GDP-linked bonds: from concept to reality?

Growth-linked sovereign bonds sound desirable in theory but have long been considered impossible in practice. Now a small group of investors, lawyers and trade bodies is working hard to turn the idea into reality. Danielle Myles reports.

When the G20 finance ministers met in July, they tasked the International Monetary Fund (IMF) with investigating the opportunities and challenges of state-contingent debt. Nestled halfway down the ministers’ list of 19 calls to action, this task is not their top priority. But the fact this suite of instruments has made it onto the international agenda at all could spark the next innovation in sovereign debt.

Among these state-contingent instruments – all of which promote greater risk-sharing between government issuers and investors – bonds linked to gross domestic product (GDP) are the most promising. Their underlying premise is that the issuer’s debt obligations grow and shrink in line with its economic growth. This discourages procyclical fiscal policies (by freeing up funds when growth is stagnant, and limiting spending during good times) and improves long-term debt sustainability.

Gathering momentum
Academics and economists have debated the concept since the 1980s but it has never led to any agreed formulation or parameters. There has been a handful of highly tailored and complicated instruments that include pay-offs contingent on GDP – most recently the warrants issued by Argentina (2005), Greece (2012) and Ukraine (2015) – but they have all been part of debt restructures. As a result, the concept has become synonymous with distress.

Today, however, the momentum behind creating an asset class of GDP-linked bonds is stronger than ever, and it is largely thanks to an industrious London-based working group that has spent the past 12 months trying to convince the market to forget everything it thinks it knows about GDP-linked debt.

On a pro-bono basis this small group of investors, lawyers and trade body representatives – with the involvement of the Bank of England (BoE) – has created a term sheet for an instrument that is radically different to anything proposed before. The biggest initial challenge has been breaking the link between it and the stigmatised warrants seen to date.

“People need to understand that this is a new tool for the sovereign debt toolbox, and they need to understand how it works before assessing it,” says Starla Griffin, managing director at Slaney Advisors, and a key member of the working group. “It won’t be everyone’s cup of tea as investors are different, but if you are only looking at it through the prism of what has come before, you need to know it’s definitely not that.”

How it is different
The term sheet is designed to address criticisms directed at the previous incarnations of GDP-linked debt, including that it can be issued outside of a debt restructure. At the time of writing, the term sheet was still being refined but, in essence, it describes a local currency-denominated bond where principal and coupon are both indexed to nominal GDP. It has a fully symmetrical payout profile with no caps, floors or thresholds, and a payment formula modelled on inflation-linked bonds. It has a bullet repayment and a robust process for ensuring the reliability of GDP data used to calculate payments.

The over-riding goal is to create an instrument the market is familiar with. “We’re trying to design something that looks like, talks like and walks like a normal bond – something people are used to: bullet repayment, a very simple stream of cashflows and a plain vanilla structure, so the bond can be used as a store of value,” says Christian Kopf, partner at Spinnaker Capital and another working group leader. “That’s why we think GDP-linked bonds can really take off.”

The repayment calculation is critical. It compares GDP at the date of payment to GDP at the date of the bond’s inception, meaning there will always be a coupon. But if in any given year GDP is lower than at the issuance date, it will be less than the base coupon. It is the same for principal repayment.

The very fact there is a term sheet takes GDP-linked bonds to a level of practicality never seen before, and draws on previous sovereign debt experiments. “Based on what we’ve seen in the inflation-linked market, which has become reasonably successful, it would make sense to standardise it as much as possible – don’t go for an unnecessarily complex structure,” says Gianluca Salford, rates strategist at JPMorgan.

It is an iterative process. Since being launched at a November 2015 workshop at the BoE and attended by nearly 100 investors, bankers, policy-makers and academics, the term sheet has been discussed at the IMF spring meeting, central bank consultations and investor roundtables. It has been refined over time, based on feedback. The International Capital Market Association (ICMA) and other trade bodies have backed the initiative and the goal is to have them endorse the term sheet by the IMF annual meetings in October.

The premium debate
A major challenge still being worked through is the premium over conventional sovereign bonds that governments must pay investors to compensate them for taking on the exposures inherent in GDP-linked bonds. “There are various types of risk but they aren’t unique only to growth-linked bonds – default risk, novelty risk, growth risk, liquidity risk. So we are trying to find a new equilibrium with a sufficiently low premium to make the instruments attractive to both governments and investors,” says ICMA managing director and general counsel Leland Goss, who adds: “I think we are getting closer to that.”

The premium will differ depending on the final terms and issuer, but ballpark estimates vary wildly. Patty Cao, a research analyst at Aberdeen Asset Management who has been consulted by the working group, suggests between 100 and 200 basis points (bps). Mr Kopf disagrees. “My guess is that it will be much less. Of course, it will be determined in the marketplace at the end of the day. But for the UK or France, we are probably talking about 20bps or 30bps,” he says. That is where inflation linked-bonds have stabilised so he expects something in that neighbourhood, even if rates were to rise.

Proponents warn against conflating the types of premium that will diminish over time. “Maybe investors don’t distinguish between these but from an analytical perspective it is good to do so,” says Ali Abbas, deputy chief of the debt policy division in the IMF’s strategy, policy and review department. “If the structure is sound, and novelty and liquidity issues can be resolved over time, then what premium would they demand?”

Critics query why governments would print debt that is more costly than conventional bonds. To measurably improve debt sustainability, they suggest a large chunk of outstanding bonds must be in GDP-linked format. But it is wrong to assess the impact of these instruments in a vacuum. “You can’t just look at the risk premium for the GDP-linked bond,” says Ms Griffin. “You must look at the broader debt profile of the issuer to see how it affects their conventional bonds.” Growth-linked bonds reduce the government’s default risk, including under its conventional bonds, which brings down the costs of their overall debt portfolio.

Investor concerns
The buyside has a tendency to be sceptical of new asset classes, and it has flagged problems other than the premium. For instance, GDP-linked bonds conflict with basic principles of credit investing. Some corporate bonds have step-up coupons, which see the coupon increase if the issuer or note is downgraded; GDP-linked bonds do the opposite.

“Instead of having an asset that offers support if the credit quality deteriorates, here you are having to deal with a coupon that steps down,” says Mark Dowding, co-head of investment grade at BlueBay Asset Management. “It doesn’t give you any protection. In fact, it effectively exacerbates your downside if things start to go in the wrong direction. So for those reasons you would see why investors aren’t wild about the structure and design.”

As for the term sheet, investors are concerned about the local currency element when it is an emerging market issuer. It makes foreign exchange (FX) yet another risk they must bear. However, according to Allen & Overy partner Yannis Manuelides, at least one central bank involved in consultations says this has not been problematic for foreign investors in their existing local currency instruments.

“They’ve said investors don’t worry short term about the FX situation, provided there is some confidence about there being no capital or exchange controls, and there are instruments to hedge,” he says. “There isn’t necessarily huge fluctuations in prices when FX moves.”

Aberdeen’s Ms Cao believes the biggest hurdles are valuation and liquidity. The latter would be helped by the bonds’ inclusion in indices, given money managers use these as a benchmark. But it is a chicken-and-egg scenario. “If you look at the typical index rules, the product doesn’t really belong in any index, and it will take a while before you get a critical mass to create one. It was years between the creation of inflation-linked bonds and the creation of indices,” says JPMorgan’s Mr Salford.

Lack of liquidity is one reason the working group is targeting long-term investors, similar to those active in the inflation-linked market. GDP-linked bonds are designed to be held over a number of business cycles, as this usually guarantees growth. The concept has most traction with insurers, sovereign wealth funds, pension funds (whose liabilities tend to increase in line with economic growth) and some mutual funds. Initially there were concerns that even buy-and-hold investors wouldn’t be interested, but Mr Kopf believes this has now been proved wrong. “In a situation like today when you have negative yields on government bonds, there would be a tremendous amount of investor interest in this paper and this is the feedback we’re getting from them,” he says.

Issuer concerns
Clearly, GDP-linked bonds are designed to benefit issuers but enthusiasm varies between different public sector groups. From the perspective of politicians, former IMF economist Eduardo Borensztein says it is like buying insurance: “When everything goes well, they have to pay, meaning it can seem like a waste. So for them the trade-offs are not very attractive.” Cynics also point out that sitting governments have little incentive to buy protection that benefits their successor.

It has been an easier sell to central banks. In a June speech, the president of Germany’s Bundesbank described GDP-linked bonds as an avenue worth exploring. But it has been tougher to convince treasuries, which would have to actually issue the instruments. Their public debt management offices (DMOs) have been the biggest stumbling block. DMOs are concerned about overpaying, which Mr Kopf says is fair. Their mandate is to minimise interest payments and keep within budget, not to ensure long-term debt sustainability.

To get treasuries on board, Ms Griffin wants some intense economic modelling based on an industry-endorsed term sheet. “We would like the IMF to incorporate this idea when doing its baseline projections of debt sustainability, to see how that would change if the sovereign had included GDP-linked bonds in its debt portfolio,” she says. It is hoped investors will do the same, and suggest changes to reduce the premium.

The CAC precedent
Growth-linked bonds are clearly an ambitious undertaking, but they are not without precedent. Both inflation-linked bonds, which have grown into an established market since the UK government’s first issuance in 1981, and collective action clauses (CACs) prove it is possible to change the rules of the game for sovereign debt.

CACs stipulate that a super-majority of bondholders consenting to a restructure will bind all bondholders. After a controversial US court ruling in 2014 that increased the leverage of holdouts, ICMA launched a new single-limb, aggregated CAC. This strengthened the position of sovereigns by setting the consent threshold at 75% of all outstanding bonds, rather than majority support within each series.

“It was controversial at first – initially some of the buyside were opposed to it as they thought it favoured borrowers too much,” says Mr Goss. “But after further consultations and building in certain safeguards, in only about 18 months we reached a consensus between the public and private sector, to the surprise of a number of people.”

Many balked at the idea and thought it would drive up borrowing costs, but it is now market standard. GDP-linked bonds do not have the same urgency as CACs did two years ago – a time when Argentina was locked in a battle with holdouts and Greece leaving the EU was on the cards – but the working group takes some comfort from the CAC story.

The first-mover problem
While it is right to embrace investor feedback, Ms Griffin insists issuers must drive this initiative. “If they issue it, then investors will buy it. They will be able to price it, they will figure out how to trade it. What we really need is more engagement with the issuer side,” she says. The working group is exploring various angles on how to open the market, but many members believe the best strategy is a simultaneous issuance by a group of credible sovereigns.

This would instantly boost liquidity and enable investors to diversify among countries with uncorrelated growth, which would depress premiums. Germany taking over the G20 presidency next year and adopting a theme of ‘financial resilience’ is serendipitous. Its public sector is among the most engaged in the working group’s initiatives, the eurozone’s debt-to-GDP ratio is 92% and GDP-linked bonds are particularly useful in currency unions where governments have less control over monetary policy. Germany’s borrowing costs dipping into negative territory means it has little self-interest in issuing, but in the spirit of the G20, its involvement in a simultaneous debut in future years is not being ruled out.

“If the price of seeing GDP-indexed bonds introduced in other countries is that Germany participates in a coordinated issue, that might be something worth considering,” says Dr Jeromin Zettelmeyer, former chief economist at Germany's Federal Ministry for Economic Affairs. “The idea of several countries introducing an identical bond at the same time – for a relatively small amount, a few hundred million – is not such an absurd idea. I’m not saying we are there yet, but it’s not too implausible.” If six or seven advanced countries went ahead with the plan, others are tipped to jump on the bandwagon.

The hurdles are high but not insurmountable, and the working group – particularly Mr Kopf – is sanguine. Others are also being won over. “This seems to be one of those ideas that sounds appealing when you think about it naively, then you realise the practical problems. You think through those, and after that you still find it appealing,” says Mr Zettelmeyer. “It is quite a robust idea, even if it’s not as easy as it looks.”

TheBanker.com