Do pandemic bonds work

Long before the novel coronavirus appeared, Olga Jonas was fretting about the next global health crisis. As a World Bank macroeconomist in 2005, she was in the department that helped coordinate the response to avian flu and saw how ill-equipped the poorest countries were to fight a serious epidemic.

When an Ebola outbreak afflicted West Africa in 2014, she hoped it would fuel a global consensus behind building local capacities to intervene early. Instead, she says, she was disappointed that the World Bank opted to issue “pandemic bonds” designed to bring in investors to bear some of the cost of containing future outbreaks. The 2017 sale of $320 million of the bonds offered premiums of 11.5% or 6.9%, depending on the risk investors were willing to bear, with the cost borne by Japan, Germany, and funds from the International Development Association. In this edited excerpt of their February discussion on the Odd Lots podcast, Tracy Alloway spoke with Jonas, now at the Harvard Global Health Institute, about the idea behind the bonds and the damage pandemics can wreak on economies. (Run BPOD to find Odd Lots and other Bloomberg podcasts.)

So how exactly are these bonds supposed to work? How does the money get to the World Bank?

There are parametric triggers, but it proves to be very challenging to define the triggers because it’s very difficult to anticipate how an epidemic starts. That’s why in designing it they chose triggers that are much later. It has to be at least 12 weeks after the beginning of the outbreak before anything can be triggered, as well as a high number of deaths and the growing rate of the outbreak. So that means it’s triggered much too late. But if it’s triggered earlier, the price of the insurance would be much higher because there is just so much uncertainty in the modelling. Much of the uncertainty is due to the lack of data on these kinds of events. The lack of data is due to the lack of public-health systems in developing countries, which is what is needed to invest in. And that’s not very expensive. It is, in fact, highly affordable compared to the benefit. And that’s what’s been sidelined.

Who’s the arbiter of when these bonds pay out? How do they verify that the triggers have been met?

Verification of the triggers is spelled out in the prospectus for the bonds, which is 386 pages long, and there is a verification agent [AIR Worldwide]. It’s a commercial contract between the World Bank and the verification agent, and they are going to ascertain whether all the triggers have been met. This is not a trivial exercise to verify these triggers, because it’s really quite complex. So when the verification agent notifies the World Bank that the triggers have been met, then the World Bank would get the money from the bonds, because it’s holding that money.

What’s the maximum payout the World Bank could get?

That’s the other issue that’s very disappointing in this whole experience. For coronavirus, when you look at it, the first payout — if it happens, there’s no way of telling at least from where I sit whether it will happen — it will be $131 million, and the maximum payout is $196 million. And that will have to be divided among the 76 poorest countries. So you can see if the poorest countries in the world — with a bigger population than China all together — they will get only a fraction of what China is already spending.

One of the sort of tragic aspects of this is, in fact, that the payment for the cost of the bonds, the premiums and the interest and the fees that were associated with this pretty complicated transaction, that those add up to $115 million. And those funds actually came from funds that were intended for the poorest countries. They came from IDA [International Development Association], which is money that donors give to the World Bank to finance productive projects in the poorest countries. So that was $50 million from IDA. Then $50 million was donated by Japan, but I’m sure the Japanese government intended that their donation of $50 million benefit the developing countries, benefit the poorest countries, and protect them from pandemics. And then $15 million was donated by Germany. Taxpayers in Germany, taxpayers in Japan. So all together $115 million has been paid for premiums and for interest and for the fees to recipients who are not poor, who are in high-income countries — these are investors who, of course, they invest their funds and they are at risk of losing some of this money because of the triggers, but that is a very high return. The idea was that investors or the private market would share some of the risks of a pandemic and thereby contribute. When a pandemic worsens, the markets will decline and the prices of assets fall, so investors are already going to be losing a lot of money just because there is a pandemic.

So these were pitched as something that should be uncorrelated with the broader market, but if they trigger, it would probably be because something quite serious was happening and therefore markets around the world would be falling anyway.

Exactly.

Do you think there’s any way to structure a bond that would be attractive to investors but also get extra money to the World Bank when it’s needed?

The World Bank does not need extra money to respond to pandemics. The World Bank is not a budget-constrained entity. It’s a bank. And IDA, the fund for the poorest countries, is the largest multilateral public fund to support development in poor countries, which includes — for the last 50 years — responding to emergencies. IDA has very ample liquid assets, reserves. It makes new loans now worth $27 billion every year. The money is not the issue, it’s the preparedness of the World Bank to respond, deliver the financing on the ground, and the preparedness of the country to implement the activities that are necessary to control the outbreak. But money is not, has never been, the issue. So this was more a sort of attention-getting initiative to try to innovate in this space, but it was not needed and it did not work.

What lessons did you learn as an economist dealing with those sorts of epidemics?

What is astonishing is how underappreciated the economics of epidemics were or still are. There’s very little realisation that if you act early, if you are prepared to stop the outbreak when it’s just a few cases, before it spreads, then in fact you are avoiding a huge cost due to the exponential growth that can happen with these diseases. That is not understood by the sort of bureaucratic processes that we have in place. Usually it is a disaster that occurs, and then there is an estimate of the cost of rebuilding like in a hurricane or an earthquake, but in this case you are averting something that hasn’t happened yet. So that’s why there’s this repetition of panic, and then it’s forgotten, because in the health sector there are so many unmet needs that people’s attention shifts somewhere else. So that accounts for the high costs of responding next time.

Does the economic recovery after a pandemic look different from the economic recovery after an earthquake or a hurricane?

Macroeconomically, it’s a temporary shock that passes through disruptions of travel and trade and supply chains. And when that comes to an end, things return to normal, and there is no permanent effect. However, when this happens in very poor countries — like what happened in West Africa in 2014 to 2016, the Ebola outbreak had a very severe effect on the health-care system. A large number of doctors and nurses died. It takes many years and is expensive to train new doctors and nurses, so there you would see development set back by a decade. Losing a decade of growth in very poor countries is very serious.

Alloway is an executive editor at Bloomberg News in Hong Kong and co-host of the Odd Lots podcast with Joe Weisenthal.