You might think this is a low probability. Maybe: it's one we can't quantify. But it's a nasty risk for at least four reasons. For one thing, it means bad long-term returns, so we cannot rely upon future recoveries to offset near-term losses. Second, the risk isn't confined to the UK. It applies to the eurozone, Japan and the US, so international diversification won't necessarily protect us. Third, it means lousy returns on cash because central banks could fight against the threat by imposing negative interest rates. Fourth, a world of secular stagnation is bad for jobs, too, so younger investors face the prospect of low wage growth and periods of unemployment as well as losses on shares.
Secular stagnation might (or might not) be a small danger. But investors must consider the small risk of horrible events.
And herein lies some good luck. We have a way of protecting ourselves from this risk at little cost. We can do so by using the 10-month average rule, which says we should hold equities if their prices are above their 10-month (or 200-day) average, and sell them if they are below them.
To check how such a rule would work in secular stagnation, I applied it to the Japanese market since December 1989: I assumed that one would have earned a zero return if not in equities - to capture the fact that secular stagnation is bad for cash returns too.
Such a rule would have actually made you money. ¥100 invested in the rule in December 1989 would have grown to ¥108 by now. This is a paltry return over 26 years. But the same sum left in the market would have shrunk to just ¥48. In effect, the 10-month rule would have avoided all of Japan's decades-long bear market.
This is because it got us out of the market for most of the tech crash of 2000-03 and the crisis of 2008, but it got us into equities in time to enjoy the 2003-04 recovery and the post-2012 Abenomics rally.
Returns on Japanese equities
Of course, this is only one test of how the rule performs in long bear markets. To get a broader picture of how it might work, think about three possible shapes of such a market:
1. A steady decline. In this event, the rule keeps us out of shares all the time, because prices are always below their 10-month average.
2. A saw-tooth pattern, in which the market rises gradually, interrupted by crashes. In this case, the rule would do terribly. It would keep us out of the early phases of upturns, but get us into the market just before the crashes.
3. Rallies interspersed with long declines. The rule does well here because it avoids the later stages of those declines, while getting us into the later phases of the rallies.
The question is: would secular stagnation be more likely to give us pattern three than pattern two? My hunch is 'yes', for two reasons.
First, news of secular stagnation would not arrive as a single big shock of the sort that would cause a crash. Instead, the evidence for it would consist of several disappointments about economic growth - the sort of things that cause shares to sag over time rather than crash dramatically.
Second, the possibility of secular stagnation creates uncertainty. And this generates momentum, which is good for the 10-month rule.
Put it this way. If we knew what future growth will be, we'd be value investors. We'd know that (say) a 5 per cent yield was a bargain and (say) a 3 per cent yield was not. We'd then buy on dips and sell on highs. These are circumstances in which the 10-month rule fails.
However, if we don't know what future growth will be we cannot be value investors. That 5 per cent yield might be a sign not that the share is cheap but that future growth will be low. And if there are fewer value investors who buy on dips there will be more momentum investors who try to guess future market sentiment and to ride rallies. Momentum-driven markets are ones in which the 10-month rule does well.
What I'm saying here is doubly good news. It's good because we have a way of reducing the nasty risk of secular stagnation. Better still, this method also works - on average - in normal markets, so it costs us nothing to adopt it.
What we have here, therefore, is a rare example of how we can manage risk without reducing returns. Sometimes, there isn't a trade-off between risk and return.
