Both government and corporate issuers of inflation-linked bonds have had great success. Investors seem to want more and more of the product. So why aren’t more borrowers coming to market?
Inflation-linked debate Participants
MR, Total Derivatives The inflation market has grown dramatically in size and sophistication in the past few years. What have been the drivers?
AM, DMO This financial year we are issuing £17.25 billion of index-linked gilts, which is roughly as much as in the first four years of existence of the DMO put together. This is partly a reflection of the overall increase in the borrowing requirement but also an indication of the value that we attach to index-linked debt in terms of portfolio diversification, and in terms of the fiscal insurance properties it brings to government. One can also add another argument in favour of index-linked issuance, in a context of asset-liability management, which is that many government income streams are more or less indexed, and therefore having index liabilities is not unreasonable.
MR, Total Derivatives How do your two debt agencies decide where to issue on the curve?
AM, DMO Our objective is to minimize the government’s cost of funding, subject to risk. To do so, we aim at extracting a liquidity premium – which is why we issue large benchmark bonds – and a preferred habitat premium where we can. To achieve the latter, we try to assess at which part of the curve investors are willing to give up a premium to hold our bonds, and that informs our annual financing programme. We also take into account that we are a repeat issuer, so place our decisions in a long-term context. We put particular emphasis on regularity of issuance, consistency of policy over time, as well as ensuring that we maintain an efficiently functioning market structure.
AS, National Grid Why don’t you take any views on the shape of the yield curve in terms of where absolute rates are?
AM, DMO Because we do not think that we are better informed than the rest of the markets as to where growth and inflation will be in the future.
AS, National Grid There could be structural reasons, though, why the curve is in a different shape.
AM, DMO Indeed, they would result in, for instance, a negative term premium. We do form a view on that – we implicitly form a view on where we believe the term premium to be, and that affects the skew of issuance every year.
BC, AFT Our approach is very similar. Our issuance policy is constrained by the granularity of the market, that is, we are issuing large benchmarks. Now, how big is enough? The market tells you, and it has to do with the liquidity premium that Arnaud mentioned. There is an optimal size for a benchmark, and we found out that it was somewhat smaller in the inflation market than the conventional market, say, around €15 billion. So we have to commit to bring liquidity in turn to all segments of the curve and then at the margin we have to take timing decisions based on market demand. But it’s only at the margin.
AM, DMO This year we have increasingly focused issuance on benchmarks at key maturities, which means currently at 10, 20, 30 and 50-year in the index-linked curve. We have become ever more explicit on that and this has brought our issuance in the real curve more in line with the issuance policy that has been in place in the nominal curve for a while. It is, to a certain extent, a normalization of the index-linked market as it has matured.
MC, RBS Yes, initially issuers chose to issue little and often, at least relative to conventionals. But in markets such as the UK, where there has been a greater supply of inflation-linked, we should now be able to tolerate larger size of average issuance. Counter-intuitively, the fact that inflation-linked issues are a little less liquid than nominals perhaps means that the market needs a greater average outstanding issue size in inflation-linked. But having said that, redemption management is an issue. It’s not just the cashflow impact of large infrequent redemptions – it is also the issue of concentrated RPI print or HICP print risk. So maybe the governments could issue large-ish infrequent points along the curve to maintain benchmark size in inflation but use inflation swaps to disperse that inflation print risk across the curve.
OPB, Santander Well, I think the DMO structure is quite transparent. From the market-making point of view, the only question from some traders is why don’t DMOs issue more short-term linkers, to cover and trade the seasonality. This is going to be a problem for derivatives books from 2008 onwards.
BC, AFT In the eurozone both Italy and France have issued short linkers. We have an EI10, a three-year linker, which we might tap any time this year. I suspect that the answer to Oscar’s question also has to do with the lack of a futures market for inflation.
OPB, Santander Yes, none of the futures is liquid enough for the market-makers on the derivatives desk. It would be helpful to have larger issuance in the short term.
MR, Total Derivatives From the investors’ side, when you go to meetings with the debt agencies, is there a consistent plea for more long-dated gilts or more short-dated gilts?
SJ, BGI We don’t have any particular criticism of the way the DMO is handling the demand for long-dated inflation.
MC, RBS In the UK, there’s an impression from both investors and banks that the market needs maximum size and maximum duration in both conventionals and nominals. Last January was perceived as a time of high stress, and since then long-dated yields have risen somewhat. However, the distant forward nominal yields, say between 30 years and 50 years, are very close to the levels last January. So those long-dated forward rates are as stressed as they were a year ago. But in the inflation-linked side of things, that’s not true beyond 20 and 30 years, because we’ve had a lot of non-government supply of inflation in the 40 to 50 year. What is highly visible in the inflation swaps curve is forward break-even inflation rates which are very high out to 20 years, but then fall quite sharply. This suggests the market needs a lot of ultra long-dated nominals, but it’s not at all clear it needs a lot of ultra long-dated inflation. It needs more 20 to 30 year. The UK government’s liabilities are growing in duration quite aggressively year by year, and this strategy would temper that rate of growth.