Inflation-Indexed Bonds

This FT story reports that Ireland may issues inflation-indexed bonds.  I have advocated such a move in  this paper;  an interesting dimension is whether such bonds would be indexed to euro area inflation or domestic inflation.

25 replies on “Inflation-Indexed Bonds”

Inflation linked bonds appear to have the following benefits:
1. Investors are protected against default through inflation.
2. Sovereigns get to access funds for longer terms allowing them to ride out the bad times.

However, there are also risks:
A. Countries don’t benefit from inflation in the way that they traditionally benefit.
B. They may amplify problems associated with stagflation, which is surely a real possibility for Ireland.
C. The rate of inflation for the Eurozone may not be linked to economic performance of the individual country issuing the bonds.
D. Economic health and inflation are not necessary directly linked in the way that interest rates may be.
E. Ultimately, de facto default through inflation is ruled out which may lead to more de jure defaults if countries issue more and more of such bonds.
F. Very large inflation is a possibliity if the Eurozone cracks up. This is not as much of a risk for Germany which may risk currency appreciation rather than devaluation in the case of Ireland.

Obviously, a low proportion of inflation linked bonds compared to overall issued bonds may mitigate against such risks.

My view is that any issue of inflation linked bonds should be seen as the Sovereign trying to play the markets to its advantage, rather than the issue of inflation linked bonds being a general good and risk reducing action.

BTW – How does deflation typically affect inflation linked bonds? I assume it is an “upward only” adjustment?

Philip, France issues bonds linked to French HICP and Euro-area HICP. There seem to be distinct markets for both, and no reason why Ireland should not issue both.

It strikes me that countries should be trying to co-operate on regulation of the bond markets to reduce contagion and amplification problems. Inflation linked bonds might be seen as evidence greater competition between nations whereby some are offering enhanced rights for bondholders when selling of bonds. Such divergence and competition does not bode well for greater international co-operation and regulation of bond markets. This in turn does not bode well for greater international financial stability.

It does make sense that some would like to be insured (for lack of a better term) against inflation risk. However, introducing another complexity (should a bond be issued like this or not) is unlikely to lead to greater efficiency.
I’d expect any efficiency gain to be countered with an equal or greater loss from higher transaction costs. I’ve no doubt that bondtraders, funds & banks will argue that due to greater complexity of the transaction they should get paid more as the higher complexity requires more skill and skill has to be paid for.

It is to be noted that the transaction cost associated with de facto default through inflation is very little. One expects that the transaction costs involved in de jure defaults are much greater. Obviously no-one likes to talk about default but it is a regular occurence as shown by Rogoff and Reinhart.

@colm mccarthy – there is reckoned to be a demand for specifically Irish index-linkers, coming from asset/liability matchers. As their requirements for a eurozone linker are already met elsewhere, demand for such issuance from the Irish government would be a lot more price-sensitive (i.e. costly) than for one linked to domestic inflation, where the gov’t would be effectively the only “quality” issuer in town, facing no competition.

To estimate whether or not an inflation linked bond is a good investment it is required to be able to estimate future inflation. With or without the inflation linked bonds, future inflation has to be estimated. Where will the benefits be?

This is only leading to additional complexity & more complexity leads to higher transaction costs & speculation. It is not possible to remove a risk associated cost for free. Risk-associated costs will not go away by clever modelling.

Is it in anyway linked to the state starting to sell annuities as part of the pension protection scheme?

This highlights that Irish economists need to rethink what they mean by “domestic inflation” now that we are in a currency union with increasingly close economic integration. This affects how inflation should be measured, how it should be reflected in contracts (including government bond contracts), inflation’s business cycle links, and best policy responses.

In the USA, state-specific inflation gets little attention, although many US states are substantially larger that the Republic of Ireland in terms of population and output. But of course state-specific inflation (e.g. California) matters a lot and can have nasty effects on state-specific income and unemployment. There are no US state government bonds tied to the state-specific inflation rate.

Eichengreen has an old paper on inflation differentials in the case of Canadian provinces (these provinces have very different inflation rates in some time periods).

I cannot help thinking that issuing “linkers 2 would not be a good long term strategy giving all the quantitative easing in the works. It could end up being very expensive funding. Far better to stuff the market with long term fixed rate paper.

While inflation-linked bonds may not be the optimum solution for everybody, (buyer or seller), they would offer real attractions for some cautious pension investors. After all, we know now that equity risks are real and returns above inflation are not reliable, even in the medium-term.

To make sense for pension investors, charges by the pensions industry would need to be low relative to the nominal return.

The government’s solidarity bonds, promised for early launch, could be an opportunity for a test of acceptability at the retail level.

There is a large latent market for Irish inflation.

I suspect there would be some significant advantage to such a bond issue, but maybe difficult to sell just at the moment, given the ongoing deflation risks in Ireland.

The tax situation of these bonds are interesting. In the UK linkers are only taxed on the real yield. If this is implemented it would make such Irish bonds very attractive from a retail perspective.

As for the sporadic deflation comments in this thread: This “fear of deflation” is largely nonsensical. Deflation does not keep people from spending – they always spend what’s necessary. And money NOT “spent” is then saved which means it is credit to someone who invests it for capital goods etc. thus it is again being spent, only not for consumption. Money never lies completely idle to any extent whether there’s inflation, deflation, stability or a solar eclipse. For deflation to seriously happen, not only the current extreme credit expansion by the central banks and states (through “quantitative easing”, stimulus packages, monetising and then spending national debt etc.) but also the money that was released into the economy PRIOR to the collapse would have to be “mopped up” again. This is nowhere to be seen nor would it be technically possible (confiscation aside) so we will rather see inflation than deflation.

Aiman, fair point about Euro HICP, the demand would be greatest I am sure for an Irish HICP issue. It would also permit the replacement of lump-sum court awards with index-linked periodic payments. The natural holders would favour long-dated issues, so there could also be an opportunity to increase duration at possibly lower cost than with conventional bonds.

By the way, the FT story looks like pure speculation from Barcap – has NTMA actually made an announcement?

@jesper – I don’t believe there’s any intent to replace fixed-rate funding with index-linked. It is intended, I’d imagine, to supplement the existing stock of fixed debt and thus reduce the risk somewhat. I would also expect that actuarial-driven demand for duration and certainty might cloud the purchasers’ judgment when it comes to pricing, to the benefit of the issuer.

(A reductio ad absurdum on your own logic would have the State borrowing only overnight funds, due to uncertainties about the future course of interest rates).

@Aiman,

The inflation risk will always be there, even for indexed bonds:

Assume you have an expectation of inflation of 3% & assume the unindexed bond rate is 4%. An indexed bond of inflation +1% will then give the same interest. However, inflation might turnout to be 3.5% and then the indexed bond is better for the bondbuyer. If inflation turns out to be 2.5% then the indexed bond is worse for the bondbuyer. With the uncertainty of the estimation it is impossible to accurately evaluate which option is best, indexed or not.

Risk & uncertainty cannot be modelled away. There are plenty of hedgefunds who’d agree with you, however there will also be those who will make massive losses in believing they have the best model for estimating inflation. I.e. issuing fixed rate to buy indexed or the other way around and getting the estimation of inflation wrong. Counterparty default risk and insurance against that & away we go 😉

Indexed bonds are just another speculative instrument. The only guaranteed winner in a casino is the house who always takes its cut. The speculators will gamble for the rest.

@jesper

Fixed rate bonds are equally a speculative instrument, for both issuer and lender. Who knows where either interest rates or inflation will be over the lifetime of the bond? Either the issuer or the borrower gets stung (assuming that the coin never lands on its edge).

Like I said above, the sure way of avoiding this nasty reality is to stick to overnight borrowing/lending…but that’s not really practical, is it?

@Aiman,

I’m not going to defend a position I never took 🙂

What I am saying is that risk will always exist. Adding complexity will only add complexity, the risk might be more difficult to see but it will still be there. Risk is part of life. The choice isn’t to avoid risk completely, the choice is which choice (risk) is the better option.

There is no magic solution & therefore I’m quite confident in stating that inflation indexed bonds is not a magic solution for anyone except lazy investment managers. The rest will do their due diligence and take advantage of those who believe their complicated model makes due diligence unnecessary.

@jesper – not looking for a row! I think the issuance of linkers will be small relative to total funding, but will tap in to a source of funds not currently available at what, by present servicing cost measures, will be cheap money. Time alone will tell whether it actually is cheap, but this is also the case with fixed rate funding.

Going by my own experiences with actuaries and their world view, they’ll be queuing up to buy the duration/liability-matching element of such issuance, with less regard for price than their underling fund managers would have.

Comments are closed.