Join our community of smart investors

The case for GDP-linked bonds

Governments should issue GDP-linked bonds – but not because they will reduce debt servicing costs
July 2, 2020

If you live long enough, you’ll see everything repeated. So it is with calls for governments to issue GDP-linked bonds.

These were first proposed by Yale University’s Robert Shiller back in 1993, in his book Macro Markets, but few countries took up his suggestion. With governments now facing soaring debt as a result of the lockdowns, they are back in fashion.

And there’s a logic for them. These bonds link interest payments to GDP growth: when the economy does well the government therefore pays out more interest and when we get a recession it pays less. This means the cost of servicing debt falls when tax revenues fall, which makes high debt more sustainable. Better still, economists at the Bank of England estimate that such bonds will reduce the government’s interest bill in normal times, because investors will accept lower interest payments in exchange for the prospect of a rising payout as the economy grows.

This, however, misses the purpose of such bonds. Governments in developed economies that can print their own money don’t face any meaningful debt constraint: reports that the government “nearly went bust” in March are hysterical gibberish, as the excellent finance writer Frances Coppola has pointed out. GDP-linked bonds therefore mitigate a problem they don’t have. This, perhaps, is one reason why it has so far been only governments in poorer countries that have issued them.

But, but, but. There are in fact much stronger reasons why we need GDP-linked securities.

One is their information value. As the IMF’s Prakash Loungani has shown, economists are terrible at forecasting recessions and recoveries. We need, therefore, some other way of predicting activity. If there is indeed wisdom in crowds, GDP-linked bonds will do the job.

Also, GDP-linked bonds would give investors a better way of betting on a country’s growth prospects. Stock markets don’t offer this, because there is little correlation over longer periods between GDP growth and equity returns. Such bonds would let us bet on a country without having to back its listed companies.

Perhaps the greatest benefit of GDP-linked securities, however, would be if we could take a short position in them. Doing so means we would profit if there is a recession or just fears of one. Which in turn means we could take out insurance against economic downturns. Given that governments are unwilling to offer such insurance – the pandemic has revealed the inadequacy of welfare states and countercyclical macroeconomic policies – there’s room for the private sector to do so.

Granted, GDP-linked bonds as they are currently proposed are not easily shortable by those who need insurance against recession. But Professor Shiller has shown that they can easily be tweaked to make them so. He has proposed that a pair of securities be issued linked to GDP, one of which rises as GDP rises while the other falls leaving the value of the pair constant.

All this poses the question: if GDP-linked securities are such a good idea, why don’t they exist?

Simple. There’s a collective action problem. The benefits of useful financial innovation go to millions of us, but it is difficult for innovators to capture them for themselves: Yale University’s William Nordhaus has shown that they typically get “only a minuscule fraction” of them. This means innovative activity is skewed towards products of high profit but little or no social utility, such as high-charging funds or mortgage derivatives.

It’s in this sense that we should welcome the discussion of GDP-linked bonds – not because these provide a cheap way of financing government spending, but because the debate highlights a flaw at the heart of our financial system.