Main index Financial Markets index About author

Gilt Asset Swaps: DMO Should Profit

Julian D. A. Wiseman

Abstract: rather than selling long gilts at swaps+40bp, the UK Debt Management Office should sell short-dated gilts, and pay fixed on the forward-starting swap.

Publication history: PDF version sent to the UK Debt Management Office, and simultaneously published at jdawiseman.com, also in HTML form. Usual disclaimer and copyright terms apply.


Even without considering the unprecedented actions by the authorities, these are interesting times. Amongst the many strange and undesirable outcomes of the credit crisis is the fact that long gilts are trading above swaps: the newly issued 4¼% Dec 2049, for example, by about +40bp at time of writing.

This has two bad effects for the UK government.

But both these points can be ameliorated, simply and profitably, by the UK Debt Management Office. Issuance of gilts is set to increase, and substantially. The DMO should make whatever decision it would take about how the extra funding is to be split between the maturity bands. But then the DMO should sell no long gilts. None at all. Whatever 2049 gilts would have been sold should be replaced with a sale of short-dated gilts (probably maturing 7th December 2011), and also paying fixed on an interest-rate swap forward starting on the maturity date of the short gilt, and ending on 7th December 2049.

There are some relevant details.

  1. Before announcing the scheme the DMO should ask LCH.Clearnet to extend their GBP SwapClear service to ≥2049 or to ≥41¼ years. (Happily, this goes with the general regulatory preference for centralised clearing.)

  2. All swaps traded by the DMO would be cleared through SwapClear, which therefore would be the DMO’s sole credit exposure.

  3. The swap auction should be open to all dealers able to novate positions to LCH.Clearnet.

  4. The swap auction does not need to be done simultaneously with the ‘stub gilt’, but might be done at a nearby time.

  5. Because of the novelty of a swap auction, they might be smaller than has been typical for a gilt auction, perhaps being only £1bn, but more frequent.

  6. The DMO’s announcement should make clear that this action is caused by the dislocation in the gilt swap spread. An easing of the dislocation, and certainly gilts trading dearer than swaps-25bp, would cause a reversion to the more usual funding policy.

  7. Further, the DMO should say that it envisages that, after normality has returned and stayed for a while, there would be auctions for the DMO to exchange its paying-fixed positions with the creation and sale of 2049 gilts. However the DMO anticipates that it would consult with the market before so doing.

Substituting the sale of gilt duration for the sale of swap duration would have two advantages:

It would be unprecedented, but so are many of the market moves and the authorities’ responses to them.

Julian D. A. Wiseman
New York, October 2008


Main index Top About author