Given that I think most new fund flavors are garbage – in particular, I love to hate offerings like inverse-levered ETFs – and that I haven’t liked a new fund in years, why on earth would I propose the creation of a new fund? Because I think a Behavioral Index Fund has merit. Indeed, I can say for sure that I’d invest in it if only a provider would offer it.
Why would it make sense? Let’s delve into the issue of human behavior, the solution to exploit it and how that solution could be used by investors.
That investors are really poor at timing the market is not exactly breaking news. When markets surge, greed compels many of us to buy stocks. When they plunge, panic often takes over and propels us to sell. As unpredictable as markets are, human behavior is quite predictable. As the influential book Predictably Irrational, by Duke University psychology and behavioral economics professor Dan Ariely, states, that’s because humans are predictably irrational. Various studies show the results of poor market timing lead many of us to underperform the markets by anywhere from one to several percentage points annually.
Can we count on financial advisers to put a stop to this irrational behavior? Apparently not – as a TD Ameritrade Institutional chart shows, many advisers time the market just as poorly as nonprofessionals.
While many investors are behaving badly individually, it is mathematically impossible for investors in the aggregate to behave badly. I realize that most mutual funds and ETFs are open-ended, meaning nobody has to be on the other side of a transaction when an investor sells. However, a fund has to raise cash for large redemptions, so it must sell securities and, in the secondary market, some entity must buy. If, for example, investors sold large quantities of the Vanguard Total Stock Index Fund (VTI), Vanguard would have to sell some Apple, Google, ExxonMobil, etc., and someone would be on the other side of the transaction to buy it. Thus, in the aggregate, their timing would be neither good nor bad.
These funds must buy when the herd is selling and vice-versa. Yet I don’t think these funds can’t account for the total differences and, to be frank, I’ve never been able to determine who is profiting. It doesn’t appear to be foundations, pensions, or even hedge funds.
Measuring Human Behavior
Rather than create indexes based on market cap or even smart beta factors, an index that would measure human behavior – and do the opposite – could exploit the irrational behavior. It turns out that data to measure this is readily available: the Investment Company Institute releases fund flow information weekly, so it can be quickly discerned whether investors are moving into or out of U.S. stocks, international stocks or bonds.
Similar to a balanced fund, a Behavioral Index Fund could be a fund of funds. (As much fun as it would be to call it BIF, a different ticker would be needed – the Boulder Growth & Income Fund has already claimed BIF.) All of the funds used to create the Behavioral Index Fund would need to be broad and ultralow cost; a preferred asset allocation would be 30% of a total U.S. stock fund like VTI or the Admiral share class of the Vanguard Total Stock Market Index Fund (VTSAX), 30% of a fund like the Vanguard Total International Stock ETF (VXUS), and 40% of a fund like the Vanguard Total Bond Market ETF (BND).
A possible formula might be something like VTI and VXUS adjusted one percentage point for every $1.35 billion in quarterly fund flows versus the prior eight-year average. If, for example, U.S. stocks had a negative $5.4 billion delta, the weighting of VTSAX would be 34% (30% plus four percentage points). If investors poured an additional $4.05 billion over the historical average into international stocks, the weighting of VXUS would be 27% (30% minus three points).
The result is that BIF (my BIF, not the Boulder fund) automatically implements one of Warren Buffett’s famous sayings: Be greedy when others are fearful and fearful when others are greedy. In practice, I’ve found it very hard to get investors to buy stocks after a plunge and only slightly less difficult to get them to sell after equities surge.
BIF solves this by harnessing what I think is the most powerful force in the universe: inertia. The BIF investor is required to do nothing after buying the fund.
If it existed, would a Behavioral Index Fund be profitable?
About five years ago, I did a lot of research on the idea. I back-tested the formula to cover the period from Dec. 31, 1999, through June 30, 2011, and presented my findings to Vanguard. I used the investor mutual fund share class, since they were the classes that existed throughout the period. I am aware of the limitations of back-testing and doubt many funds have been launched that didn’t work on a backward-looking view.
From 2000 to mid-2011, bonds comprised as little as 19% and as much as 58% of the BIF portfolio. Note how the portfolio was lighter in stocks in 2008 (after five consecutive up markets) and very heavy into stocks in 2009 (after the plunge).
I compared BIF to both a static buy-and-hold strategy and a rebalanced strategy. Though all started with the same allocation, the static strategy did no rebalancing, while the rebalanced reset to the target at the end of every quarter.
The static portfolio earned 3.94% annually versus 4.37% annually for the rebalanced portfolio. BIF did significantly better, earning 5.46% annually. However, over long periods of time, BIF lagged both portfolios. For example, from 2002 to 2007, when U.S. stocks doubled and international stocks tripled, BIF was getting lighter in stocks since money was pouring in. But since stocks did well, BIF gave up most of the ground it gained versus the other two strategies. Finally, this was just one version of a formula to weight the three funds; better ones probably exist.
I was pleased Vanguard was willing to look at my proposal, and even did additional research that showed the performance differential was not statistically significant. Despite the rejection, I still haven’t given up on the concept.
Here’s why: BIF would not be a replacement for index funds or any low-cost investment strategy. It might be right for a small proportion of one’s portfolio – say 10%, for two reasons. First, advisors and clients should control the amount of risk taken. A portfolio that could theoretically go from 100% stocks to zero is, obviously, inappropriate. If, for example, an investor had a target 60% stock portfolio, having 10% in BIF would move the entire portfolio to a theoretical maximum of 64% stocks and minimum of 54%. Second, BIF would be very tax-inefficient. Since the fund would be selling securities regularly, it would create taxable gains and losses and would best be located in a Roth account or tax-deferred wrapper. The Roth wrapper would be ideal since the fund has high growth potential. Locating such a fund in only part of the portfolio also limits how much can be invested.
A Message to Fund Providers
I’d love to see BIF launched, but only if these four conditions are met:
· It should be ultralow cost. A fund with a 1% expense ratio defeats the potential gain.
· It should be broad – narrow index funds only increase risk.
· It should be easy to explain – a simple formula should be used to set asset allocation. That doesn’t mean the formula I back-tested is best, but complex formulas defeat the simplicity of exploiting human behavior.
· It must be marketed with brutal honesty. BIF is not a solution to an entire portfolio. It will not work every year and can underperform the static and rebalance portfolio many years in a row. BIF is a long-term strategy for a small part of one’s portfolio.
Of course, I can’t even guarantee that BIF will work in the long-run. But a change in human nature isn’t likely, so I suspect the cycle of human fear and greed will last indefinitely.
So fund providers (advisers, too) – what do you think? Post your comments on this page. Don’t worry – I don’t have my hopes up.