Many of us will regard 2016 in the same way the queen saw her own annus horribilis - as “not a year on which I shall look back with undiluted pleasure”. There is, though, a paradox about it: whilst it was an extraordinary year for politics, it was much less remarkable in finance.
After inflation, the All-Share index has given a total return of just under 15 per cent this year. This is better than the long-term average of 5 per cent, but not freakishly so. Using post-1990 volatility as a guide, we should expect to see such a return almost one year in every three. The half-point drop in ten year gilt yields is also the sort of thing we’d see almost once every three years. Among major assets, only the drop in sterling has been noteworthy. The 13 per cent fall in its trade-weighted index is the sort of move we’d expect around once every 30 years.
Although the year hasn’t been terribly exceptional financially speaking, it’s worth asking: what did I get wrong and right, and what can I learn from this?
One mistake I made was to believe betting markets, which told us that the odds were against both Brexit and a Trump presidency.
My mistake here was to confuse prices with the wisdom of crowds. The prices on a Remain vote and Clinton presidency were inflated by some big bets made by a few people who might have been motivated more by wishful thinking than ground truth. In both cases, the median bet seems to have been nearer the truth, which is consistent with the original thinking behind the idea of the wisdom of crowds as proposed by Francis Galton.
The lesson I take from this is not to trust prices alone, but to ask: what are gamblers doing?
Another thing I got wrong was to believe the “sell on May Day, buy on Halloween” rule. Both legs of this let me down. From Halloween 2015 to May Day 2016 the All-share index gave a zero total return, and from May Day to Halloween it delivered 12.2 per cent.
In February I wrote that gold’s attraction lay in the fact that it protected us against falls in expected interest rates, because gold tended to rise as real yields fell. Events have corroborated this
However, I don’t know what to infer from this. “Sell on May Day, buy on Halloween” was never an iron law but only a statistical tendency (albeit a strong one) with a few exceptions. So does 2016’s failure represent a breakdown of the tendency? Or is it just one of the exceptions to a tendency that will continue?
I’m inclined towards the latter. One reason for this is that the theory behind the tendency seems robust: we often get too gloomy in the autumn and too cheerful in the summer.
But there’s another reason to stick with the rule, at least for now. It’s that 2016 has vindicated one of my beliefs - in the ten-month average rule. This says that we should buy when prices are above their ten-month average and sell when they are below them.
Such a rule would have kept us out of equities earlier this year, thereby protecting us from a 12 per cent drop between the start of the year and mid-February, but would have got us into the market in July, since when prices have risen. With the All-share index still well above its ten-month average, this rule tells us to stick with equities.
My confidence in this is, however, tempered by the fact that the rule has done better as a sell signal than a buy one: its greatest virtue is that it protects us from prolonged bear markets.
A second thing I got right in 2016 was gold. Back in February I wrote that gold’s attraction lay in the fact that it protected us against falls in expected interest rates, because gold tended to rise as real yields fell. Events have corroborated this. Gold rose as real yields fell from February to late summer, but has fallen since then as yields have turned less negative.
This implies that there’s still a case for holding gold as insurance. It’s possible that real yields will fall in 2017 if growth expectations or appetite for risk decline. This is only a possibility rather than a probability. But it is a nasty one because the circumstances in which real yields fall might well be ones in which equities do badly. And low-probability, high-cost events are the sort of things we should insure against.
There’s still a case for holding gold as insurance. It’s possible that real yields will fall in 2017 if growth expectations or appetite for risk decline. This is only a possibility rather than a probability. But it is a nasty one
Insurance, however, is expensive. The ten-month average rule works for gold as well as shares. And with gold now below its ten-month average, the rule is warning us of the danger of gold prices falling.
A third thing I might have got right was sterling. In January I said foreign currency was insurance against crises as sterling tended to fall in these. Those of us who regard Brexit as a bad thing will see sterling’s fall this year as vindication of this: our profits on foreign currency softened the blow of the plebiscite result.
And remember why sterling fell. It’s because markets believe that Brexit will reduce future GDP (relative to what it would otherwise have been). In this sense, foreign currency does indeed hedge against bad news for the UK economy.
2017 was, therefore, a year in which I got some things right and some wrong, which is pretty normal. But of course, being right in the past is no guarantee of being right in future. One of the easiest mistakes we can make is to become overconfident because we got some things right. I’m sure that 2017 will bring some surprises and more errors. However, if we avoid taking strong positions and try to diversify we can at least limit the costs of those mistakes.
MORE FROM CHRIS DILLOW...
Chris blogs at http://stumblingandmumbling.typepad.com