|Main index||Financial Markets index||About author|
Julian D. A. Wiseman
Abstract: On Friday 17th September 1999 the UK’s Debt Management Office published a paper (local copy), consulting about “a possible non-discretionary special repo facility”. Below is a copy of the reply sent to the DMO. [Following the Bank of England’s habit, the DMO’s consultative document did not cite sources, including a letter on this subject sent to the Bank of England on 15th August 1996, that has been repeatedly discussed since with both the Bank and the DMO.]
Publication history: Sent to the UK DMO, then here. Usual disclaimer and copyright terms apply.
Congratulations to the DMO on its publication of its paper on suggesting a ‘possible’ non-discretionary special repo facility. Needless to say, [my employer] whole-heartedly supports the principle. However we have some suggestions about the details.
The Principle of a Non-Discretionary Special Repo Facility
In general, we strongly believe that the market should set the price, and that governments are inefficient allocators of capital. However, this is not to advocate raw capitalism, famously red in tooth and claw. The last century’s most successful economy has had a regime of fierce trust-busting. An independent Monopolies and Mergers Commission is widely accepted to be beneficial to consumers and the economy at large, and the same can be true in financial markets.
If a security is tight in repo, it reflects the fact that there is a shortage of supply. This shortage might be driven by investor-regulatory reasons (excluding some holders of the security from the repo market), or by exchange-regulatory reasons (such as the penalties attached to futures non-delivery), or by a deliberate attempt on the part of some market participants to acquire a controlling stake. Of course, if either of the first two are present, it encourages the third.
A well-designed non-discretionary special repo facility, by preventing the creation of monopolies and deterring the attempt, will lead to a more efficient market. In function it would be very different to the Monopolies and Mergers Commission, but the benefit would be similar in nature. Further, in the UK, we suspect that the past market manipulation of the 9% Oct 2008 has contributed to the decline in liquidity, and hence would particularly welcome such a facility.
We follow the numbering of the paragraphs in the Debt Management Office’s paper of 17th September.
7. The DMO rightly states that such a non-discretionary facility would “reduce the uncertainty about the circumstances and terms of any operation by the DMO”. We agree that “availability of this facility [sh]ould not be linked with the indicators of market breakdown”, and it should be “automatically available”.
However, we are puzzled by the sentence that reads “Non-GEMMs would be able to have access to such an operation through a GEMM, and the DMO would expect GEMMs to be able to present proof, if asked, that they were ‘putting through’ any trades done for a non-GEMM”. Why? If a stock is tight in repo, the GEMMs will access supply via the DMO’s non-discretionary window, and the repo rate will become less tight. It may well be — it will usually be — that the GEMMs will on-lend borrowed stock to clients. If the DMO wishes to imposes evidential rules on ‘put-throughs’, but not on ‘on-lends’, then all of the former will become the latter, and to no benefit. Perhaps the DMO could replace this with a statement reminding the GEMMs of their obligation to provide the DMO with timely market intelligence.
Very importantly, the DMO has stated that “Such a facility would need to be unlimited in amount”, and we urge the DMO to repeat this clearly in its final operational notice.
Below, we suggest that the repos be for a longer term than overnight.
9. Pricing. We urge the DMO to adopt a pricing regime that is sustainably credible, and yet effective. In particular, the DMO’s pricing methodology should function properly under all plausible market circumstances. Under the current monetary policy regime, circumstances to consider include very low rates (Japan and Switzerland currently have sub-1% policy rates), and also include very high rates (say, in double digits). Other circumstances to consider include a very steep or a very inverted sub-2-week money-market curve; and also a run-up to the replacement of the pound (whether by dollar or euro), during which time rates might vary intra-day.
Before discussing the pricing, we remark on its quotation. The DMO is proposing to lend specific stocks against other collateral. So the cashflow in the trade depends on the difference in the two repo rates. We therefore urge the DMO to quote the rate as an amount of ‘specialness’, equivalent to the spread to GC. Note that this would avoid the need for the DMO to arrange a formal ‘fixing’ of GC. (For settlement purposes the DMO might wish to settle one leg at a rate x, and the other at the rate x-specialness. In the absence of a haircut, x wouldn’t matter, so we suggest setting x to the Bank’s monetary policy rate.)
Indeed, for the same reasons that the rate should be quoted in terms of its specialness, it cannot be a fixed rate. A fixed positive repo rate would malfunction if policy rates dropped lower than it, and a fixed zero rate would be ineffective (in that, except for futures deliveries, the DMO’s window would be no cheaper than just failing-to-deliver).
Likewise, the rate cannot be a fixed spread below GC: if GC is smaller than the spread, then failing would be cheaper (possibly substantially cheaper) than using the DMO’s facility. Again on grounds of effectiveness, this pricing can be rejected.
So the natural course is to quote the degree of specialness as a proportion of policy rates. But as a proportion of which policy rate: the rate with which the Bank starts the day, the rate to which it changes rates at 11am, or the rate to which it changes rates at 13:30? We suggest that the DMO choose the specialness as a proportion of the highest policy rate in use by the Bank that day.
But the Bank’s monetary policy interventions aren’t overnight, they typically have a 2-week tenor. Matching this would have several advantages. First, it would ensure that the special repo facility functions naturally even when the money-market curve was highly distorted (as would happen under an FX-targeting regime, and as happens now in the run-up to an MPC meeting which is widely expected to result in a rate move). Second, it would reduce the DMO’s administrative overhead: roll-overs would occur every 10 business days rather than every 1 business day. Third, by increasing the cost of using the facility, the it allows the DMO to choose a more generous degree of specialness.
So we suggest that stock be lent for stock, for a term equal to that of the term of the Bank of England’s monetary policy operations, at a degree of specialness which is a fixed proportion of the highest monetary policy rate in force on that day.
And what proportion should that be? The more generous (the smaller) the degree of specialness, the ‘better-behaved’ will be term repo. Markets in listed futures and unlisted options are highly dependent on functioning term repo. Because of our desire that term repo should be liquid and well-behaved, we believe that a specialness of 25% of the Bank’s monetary policy rate would be appropriate. For example, with policy rates at 5.25%, the DMO would supply specific stocks at 131.25bp special. Allowing for compounding, this is 137 pence over a year, worth some 18.3bp of yield on a 5.75% par 10-year gilt. This is substantially in excess of a typical benchmark premium, and we believe it to be a reasonable floor for the DMO to impose.
So we advise that stock be lent for stock, for a term equal to that of the term of the Bank of England’s monetary policy operations, at a degree of specialness which is one quarter of the highest monetary policy rate in force on that day.
• For how long will this facility be made available? Constitutionally, the DMO cannot commit to it continuing in perpetuity. But the policy would be more credible if it were believed that it would last for a long time. Perhaps the DMO could consider a phrasing such as: “Constitutionally, the DMO cannot commit to the indefinite availability of this special repo facility, and hence the DMO reserves the right to withdraw or alter the facility at no notice. However, it is the intention of the DMO to give substantial notice before withdrawing, restricting the availability of, or increasing the cost of this facility. This leaves open the possibility that the facility might be cheapened at no notice”.
• Which stocks should be available under this facility. We note that some stocks, such as 2¾% Annuities 1905+, have conditions attached to the call that depend on the amount outstanding, and hence there may be legal implications associated with any change in the issue size. The simplest arrangement would be to make available to the GEMMs all non-rump non-ILG gilts, and to make available these stocks plus the non-rump ILGs to the IL-GEMMs.
Strips would be included, and the DMO’s operational notice should state that all strips that are not explicitly excluded are available. (It should be borne in mind that the DMO might wish to exclude the principal of a strippable rump.) The broad principle is that a GEMM can access any stock in which it is expected to make a market.
We make no comment on the possibility of extending this facility to Treasury Bills.
The moving of a rump to the excluded list requires some care. The DMO should commit to consulting with market participants before doing so, and to “being mindful of any GEMM’s bone fide reasons for delaying exclusion”. This avoids tying the DMO’s hands, whilst reassuring the GEMMs that a position in a particular stock will not suddenly become stranded.
The DMO’s paper finished by asking a number of questions, which we answer here.
• Do gilt and repo market participants believe that the introduction of an automatic special repo facility, at a known premium, is necessary or warranted? Yes, very much so.
• Should such a facility be available only to the to GEMMs who had signed the appropriate agreement with the DMO? Should the facility be extended to other interested repo counterparties who had signed the relevant documentation? Only to the GEMMs. This special repo facility is not a cash-management tool, but is a gilt market tool. The GEMMs enjoy the privilege of an exclusive gilt dealing arrangement with the DMO, and we believe that this special repo facility will become a key privilege of GEMMhood.
• Given the need for an automatic facility to be sufficiently penal to avoid substituting for normal market conditions, how penal should the rate be? Should it be expressed as a fixed overnight rate (eg. 0%) well below GC overnight rates; a fixed number of basis points below either GC or the Bank of England repo rate (eg. GC-500bps); or possibly a fixed percentage (eg.10%) of GC or Bank of England base rates (thus, with current base rates at 5.25%, the DMO’ s special repo rate could be fixed at 0.525% overnight)? The facility should have the same tenor as the Bank’s monetary policy window, and should be at a specialness of one quarter of the highest monetary policy rate in operation that day.
Julian D. A. Wiseman, September 1999
[Postscript: 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 such a non-discretionary repo switch facility; see also this author’s commentary on the DMO’s February 2000 decision.]
|Main index||Top||About author|