In truth, this is a long-standing problem. MSCI researchers and Jay Ritter at the University of Florida have both shown that economic growth only rarely translates into equity returns.
In theory, there is a simple solution to this problem. Investors should be able to buy GDP-linked bonds. These are government bonds whose interest and principal are tied to GDP growth: they pay out more when the economy does well and less when it contracts, much like index-linked bonds' payouts are tied to the inflation rate.
A recent paper from the Bank of England describes the benefits of such bonds. They would allow governments to cut interest payments in a recession, which would give them more "fiscal space" in which to support the economy by cutting taxes or raising spending without increasing public debt.
And investors would get an asset tied to GDP growth. This would allow them to bet on economic growth whilst avoiding some of the risks to equities, such as incomes shifting from profits to wages or from quoted to unquoted companies.
Such an asset should offer higher returns than gilts, to compensate investors for taking on cyclical risk, but lower returns than equities because GDP-linked bonds protect them from distribution risk. Bank economists estimate that they should return around 0.35 percentage points per year more than gilts.
There would be two other benefits of such bonds.
One is that they would allow people to buy insurance against recession. If you feared a downturn, you could go short of GDP bonds and so profit when their prices fall in the recession. Or at least, financial services companies could short-sell such bonds on their clients' behalf and so offer recession insurance.
Another benefit would be better information about the economy's prospects. The gap between yields on GDP bonds and conventional gilts would tell us what investors believe about the outlook for growth, just as the gap between index-linked and conventional gilts tells us about their view on future inflation. This means we could use the wisdom of crowds rather than professional economic forecasters to help foresee the future.
All this raises an obvious question. Given the benefits of GDP bonds, why don't they exist?
It's certainly not because they are a new idea. Nobel laureate Robert Shiller proposed them more than 20 years ago. And he was simply building on the idea of complete contingent markets developed by Kenneth Arrow in the 1950s.
One reason is that there are practical problems with them. One of these is that GDP data are subject to quite big revisions. This poses the dilemma: should payouts on GDP bonds be based on early data or later ones? The former are unreliable, but the latter would mean investors face years of uncertainty about their prospective payoffs, which would greatly reduce the benefits of such bonds.
This problem, though, is surely ameliorable by better data-gathering.
Instead, there might be another reason why GDP bonds don't exist. It's that the first government to issue them would suffer a disadvantage. It would have to pay higher borrowing costs partly because investors would find such products unfamiliar, and partly because they might fear that governments were issuing them precisely because they expected a recession.
The solution to this, say the Bank's economists, is simply for a group of national governments to agree to issue them at the same time.
Whatever the reason for the absence of GDP bonds, the fact is that a potentially useful asset doesn't exist. Cynics might say this is yet more evidence that the financial services industry does a better job of enriching its senior managers than it does of serving genuine needs.
