|Financial Markets index
Julian D. A. Wiseman
Abstract: On 22nd February 2000 the UK’s Debt Management Office published a “Response to DMO Consultation Document on ‘Special’ Gilt Repo Operations” (local copy), announcing the introduction of a non-discretionary repo switch facility. This excellent decision should be emulated by others managers of government debt. However, the price of the DMO’s facility is excessively far from generous, and the term of the facility should have matched that of the central bank’s monetary policy operations.
Publication history: Only here. Usual disclaimer and copyright terms apply.
Official Intervention In The ‘Specials’: the DMO’s decision
On 22nd February 2000 the UK’s Debt Management Office published a “Response to DMO Consultation Document on ‘Special’ Gilt Repo Operations” (local copy), announcing the introduction of a non-discretionary repo switch facility. The DMO’s February 2000 paper follows from its paper of 17th September 1999 (local copy), to which this author replied on 20th September 1999. It is believed that the DMO’s original source was a letter to the Bank of England dated 15th August 1996, which was subsequently the subject of a number of discussions with both the Bank of England and the DMO.
The DMO’s facility allows market participants to borrow any gilt from the DMO, in unlimited size. In return the market participant must give the DMO another gilt as collateral, with a spread between the repo rates of 90% of the BoE’s official monetary policy rate. The facility is overnight, “with the option of rolling-over the facility indefinitely. However, the DMO does not expect any participant to require the facility to be rolled over for more than a continuous two-week period and would review its use if it did.”
This facility ensures that there is a price at which market participants can access any gilt. Because this facility exists, no market participant (or consortium thereof) can corner the entire supply of one stock; because such corners are impossible, such ‘corners’ or ‘squeezes’ are less likely to be attempted; and hence one would expect the facility to be used at most rarely.
This author has some minor squabbles with the details of the DMO’s facilities. However, it should be stressed these squabbles are minor in comparison to the benefits that flow from the existence of the facility.
At 90% of the BoE’s repo rate, the price of the facility is near-maximal. It will deter extreme squeezes, particularly those relating to futures delivery, but a more generous price would have encouraged a more orderly market. This author advises other debt management agencies to lend at a specialness of 25% of the official monetary policy rate.
The Bank of England intervenes at a two-week maturity (which she shouldn’t, but that’s a separate story). Overnight money can drift below the BoE’s monetary policy level. Indeed, the fixing of the average cost of unsecured money, SONIA, traded below 90% of the BoE’s policy rate on 31 of 663 non-MPC business days between 6th June 1997 and 10th Feb 2000. Secured overnight money (for which this author lacks quality data) would have traded below 90% of the BoE’s much more often. So the DMO’s implied repo rate for specials (market GC minus 90% of the BoE’s repo rate) would have been negative more often the 5% of the time. The real solution would be for the Old Lady of Threadneedle Street to intervene in overnight rather two-week money — perhaps one day she will be persuaded.
The DMO clearly desires that this facility only be used over short horizons: see above quotation. Why? Private-sector use of the facility would be profitable to the DMO.
Perhaps the DMO was concerned about the effect on gilt indices. But the providers of gilt indices could not include this temporary stock in their indices, for two reasons. First, the DMO does not publish details of the collateral proffered to it, so the index provider would be unable to do both parts of the substitution. Second, to include the created stock would be to allow private-sector manipulation of the quantity of government debt in the indices: not something that the index providers would welcome.
Thus there is no reason why the DMO should have discouraged long-term use of this facility.
However, despite these squabbles, the DMO is to be congratulated on the introduction of this facility. Other government-debt managers are encouraged to introduce a similar facility, perhaps even incorporating the above objections.
Julian D. A. Wiseman, March 2000
Postscript added January 2001. This standing repo facility was first used on 29 December 2000, when the DMO announced that
An additional STG 1146.5 million of 5 3/4 % Treasury Stock 2009 has today been created and made available to the market under the terms of the DMO’s standing repo facility that was introduced on the 1st June 2000. … The DMO has been given general collateral at the Bank of England’s repo rate against the stock it has lent.
So the DMO borrowed money at 0.60% (against collateral of the 5¾% Dec 2009), and lent at 6% (against general collateral), from Friday 29 December 2000 until Tuesday 02 January 2001. In size of almost £1.2bn, 540bp for 4 days produced a profit for the taxpayer of about £0.7 million.