Jose Vicente Martinez at the Swedish Institute for Financial Research studied the performance of naïve diversification - simple investment strategies that consisted merely of spreading money evenly across the funds offered by the defined contribution pension plans of 10 big US employers. He compared these with more sophisticated strategies recommended by professional advisors. He found that the costs of simple naïve diversification were "insignificantly small". And, he added, the failure to rebalance regularly "may not be costly at all".
This conflicts with a famous paper by Sholomo Benartzi and Richard Thaler. They found that naïve diversification was a terrible idea. However, this was because of a framing effect: when investors were offered lots of equity funds and few bond funds they over-invested in equities, and when they were offered lots of bond funds they over-invested in bonds. Without this framing distortion, naïve diversification isn't so bad.
There's a reason for this. Optimisation requires knowledge of future returns and volatility which simply isn't available, and which even the best experts can infer only vaguely. Without such knowledge, optimising can go badly wrong. Imagine - which shouldn't be difficult - an investor in 1990. He looks back at history and infers that equities are a great investment, because they outperformed gilts by five percentage points per year between 1900 and 1990. So he invests heavily in them. Since then, however, the All-Share index has outperformed gilts by only 0.5 percentage points a year, and given some nasty losses in between. A more naïve investor, who knew nothing of past returns, might well have had a more balanced portfolio. For some levels of risk-aversion, he'd have done better in risk-adjusted terms than the smarter investor.
Imprecise knowledge isn't a minor problem for optimisation. It can be fatal. As the late George Shackle - one of the first economists to study the economics of knowledge - wrote: "For the traveller in the dark, a bridge with a missing span is worse than merely useless".
By contrast, that old Orwellian slogan "ignorance is strength" can actually be true for people using rules of thumb. Gerd Gigerenzer at the Max Planck Institute points out that when people were asked in the 1990s 'which is the larger city, San Diego or San Antonio?' Germans were more likely to give the right answer than Americans. This is because they figured 'I've heard more about San Diego than San Antonio, so it must be bigger' and so got the answer right. Americans, knowing more about both cities, just got confused. Ignorance, then, can be a useful cue. (In fact, since then San Antonio has overtaken San Diego.) Similarly, the naïve investor can use his ignorance of future returns to his advantage, by spreading his money more evenly.
However, this might not be the only way in which sophistication can cost money. If you think investing is a tricky subject requiring lots of research and thought, you can easily postpone doing so until you feel you know more. If this causes you to delay starting a pension scheme, you could end up with a smaller pot simply because you've made fewer contributions. In the long run, how much you save and for how long can easily matter more for your wealth than a few percentage points here or there on asset allocation. Even if naïve diversification is second-best, it can be better than doing nothing.
Perhaps, then, we shouldn't worry so much about doing the precisely right thing. Rough and ready diversification and simple rules of thumb can work pretty well.
