors then look at investor cash flows. They find that investors do worse than if they based their decisions purely on past returns or investing based on fund size. One interesting variable is the load, or an initial fee – often up to 5% – that investors have to pay to enter a fund. Many funds don’t charge a load, so at least in this subset of funds (given the importance of costs) it would appear rational to only buy no-load funds. The authors actually find a positive coefficient on their load variable: increasing the load tends to increase investor inflows – something which could never be rational in this scenario.

The authors are fairly damning in their results’ implications for investor rationality. But why are investors making such bad decisions? One potential explanation which the paper hints at is that it could be due to marketing efforts. Funds with high fees and high loads tend to be the most heavily advertised and pushed by financial advisers. It could be that many investors are effectively “choosing” only from the subset of high cost funds.

My own hypothesis is slightly different. People are usually good at not overpaying for products. In fact, we tend to come with a whole set of norms over “fair” prices: see the paper by Thaler on fair prices for a beer from either a fancy hotel or a run-down shop. People will willingly pay much more for a beer in the former case than the latter, even if they are not “consuming“ any secondary products such as enjoying the hotel’s ambiance, and in fact the product is exactly the same. We will pay much more from a hotel since that seems fair given their higher fixed costs.

I think that in the index fund case something is going wrong with the communication of fund expenses. First, norms over a “fair” fund expense are likely not fully developed. Second, all the funds have expenses that are very low numbers – 0.06% and 1.35% both sound like low expenses, although the latter is over 22 times as expensive as the latter. If we were offered side by side prices of £1 or £22 for a beer I think I know the one we would pick! And third, we are used to paying for products with a direct exchange of cash. In investing the fund management charge is taken straight out of the fund’s gross return, so many people may not realise that they’re being ripped off. I think investors often display irrational behaviour because of a poor fit between the world of investing and the heuristics and techniques they have grown accustomed to in consumer behaviour. The crucial point is that many mistakes could well be reduced with a helpful nudge or two.