Foreign currency as a safe asset

One solution might be to hold some foreign currency. My table shows the point. It shows the average annual change in sterling since 1991 during three different types of bad times – falls: in US industrial production (an indicator of global recessions); in the All-Share index; and in UK house prices.

During these times sterling falls significantly on average, delivering nice returns on major foreign currencies.

It’s true that gilts have also done well during these times. But this is partly because they’ve given good returns almost all the time since the early 1990s. In fact, their returns during times of falling US output and UK share prices have actually been slightly lower than their long-term average returns. If you believe – as many do – that gilts are over priced, they will not deliver these past returns in the future.

But why should the history of foreign currency doing well during bad times repeat itself?

Simple. Sterling is regarded by global investors as a riskier currency than US dollars, euros, yen or Swiss francs. And in bad times, investors dump riskier assets in favour of safer ones. So sterling falls in such times, giving sterling-based investors nice returns on other currencies.

One of the privileges of being British is that we have a form of insurance not available to European or US investors. We can use sterling’s status as a risky currency to hold foreign currencies, thereby getting some insurance against bear markets and recessions.

There is, of course, a price to be paid for such insurance.

One is that in good times the pound will rise and we’ll lose on foreign currency simply because during such times investors become more willing to take risk so they buy sterling.

In such events, however, share prices would probably rise, offsetting our losses on dollars or euros. What matters is the performance of our portfolios as a whole. There is no single asset that will perform well in all possible states of the world.

Another cost is that foreign currency will do badly on average. Our central expectation for future exchange rates should be their current rates, give or take volatility: history suggests there’s a two-thirds chance of the euro or US dollar being within 7-8 per cent of their current rates against sterling in 12-months’ time. With interest rates on euros negative even in nominal terms, this means that we should expect to lose a little on average.

This, however, is an unavoidable fact about insurance. If you don’t make a claim on your home insurance, you lose your premium. But this loss is worth it for peace of mind. The same is true for assets that insure us against losses. We shouldn’t expect profits on these: peace of mind is sufficient reward. Perhaps the 1980s and 1990s gave us over-inflated expectations of what returns on safe assets should be.

Instead, there’s a bigger problem with foreign currency. It’s not inconceivable that it might not actually work as insurance.

One complication here is that sterling is cheap. Adjusted for inflation, it is low against the US dollar and euro – a fact which has in the past led to it rising, producing losses on foreign currency.

In itself, I don’t think this is a problem. Sterling has been cheap by this standard since the EU referendum result in 2016. If this fact alone were going to cause it to rise, it would have done so by now. Instead, of course, the pound has stayed low because investors fear that Brexit will depress long-term growth, relative to what would otherwise be the case.

Which raises one risk. If Brexit doesn’t happen (or if investors were to expect it not to) sterling might well rise causing losses on foreign currency. If this happens at the same time as world share prices were to fall, foreign currency would lose its insurance value.

There’s another way in which this might happen. If the Bank of England raises rates while other central banks don’t, sterling could rise and we might see a reversal of UK investors’ “reach for yield”. Again, this would see losses on both foreign currency and equities.

I don’t know how likely these scenarios are. Many of you, though, might think them no more likely than a scenario in which bonds and equities both fall – say because investors revise their current belief that the Federal Reserve and European Central Bank will not raise rates for a long time.

Assessing probabilities is not, though, the point. The point is that gilts and sterling cash might not be the only safe-ish assets. It’s very possible that major foreign currencies are also safe-ish in the sense that they might do well if we see losses on equities or housing. There are, therefore, more ways of spreading risk than you might think.