|Financial Markets index
Julian D. A. Wiseman
Abstract: On 29 June 2011 the UK DMO published CPI-linked Gilts: A Consultation Document, which asked whether and how the DMO should sell gilts linked to CPI. This is the reply from Julian Wiseman, soon-to-be the Head of UK Rates Strategy at Société Générale. Société Générale ask that this also do duty as the GEMM’s official reply.
National Statistics are in the process of building CPIH, a variant of CPI that includes housing costs. It might be that the Department for Work and Pensions will adopt this as the statutory measure by which pensions are to be revalued and indexed. It might be that the Chancellor of Exchequer might adopt this as the variable targeted by the MPC. If these were to happen, the (premature) issuance of CPI-linked gilts would be regretted. Hence the DMO should wait.
Publication history: sent to the DMO as a PDF; also at www.jdawiseman.com/papers/finmkts/20110919_cpih_gilts.html. Usual disclaimer and copyright terms apply.
Contents: National Statistics and CPIH; The DWP and the BoE’s MPC; The DMO’s decision; Conclusion: Wait; A Lesser Matter: CPI revisions; Afterword.
This reply is sent by Julian Wiseman, who is soon to join Société Générale as Head of UK Rates Strategy. Because that employment has not yet commenced, the author cannot yet write in that capacity. Nonetheless, it was written at the request of Société Générale, and so the DMO is asked to allow this also to do duty as Société Générale’s reply.
This response is also published via www.jdawiseman.com/20110919.html, and the DMO is welcome to quote from or re-publish it, with appropriate attribution.
As explained in the consultation document, National Statistics is in the process of creating a variant of CPI including housing costs (‘CPIH’). Indeed, in November 2010 the Consumer Prices Advisory Committee requested that the work be completed “in the next two years”, that being the end of 2012. The construction of CPIH depends on decisions by National Statistics, but the actual timing might be influenced by the prospect or otherwise of international agreement. In particular, NS is trying two different methodologies, of which Eurostat has been pursuing only one. There is the possibility of a relatively speedy agreement on one methodology; and there is the possibility of the UK preferring the other, that slowing international concurrence.
So it is likely that, within a small number of years, National Statistics will start publishing one or more versions of CPIH.
Pensioners need to live somewhere: pensioners incur housing costs. Hence the Department for Work and Pensions might, after CPIH has existed for a year or so, prefer that as the statutory measure by which pensions are to be revalued and indexed. It is possible.
There are also macro-economic advantages in having the MPC target CPIH rather than CPI. Doing so would, at least in theory, slightly slow fast growth in house prices, and slightly soften fast falls. Hence the Chancellor of the Exchequer might change the variable targeted by the MPC. It is possible.
Given these, it is possible that, within a few years the appropriate index might change from CPI to CPIH, or to some other variant. Even if it difficult to estimate the probability of CPIH being the relevant index, for the purposes of this reasoning it suffices to conclude that the possibility is not remote. It might happen.
If the DMO were soon to issue CPI-linked gilts, and subsequently the relevant index were to become CPIH, the DMO would regret this fragmentation of the gilt market, from which recovery might be difficult, expensive, or both.
But if the DMO were to delay issuing CPI-linked gilts, and National Statistics were to decide not to produce CPIH, or the DWP were to decide not to use it, then the DMO might regret the delay in issuance.
Which regret would be worse? The former regret is not only possible, but would be a far greater regret than the latter. It would be would be a very unwise fragmentation of liquidity to have all of: (very new) CPIH gilts; (newly unloved) CPI gilts; and (two variants of) RPI gilts. Rephrased, issuing CPI-linked gilts now is taking a significant risk for insufficient gain.
So our advice is that the DMO should wait. Wait for decisions about CPIH, and if it is created, to have existed for a year. Then encourage the DWP to make a decision that the DWP says is likely to be durable. Encourage the Chancellor to give similar guidance about the MPC’s target.
If National Statistics is ready in 2013, CPIH will be a year old in 2014, and these decisions might happen in 2015.
And then, and only then, in the light of those decisions, should the DMO revisit the question of gilts linked to an index other than RPI. Until then, the DMO should wait.
On 13 July 1995 the Central Statistical Office, as it was then named, announced that the RPI levels from February to May 1995 were slightly wrong. The Bank of England then announced that the higher revised numbers would be used for the relevant coupons, affecting six gilts. This was the right thing to do: triple-A countries always decline opportunities to rob investors.
However, that Bank of England announcement also said:
The calculation of accrued interest on deals entered into after this announcement should be based on the above figures.
So, for the affected gilts, trades done early that day had different accrued interest to trades done late that day, complicating settlement. It would have been easier if the new accrued interest had taken effect at the start of the following day.
If the DMO were to issue CPI- or CPIH-linked gilts, and were to use the latest price series, revisions should take effect at the end of the day, rather than immediately.
This request would be amply and most easily satisfied by adopting the US/French practice of indexing exclusively to the first estimate of the price level.
|— Julian D. A. Wiseman
19th September 2011
(soon to be of www.sgcib.com)
On 29th November 2011 the Debt Management Office published CPI-linked Gilts: Response to Consultation (local copy).
After careful consideration and taking into account the feedback received to the DMO’s consultation, the Government has taken the decision not to issue CPI-linked gilts in 2012-13. The Government judges that issuance of CPI-linked gilts in the near-term would be unlikely to be cost-effective and would involve a number of risks, although it is possible that this could change over time. The decision not to issue CPI-linked gilts in 2012-13 does not preclude CPI-linked gilt issuance in the medium term, should there be a case to do so. The Government will therefore keep the case to issue CPI-linked gilts under review.
Reasoning included prominent mention of:
… the key uncertainties around the future composition of the Consumer Prices Index (CPI) …
An excellent decision.