Consider a simple asset allocation - one that has 40 per cent in the All-Share index, 30 per cent in gilts, 20 per cent in cash and 10 per cent in gold. And let's say you rebalanced it at the end of every year. Such a simple strategy would have beaten the typical fund in Trustnet's database of targeted absolute-return funds and done about as well as the average balanced managed fund with 20-60 per cent equity weightings.
| Returns on a simple asset allocation & funds | |||
|---|---|---|---|
| 40-30-20-10 portfolio | Target absolute return | Mixed 20-60% equities | |
| Last 12 months | 13.5 | 1.1 | 9.5 |
| 12M to Sep 2015 | 1.3 | 2.6 | 0.8 |
| 12M to Sep 2014 | 4.2 | 3.9 | 5.3 |
| 12M to Sep 2013 | 3.9 | 5.6 | 9.0 |
| 12M to Sep 2012 | 9.8 | 1.8 | 7.8 |
| Source: Trustnet & IC | |||
This is despite the fact that this portfolio is very basic. We'd have had more stable returns with a lower equity weighting, or with foreign exchange because sterling tends to fall when equities do. And we'd have had better returns if we'd used global rather than UK equities.
Very basic asset allocation strategies can, therefore, do about as well as fancy funds. Most of the time that fund managers spend in the office is wasted: my strategy requires only a few minutes work on every 31 December. This shouldn't be as radical as it seems. If returns are unpredictable, a policy of spreading your bets is as good as you can do, and time spent trying to forecast returns will be wasted.
There's a simple reason why such a strategy works. It's because its two biggest holdings - equities and gilts - have been negatively correlated so losses on one have often been mitigated by gains on the other.
However, the performance of the past five years flatters this portfolio. It would have lost you almost 10 per cent in the 12 months to October 2008 as massive losses on equities weren't offset by gilt returns.
But a few tweaks to our simple strategy could have mitigated these losses. Using the 'sell on May Day, buy on Halloween' rule would have saved us from a 28 per cent loss that summer. Using foreign currency would have given you nice profits in 2008 as sterling fell. Or following Mebane Faber's rule of selling when prices fall below their 10-month average would have kept you out of equities between late 2007 and May 2009, thus avoiding a 27 per cent loss. Applying his rule to our portfolio generally would have given us more stable but lower returns since 2000.
It doesn't follow, however, that a portfolio like this, even with tweaks, will always do well. One danger is correlation risk - that previously lowly correlated assets will do badly at the same time. This could happen if we get a surprise tightening in (global) monetary policy. This could lead to losses on bonds as investors anticipate higher future interest rates; falling gold prices as the opportunity cost of holding gold (returns on cash) increases; and falling share prices as 'reach for yield' investors sell.
We can't say how great a risk this is - although given that central banks strive for transparency and predictability it should be small. Nor can we be confident that absolute return or balanced managed funds would escape the danger if it does materialise.
For most practical purposes, however, it seems that quite simple DIY strategies work well enough - or at least well enough to make one question whether it's worth incurring fund managers' fees.
