Do gold or cryptocurrencies protect us from a fall in the value of conventional money? Recent events suggest not.
I say this because US inflation expectations have risen recently: the gap between five-year US Treasury yields and their inflation-proofed counterparts has risen to close to a 15-year high. This means that markets expect the value of money to decline at a faster rate than normal in coming years.
Such expectations should have been good for assets that are alternatives to conventional money. But they have not been. The gold price in dollar terms is much the same as it was at the start of the year, and Bitcoin has fallen by a third since early April.
Does this discredit those who had been holding these in the belief that conventional money would lose its value?
Actually, no.
What the last few weeks have shown is that gold and cryptocurrencies are uncorrelated with inflation expectations in normal times. Normal times, however, are no guide to what happens in crises. As Richard Bookstaber wrote in his excellent The End of Theory, “during crises, we see a breakdown of institutions and of assumptions that govern the normal application of economics”.
One of the assumptions that breaks down in crises is our beliefs about correlations. In crises, correlations change. Previously uncorrelated assets suddenly become correlated. This, for example, was what caused the most spectacular of hedge fund collapses, that of Long-Term Capital Management in 1998. It held assets which seemed uncorrelated such as Russian bonds and less liquid Treasuries – but these suddenly fell together. Similarly in 2007 banks found that hitherto uncorrelated mortgage derivatives lost money at the same time. The same thing happens in stock markets: when markets fall a lot, all strategies – value or growth, defensives or momentum – lose money.
It is this fact that provides one case for gold and cryptocurrencies. Sure, these aren’t much correlated with the value of conventional money in normal times. But this could change in extreme times. A full-blown crisis for conventional money could see the two do well.
In the case of gold we have hard evidence for this. Dirk Baur University of Western Australia and Thomas McDermott at the LSE show that it does well in really bad times for equities. “Gold can be seen as a panic buy in the immediate aftermath of an extreme negative market shock” they say. Also, during the inflation crisis of the 1970s – when the value of conventional money slumped - gold soared, tripling in price between 1972 and 1975.
Such safe haven status could prove invaluable. When people lose faith in conventional money they lose confidence not only in other assets such as equities and bonds but even in political institutions themselves: the inflationary 1970s saw a “crisis of democracy”. An asset that protects our wealth in such turbulent times would be very useful.
All this, however, raises some big questions.
One is: how likely is such a crisis of confidence in conventional money? Personally, I think it extremely unlikely. This could be because I have too much faith in liberal-democratic capitalism. But if this really is so fragile that such a crisis is possible, I’d expect to see serious people discussing ways to improve its resilience. The fact that this issue isn’t dominating the political discourse suggests that most people share my confidence.
Asset prices, however, should be equal to the probability-weighted sum of their values in all possible states of the world. Even a tiny probability of gold or Bitcoin being immensely valuable should therefore raise their prices today.
A second question is: if I’m wrong and conventional money does fall into disrepute would cryptocurrencies really benefit? In such a world, we’ll need are assets that do well when appetite for risk is plummeting. Whether cryptocurrencies do so is, however, doubtful. The fact that Bitcoin fell 30 per cent last March when the pandemic caused equities to slump suggests that its price is sensitive to appetite for risk and falls in a crisis. That’s the exact opposite of what we’ll need in desperate times. Sure, this could change. But the fact is that we’ve no significant evidence yet for cryptocurrencies being resilient to nasty macroeconomic shocks.
There’s a third question: what returns should we expect in normal times on assets that do protect us from a catastrophic crisis?
The answer is: negative ones. You have to pay for insurance – and the greater the chance of disaster or greater the cost of it, the more you pay.
An asset can protect us against bad times. Or it can do well in normal ones. But it is difficult to believe that it can do both.
