Seldom has it been more important for a borrower to ensure a successful start to a new funding programme. When the UKs Debt Management Office (DMO) kicked off its syndication programme for 2009 with its £7bn 25 year transaction earlier this month, the general consensus was that the benchmark was a spectacular success.
Led by Barclays Capital, Goldman Sachs, HSBC and Royal Bank of Scotland (RBS), the UKs deal ticked all the right boxes. The issue, which was the largest ever syndicated transaction from a European sovereign, generated demand of more than £15bn from a diversified range of investors, with more than 60% placed with pension funds, insurance companies and fund managers. This robust demand allowed for pricing to be set at 11bp over the 2032 Gilt, below the original guidance range of 12bp-15bp over the UKs government curve.
Just as well. Because with £220bn of Gilts to sell this year alone, this will not be the last syndicated benchmark that the DMO will issue in 2009.
Participating in the EuroWeek/UK DMO roundtable to discuss the mechanics of this months blowout deal, and the UKs future borrowing strategy, were:
Robert Stheeman, chief executive, United Kingdom Debt Management Office (DMO)
Andrew Roberts, director, global fixed income strategy, Banc of America Securities-Merrill Lynch (BAS-ML) Global Research
Francis Todd, executive director, fixed income, currency and commodities, Goldman Sachs
PJ Bye, managing director, debt capital markets syndicate, HSBC
Kevin Mountain, executive director, UK inflation trading, JPMorgan
Myles Clarke, joint global head of frequent borrower group, Royal Bank of Scotland
Peter Allwright, fund manager, currency and interest rates, Threadneedle Asset ManagementPhilip Moore, contributing editor, EuroWeek
EUROWEEK: Last weeks 25 year deal for the UKs Debt Management Office (DMO) is widely regarded as having been a phenomenal success. What was it that made the deal such a success?Stheeman, DMO: From our perspective it was very important for the deal to be a success because it was the first transaction in the syndicated part of the programme. In order to set the right tone for the rest of the programme it was essential to make sure that we had a very well-received transaction.
I think it was successful because of a number of factors. We consciously chose a maturity which we thought would be a clear favourite. We had a 2032 and a 2036 outstanding and there was a gap at the 2034 maturity. We also deliberately chose not to go for a maturity within the Bank of Englands purchase band. These were important ways of signalling to the market that this was an issue that could stand up on its own.
Clearly the size was also very large, which was important in sending out a signal about the initial liquidity of the issue. When we do non-conventional Gilts the size is usually initially in the £2bn-£2.5bn range. A deal of that size is not strippable and seldom becomes very liquid until the second or third auction.
Another slightly less definable reason for its success is that the tone in the market is now much better than it was a few months ago. Looking back almost exactly three months to the time of our uncovered auction, every debt capital market seems to be in much better shape now than it was then. And that is true not just of the Gilt market but also of US Treasuries and German Bunds. There was certainly a much better feel about the market in early June than there was in March.Bye, HSBC: The amount of work that was done in advance, together with the market consultation exercise was also very important because it helped all the banks and investors buy into the process. Because we were all in it together it was essential for the whole market that the trade worked well.
A book of £15bn-plus for a new deal in a 25 year maturity is amazing. It was the largest ever syndicated sovereign deal and all the others that come close in terms of their size have been shorter dated transactions. Ten years is typically the maturity that would attract a book of this size. It is an anomaly of the UK market that a 25 year deal can attract that amount of interest, and I think that is attributable to the work that was done by all parties in the run-up to this trade. It was meticulously planned, extensively discussed and the timing was excellent.
EUROWEEK: How long was the preparatory phase for the deal? You started talking to the market about it a long time ago, didnt you?Stheeman, DMO: The fact that a syndicated programme was part of our funding plan for 2009-2010 was originally flagged towards the end of last year.
It was after the Budget in February that we sent out a clear message that we would be launching a syndication programme, although not before the beginning of June. As usual we held our quarterly consultation meetings with all the GEMMs [Gilt-Edged Market Makers] in May, which provided a useful forum to discuss the programme and first transaction with the market. We followed that up with a series of bilateral discussions.
Once we had made the decision on which banks we wanted to appoint as lead managers, we told them on Monday June 8 and the formal announcement came the following day. So I think it was a very well flagged and transparent transaction.Todd, Goldman Sachs: It had been known since May that a 25 year gilt was coming, and I think that six or seven weeks is enough time for the GEMMs and investors to start thinking about how they might manage the demand. Roberts, BAS-ML:Whether it was coincidence or not that process coincided with the zenith of the global bond sell-off.Mountain, JPMorgan:Thats right. I think people had almost lost sight of the fact that 10 year Gilt yields had risen more than 100bp from their trough. And in modern day terms, a 100bp rise in yields is an enormous bear market. So I think that somewhat serendipitously the DMO hit a really sweet spot. Clarke, RBS: Its also worth noting that one of the main attractions of using the syndication process is that it allows you to insulate yourself from some of the timing issues that arise when you do an auction straight on to the Street. Some Gilt investors only experience of primary markets is through auctions. That contrasts with Europe where you have government bond investors who are also very au fait with how the market for syndicated issues works. So in some instances in the Gilt market there needs to be an education process as to why the syndication method is being used, explaining to investors how they can express their opinions, their biases, and their preferences in the market.Mountain, JPMorgan:I think there were some clients who initially grumbled about the syndication method. They said that effectively everyone was going to get the bonds at the same price, which would mean nobody had the chance to outperform. Theyre used to playing auctions and to using tactics applicable to the auction market. But, by and large, I think everyone came out of the process thinking that it had been a success from every perspective. Investors were allocated the bonds they wanted at a reasonable price and it was an incredibly orderly market for such a big issue. Often Gilt auctions create enormous volatility and that all seemed to be swept away by the syndication process.Bye, HSBC:You could argue that the one-week period between when the mandate was awarded to the banks and the actual trade was unusually long. As Myles says, we had to do a fair amount of education for investors who werent completely familiar with the process. I think some of them were already getting bored with that by Thursday and were asking where the deal was. So by the time we got to the launch day everyone knew everything there was to know about syndicating a long dated Gilt. We had a massive rush into the book. To get it over the £5bn mark within an hour was fantastic, and is unlikely to happen with the next deal. Nor will we have the same kind of media impact that we had. We had calls from papers like the Telegraph and The Times, proving that the transaction had caught the imagination of the general public. You never usually see that kind of national media interest in a sovereign syndication.
EUROWEEK: Was the glare of public scrutiny uncomfortable? Were there difficult questions to be answered about how good a deal it was for the UK taxpayer? Stheeman, DMO: Not directly, although there was a parliamentary question about how much was paid in fees to the syndicate.
We feel quite strongly about the debate over value for money, which is extremely important to us. It is not just a question of how much we paid explicitly in fees in a syndicated transaction relative to an auction whereby you pay no fees at all, although markets typically charge a concession in the price set at auctions. Although that is true, you have to ask how much the taxpayer would have ended up paying overall if we had tried to auction £7bn in one go. I will never be able to prove conclusively that we were able to save x or y for the taxpayer. But I do think there was a real sense that this represented good value for money overall.Allwright, Threadneedle Asset Management: From our perspective the fact that the market was able to take down £7bn also gave investors a lot of added confidence about the stability of the Gilt market. This weeks £4bn auction which was 1.7 times covered had the same effect. That was one of the best-received Gilt auctions we have seen for a very long time because the volatility in the Gilt future both in the morning running up to the auction and afterwards has been minimal. So it is clear that getting such a large deal away so effectively has injected a lot more confidence into the market. Weve seen something similar happen in the US and Germany, which underscores what Robert was saying about how confidence has returned to the market since mid-March.
EUROWEEK: Has the success of the deal led you to make any revisions to your planned use of syndications in this current programme? Stheeman, DMO: The success of the deal definitely wouldnt lead us to change our plans. What is set out in our remit, which I sometimes describe as our "marching orders", are plans to sell £25bn through syndication via up to eight offerings. But we also made explicit at the time of the Budget that 83% of the £220bn in Gilt sales in 2009-2010 will be through auctions, which remain our primary issuance method.
To make our syndication programme as a result of this deal even larger than we announced at the 2009 Budget would be committing the cardinal sin of debt managers which is to say one thing and do something else. To do that in response to a market event would look opportunistic.
Secondly, I think we fundamentally believe that auctions provide good value for money. Being predictable and being transparent actually has a monetary value to the taxpayer. So its a balancing act. We recognise that by doing a syndicated deal we are moving slightly away from reliance on the pure auction format and because of that we want to be very circumspect about how we go about it. We dont want to push it too far.
We have bidding rules and clear bidding limits in our auctions, but one element that was very helpful about the process last week was that we ensured that every bank was able to have retention of a certain amount in the bond. We also have a non-comp entitlement in the auction process but we did everything we could possibly do to make sure that once the £7bn bond was out there it wasnt tied up in any one part of the market or in any one dealers hands. That would have been detrimental and sent a very negative signal to the market.Roberts, BAS-ML: Are you going to publish, as some issuers do, a detailed breakdown of the distribution in terms of type of investors?Stheeman, DMO: We havent done that although we did announce that 6% was taken down by overseas investors. We also said that 38% went to the GEMMs and the rest to UK end investors. Thats more information than we would normally provide at an auction. But what I can say is that having that proportion of end investor interest in the book is much larger than youd ever get in an auction, which is one positive feature of the distribution.
One possible problem arising from having given out information about the split between the GEMMs and the end investors is that while the distribution to end investors seems a lot to us, in normal syndications 38% might be interpreted as being quite a lot for the dealing community. So one thing we might consider changing in the future is the sort of information we release and the way we publish it. This is because in many cases the GEMMs had client orders they were committed to and which they needed to fill with their allocation. So the portion ultimately allocated to the GEMMs was probably a lot less than the figure that we published.Bye, HSBC:When a government syndicates a trade it wouldnt typically provide details of the split between the primary dealers and end investors. The share allocated to primary dealers would generally fall within that bank categorisation. And 38% going to banks isnt a problem at all.Mountain, JPMorgan:A few people reacted quite negatively at first to that information which came out relatively late on the day of the syndication. The day after the syndication the bond traded poorly in the secondary market. It was issued at 11 over and the following day it traded in the secondary market at 12½ over, because people looking at the distribution numbers assumed that the Street had been left with a large amount of bonds and therefore marked the price down. Perhaps some more rational thought went into it afterwards. Clarke, RBS: Youre not necessarily comparing apples with apples when youre looking at the Gilt market relative to the Italian or Spanish or French syndication. It is one of the idiosyncrasies of the Gilts market that you have large amount of index orders, while the participation of GEMMs in the overall market is also high compared to Europe.Mountain, JPMorgan: I agree. Im just saying that some people interpreted the information about the distribution more negatively than they should have done. That is why the bond traded quite poorly for a brief period of time the following day.Bye, HSBC: That was useful in a way because it showed how effectively the price discovery system had worked during the bookbuilding. We had gone from a 12bp-15bp range and then we tightened to 11bp-12bp before ultimately pricing at 11bp. And when we went down to that final 11bp number some although not many investors dropped out of the book. When the bond cheapened up in the secondary market a number of those accounts that were squeezed out of the primary book took the opportunity to take down the paper they were looking for originally. Hence, the secondary market readjusted quite quickly as bonds were redistributed from GEMMs to real money. If we were still trading at 12½ over today, then wed say we had a problem with the original pricing. But weve tightened back to 10 over, which is a perfect performance on a £7bn trade.Clarke, RBS:Arguably it would have been much worse if there had been no paper available at all and it had been completely illiquid. Maybe in the 2055 there wasnt much liquidity straight after the transaction but I think weve given liquidity to the deal without sacrificing performance which is a very difficult balancing act to achieve.Stheeman, DMO:My concern when we report that 38% went to GEMMs is that some people equate GEMMs with the Street. But I think that was a misconception. These were not loose bonds. It was simply a question of finalising the allotment process as the bonds allocated to GEMMs were then fed through to meet investor orders.
EUROWEEK: Was the distribution of 94% domestic and just 6% overseas in line with your expectations, given the maturity which obviously catered to the UK institutional demand? Stheeman, DMO: I think so. The long end of the market is not necessarily the preserve of overseas investors. Having said that, there were quite a few interesting overseas names in the book although they may not have amounted to much in terms of their overall size.Todd. Goldman Sachs: As you say, its a question of maturity. If you had syndicated a five year you would have got a vastly different distribution of investors. I think it really reflected the fact that the curve beyond 10 years tends to be dominated by the domestic investor base.Roberts, BAS-ML: The 6% slightly surprises me. True, we know that overseas interest is concentrated at the front end of the curve. But we also know that overseas central banks generally have an ability to go down the yield curve and buy long dated paper, as they have in the US market. And we also know that there are many non-central bank investors such as those in Japan that have long term liabilities. So although Im not saying Id expect a lot of the deal to be taken up by overseas investors, 6% sounds relatively low especially given that about 30% of the entire Gilt market is owned by non-UK investors. Bye, HSBC:Yes. But as Robert says there was so much more granularity in the overseas book compared to the UK demand which was made up of some quite chunky tickets. We had 130 accounts in the overall book and I think over 40 of those were overseas investors. So it was more like 30% international placement in terms of number of orders.Allwright, Threadneedle:What is interesting to us is how people play in the auctions process. Some always buy in the pre-market while others will always prefer to wait and buy at the DMO. Did you see much difference in behaviour among your UK investors?Stheeman, DMO: This is one of the areas where we suffer from an information deficit. When we run an auction, the honest answer is that we dont know enough about end investors behaviour. This is because such a vast amount is taken down by the GEMMs, and the number of direct bids from clients tends to be much smaller than the 60% we saw in last weeks issue. The people who know more are the GEMMs themselves because they are the ones who are in a position to see who is buying at 10.15am or 10.45am on auction day and whats going on around the auction. We pick that up anecdotally. However, unless we actually see the order coming through from the auction close itself, its very hard for us to say.
In relation to the figure for overseas allocation, Id add that the whole purpose of introducing syndication was to enable us to continue issuing record absolute amounts in longs and also in linkers. This is primarily for the benefit of the UK domestic investor base, notably the pension fund industry. So Im not hugely surprised by the overseas interest. We had not set ourselves any target for overseas participation in the book.Todd, Goldman Sachs:Even if overseas investors werent going to participate in the deal because their preference is at the short end, a useful part of the exercise was getting the UK story out into the market. We had a lot of discussions with customers who are investors in the UK about what our views were and so on. So in terms of publicity for the UK, it was beneficial to have that discussion even with investors who chose not to participate in the deal itself. Roberts, BAS-ML:Especially since the last time the DMO was covered so much in the press was after the failed auction. That created intense media coverage because it started the whole discussion about the buyers strike.
EUROWEEK: How did the DMO deal with that barrage of poor publicity in March? Stheeman, DMO: I cant pretend it was just another day at the office. But quite genuinely weve had uncovered auctions in the past, most recently in 2002 when we held 13 auctions in the year and had a borrowing requirement of £26bn. Did that mean that in 2002 we were facing a buyers strike? Of course we werent. Thats why I think all this talk is slightly silly.
In terms of how we dealt with it, we have a very clear procedure which is that we take the unsold portion of any uncovered auction on to our books which we subsequently offer to investors by tender. The irony in March was that we ended up getting a much better price and a lower yield for those bonds than we would have done at the auction.Allwright, Threadneedle: Long yields that day actually closed lower. So it was a great buying opportunity. Stheeman, DMO:Thats right. But the other point to make about this is that Ive stuck rigidly to the words, uncovered auction, rather than failed as some of the press uses, because it wasnt a failed auction we raised £1.49bn. It was an uncovered auction. A failed auction is presumably one where you get no bids at all. Roberts, BAS-ML:If you use the criteria that some of the press uses, one in three German auctions would be classified as having failed. Stheeman, DMO:Yes. And selling 40 year nominal Gilts at a yield of about 4.5% by historical standards to me suggests anything other than a failure. Todd, Goldman Sachs:There was some chat in the market around that time that a number of the banks were balance sheet-constrained and there had also been a lot of volatility in the early part of that week. Volatility tends not to be particularly good for outcomes in the market anyway and leads to people shying away from participating. So there were circumstances specific to that week which we probably wont see again. Stheeman, DMO: And as I said before, the tone was all wrong in March. Since then the back-up in yields has enticed a few more people into the market, but at the time yields were incredibly low. At the end of February we saw yields on 10 year Gilts drop briefly below 3%, which was the lowest since 1895. For people to think twice at those levels isnt very surprising to me.
EUROWEEK: How do you think the success of the deal would have been interpreted by Continental European sovereign borrowers? Clarke, RBS:It probably just raises expectations in terms of what they can achieve in their own syndications. There is a longer tradition of doing syndications in most of the Continental European markets. I think you can sell the benefits of it particularly to any borrower wanting to do longer dated issuance, as France did today [Tuesday]. So the arguments are in tact. It proves that syndication works.
EUROWEEK: Could the transaction have been increased to more than £7bn?Bye, HSBC:The target size of a Gilt auction is £2bn-£2.5bn, so it had to be significantly bigger to justify syndication. If you double that you get £4bn-£5bn. It would have been a difficult balancing act to justify printing more than £7bn, but with a £15bn book we were still cutting people back by 50%.
Then again, you cant argue that £7bn isnt going to be liquid. There is only £25bn in this programme and there will be more wood to chop in other parts of the curve. You dont want to use all your firepower in one go. So I think we got it right. I didnt hear any complaints about £7bn being too big.Stheeman, DMO: The bigger the deal, the more wed be taking out of the rest of the programme, which is an important issue. We didnt explore increasing the size in too much depth, but I think that had we gone further it could have reduced the value for money we extracted from the deal. Had we wanted to do £8bn, which was not the case, to have priced the issue at 12bp would not have been a trade-off we would have been comfortable with. We talk a lot about taxpayers value for money because it is fundamental. So I think the deal was exactly where it should have been both in terms of size and price. Clarke, RBS: I think thats right. If there had been any under-performance on an £8bn deal it would have been easy to point to the size as the reason. Whatever may have happened the day afterwards, nobody said we shouldnt have done £7bn.
EUROWEEK: Can we move on to what lies ahead in the programme? Stheeman, DMO: As someone reminded me the other day, however big £7bn may sound, its still only 3% of this years borrowing requirement.
Weve said that in the second quarter of the financial year we will be looking to syndicate an inflation-linked bond, probably in the second half of July, in the 30-40 year area. Thats next on the agenda. So far we have not disclosed anything more than that.
The third syndication weve referred to will also be an inflation-linked issue to be held in September. Anything more than that we have deliberately not announced. One of the issues weve always found with our issuance calendar is that we make decisions ahead of the quarter, sometimes very far in advance. And sometimes as the weeks and months move on those decisions can start to look a tiny bit stale or out of date. To me that is the price we pay for being predictable and transparent. One of the benefits to be derived from syndication is to be able to step back half way through the quarter and say, lets identify where the demand lies, so we can make sure were aligning our supply with demand at the time.
EUROWEEK: What are investors telling the banks about where the demand is today? Todd, Goldman Sachs: At a big picture level theres a lot of talk about this record amount of supply. But were also seeing record levels of demand, because we have more investors in Gilts than weve ever had before. One of the fall-outs of the credit crunch has been greater risk aversion on the part of investors which is a good thing for Gilts.
Demand is also being driven by regulatory reasons, with the FSA potentially requiring that banks will be required to hold more Gilts on their balance sheets. We have also seen a continuing trend of pension funds seeking to hold Gilts as opposed to credit products or equities, or holding Gilts instead of receiving swaps, and all this is good for Gilt products, so its not just a one-sided phenomenon. There has been a very healthy increase in demand as well.Roberts, BAS-ML:From our perspective, all the key metrics have been covered. Weve had the announcement of new financial regulations that will increase banks holdings of Gilts, which we estimate amounts to about £200bn globally, just less than half of which is in the form of UK buying. That has spread along the curve and has become very much a long-end influence which I think has blindsided a lot of people who thought it would be visible purely at the short end of the curve.
At the same time we still have the ongoing pension problems which continue to be a very large-scale news story. Even yesterday, news stories were being reported about finance directors being upset about high equity weightings. Now that were seeing pension funds being far more educated about the use of repos, there has been much more of a move away from using swaps and towards the use of government bonds. In fact I would say the biggest move in the UK market in the last month and a half, without question, has been the shift from swaps to governments that has taken place. That has been a big influence on Gilts, which are trading cheap against Libor at the long end of the curve, but they have richened up on a relative basis versus swaps.
Against that backdrop the demand equation is so much more important than any worries about supply, which isnt really that excessive.Allwright, Threadneedle:Id agree that thats caused a big shift in liquidity out of the swap market into the physical Gilts market, which has to be good. In the past the swaps market has been used as an instrument of leverage with nothing underlying it, which was a case of the tail wagging the dog, but now it is clear that the Gilt market is driving interest rates across the curve.
Another important point is that the programme is helping to build up the delivery basket for the Gilt future. Having been a market maker in the past and being an end investor now, I am very aware that having a hedging tool with at least some semblance of liquidity has been very important. Weve always looked at the swap market as not being an appropriate area for the end investor to play in, because liquidity there can be ephemeral.Roberts, BAS-ML:The Trésor said in its documentation for its 30 year syndicated deal that 15 year-plus demand has been effected by investors shifting out of swaps. So even though Europe doesnt have the same proposed banking regulation as the UK does on liquidity management, the same discussion is taking place. So its not just a UK phenomenon.
EUROWEEK: Im surprised to hear people say that supply is not excessive. Do others agree? Bye, HSBC:I dont think its excessive in the sterling market, because Robert has the benefit there of having the market almost entirely to himself. If youre a European sovereign in the EU zone, you potentially face a number of problems relating to supply. For example, you may have another 10 issuers coming into the market on top of your trades wanting to attract the same investors. And then youve got the relative value problems arising from the risk that if one of those issuers is downgraded that will have a disproportionate effect on your funding. But if the UK were to be downgraded, does it matter? Probably not. Because who else is going to provide the benchmark securities in sterling? Its not going to be Network Rail as it would also be downgraded in line with Gilts, and its not going to be issuers like EIB or KfW which dont issue sterling in the necessary volumes to compete with Gilts for benchmark status. So the DMO will always be the de facto benchmark issuer in sterling regardless of its rating.
So I dont believe the rating matters too much, although I think the level and volatility of the currency is important because that will clearly have an impact on the level of overseas demand.Clarke, RBS:The UKs regulatory framework in terms of ALM matching means that more so than in any other market, with the possible exception of the Netherlands, the problem is one of lack of supply. We probably see more of this shortage in the index-linked market than in the conventional market. So one question going forward will be how do we address the shortage of index-linked issues, not only in terms of creating supply but doing so in such a way that encourages people to participate in the auctions? Latent demand has always been there, but it hasnt always had an efficient avenue through which to participate.
EUROWEEK: So where have sterling investors been going for their index-linked exposure?Clarke, RBS:Thats the question. Theyve been using swaps in many cases. There has also been a complication in terms of timing, because they have not always been able to align all their interests at short notice in an auction, which is why you see such lumpy tickets sometimes coming into the secondary market. What the syndication process allows them to do is effectively pool together all that interest and express it at the primary rather than the secondary level. Mountain, JPMorgan: There is certainly a difference between the nominal and the index-linked market in terms of auctions. In the index-linked market you always get much higher end investor participation through direct bidding in the auction than you do in nominals. This is because the end investors know that if they want to buy a big ticket, most of the time they cant do it from a market maker just with a phone call, as market liquidity is not up to the task.
So it might be interesting to see if there is less of a difference in the syndication process for the index-linked market than for the nominal market.
Because supply will be via syndication for the index-linked market in the second half of July, when demand is often high, investors will clearly be eager to participate to satisfy their usual requirements. That dictates that they will certainly be big participants in the 30-40 year linker auction.
Syndication is definitely a welcome development for the linker market as well as the nominal market, but I suspect that its going to be less of a radical development in terms of investor participation than it was for the nominal market.Stheeman, DMO:I think youre right. I also dont think were going to be doing £7bn transactions in the linker market.
EUROWEEK: What will your size targets be in the linker market? Stheeman, DMO: At this stage were not saying anything explicitly. Again, were saying that we would like to do something larger than what we would normally do via auction. But for neither the conventional nor the index-linked market are we aiming for a specific size. We need to feel our way forward.Mountain, JPMorgan: If I were to take a guess at what the market is thinking at the moment, it is probably that a new 30 or 35 year linker would be roughly £1bn if it were sold by auction. So the market would assume that the initially indicated size might be £1.5bn-£2.5bn, which would be a similar ratio to what was done with the 25 year deal last week. Then it will be a question of how much demand shows up. I think rather like in the nominal issue where initial guidance was £3bn-£5bn, but £7bn could be issued without seriously effecting the secondary market, people will assume that the DMO will be able to do £3bn or £3.5bn if that is the level of demand. Todd, Goldman Sachs:The beauty of the syndication process is that you can tailor the size depending on the demand. Mountain, JPMorgan: Yes. The single greatest benefit of it is that you dont have to pre-commit to selling too much in the way that you pre-commit in an auction, with the risk that if its not covered on the day there will be questions in the House and in the press. Syndication removes that problem.Roberts, BAS-ML:That is especially important because within the index-linked market the longest part of the market faces such a supply shortage. Supply shortage rather than anything else is what we should be discussing in the market for index-linked Gilts. The relative increase in index-linked Gilts is much smaller, year-on-year, than the increase in the nominal market. Stheeman, DMO: This year the proportion of linkers is 14%, which by chance is the same as it was last year. However, in 2007-08 the linker proportion was 26%, so it is right to say that the relative increase is smaller. Nevertheless, the planned absolute amount for 2009-10 is at a record of £30bn. It should also be noted that the issuance of nominal bonds is also at a record high. Just a few years ago, £30bn of issuance in the linker market would have raised quite a few eyebrows.Roberts, BAS-ML: But its only £10bn more than it has been over the last few years, which as a share of the £1tr of private sector pension liabilities is only 1%. Stheeman, DMO: Its actually £9bn more than 2008-09. But just two years ago, in 2007-08, the total was £15bn. So it has doubled within two years. And if we go back to 2005-06, inflation-linked issuance was only £10.8bn. Roberts, BAS-ML: Yes. But my point is that only equates to 1.5% more of pension liabilities being hedged than before.Allwright, Threadneedle:That assumes that every pension fund liability is going to be hedged. But Im concerned that there may be a degree of complacency about the UKs rating. Yes, there is very high demand now so the market can take lots of supply. But that cant go on ad infinitum. At some stage QE has to be taken off the table. What happens when every final salary scheme is shut down and when the public sector pension regime is changed or re-jigged? That final demand will then diminish as well. So this is a good time to issue as well as to buy a large amount of Gilts; it is a good time to buy a lot. It is also a good time to be in the syndicates of investment banks. But this isnt just a one-way bet and I dont think you can underestimate the importance of a triple-A rating. It is important not just for the DMO as a borrower in the market but for the whole structure of the market, as well as for confidence in the market.
The DMO has done a very good job of managing a highly transparent borrowing programme. But it is at the beck and call of its political masters. And if they suddenly say, we want to borrow another £100bn next week, then that is what you will have to do.Stheeman, DMO: True. Were debt managers. We dont set fiscal policy.
EUROWEEK: Is the figure of £220bn of planned Gilt sales in 2008-09 set in stone? Stheeman, DMO: The figure represents our financing remit at this current time. Traditionally if the remit does get tweaked that can only happen at the time of the Pre-Budget Report. Last year was an exception in that there was a one-off remit revision in October as a result of the bank recapitalisation exercise. And it was revised upwards again at the time of the Pre-Budget Report. The remit does allow for exceptional revisions if necessary, but they only happen very rarely.
At the moment £220bn is the figure we are working with, and we are doing so on an even-flow basis, to ensure that everybody can see how far we have come in the programme.
EUROWEEK: But from what people around the table are saying about a shortage of supply in some parts of the market, a minor tweak may not be that much of a problem.Allwright, Threadneedle: It might not be a problem. Bank recapitalisation was a huge amount and I think led to what was almost a delayed reaction in the market. I dont think anyone is suggesting that we are now at the tipping point that would push us into third world debt management policies, but if borrowing keeps going up we are going to get closer to it. This is why I think investor confidence is so important. Stheeman, DMO:I entirely agree with your point about investor confidence. That is indeed absolutely crucial, and it is one of the reasons why the syndication exercise was so useful. It did send an extremely strong signal to the market in every respect and that was an important achievement in its own right, quite separate from what we achieved in terms of the financing.
My feeling about the borrowing requirement and how the market takes it down is that although it may sound axiomatic, government bond markets in OECD countries tend to be what I would describe as elastic. They do expand to take down the supply and they contract when that supply begins to decline. And at some point that supply will start to decline.
We are often asked if there is a tipping point, and if so, what that tipping point is. The honest answer is, I have no idea. Logically you have to assume that there must be a limit to what the market can take, but it is not clear what that limit is.
A couple of years ago, people would have said that its simply not possible to issue £100bn or even £200bn. But its not just a question of the amount were issuing: its also a question of what is happening with monetary policy, how much liquidity is in the system, what is the overall demand for risk-free assets? All of these questions need to come into the equation at the same time.Roberts, BAS-ML:Thats what is so worrying. Right now, we know were seeing the worlds biggest ever increase in structural savings, including in the UK where the savings rate is going up dramatically. We have banks reallocation into government product and we have the ongoing pension fund story. This is all great. But well still have a 12% deficit to worry about next year. And then what happens the year after that, and the year after that, when the deficit will remain substantial?Allwright, Threadneedle:I dont think the volatility in the Gilt market is by supply and demand dynamics. Its caused by the Budget. Stheeman, DMO: I agree. And that suggests to me that it is fiscal policy that dictates what we do. Ultimately we as debt managers have to fund the deficit, whatever that may be. But how that deficit looks in the future is not a function of debt management policy. It is clearly a function of fiscal policy. Roberts, BAS-ML:It makes you wonder if the increased chances of an October election, and therefore of a retrospective budget before the end of this year is one of the reasons why Gilts have calmed down. Mountain, JPMorgan: If we had an early election that brought the Conservatives to power with a clear commitment to fiscal discipline, that could lead to quite a substantial repricing of the Gilt curve.
EUROWEEK: Will the balance of roughly 95% going to the UK and the remaining 5% overseas be maintained? Presumably it will, given that the DMO will be issuing linkers at the long end of the curve?Stheeman, DMO: I suspect that for linkers it will be even more domestic-focused. Mountain, JPMorgan:In the case of the long linkers it will probably be close to 100% UK placement. And within that, 20 or so investors will be dominant.
EUROWEEK: Its an unusual luxury, isnt it, to be able to depend so heavily on what is effectively a captive market of investors? That is quite different from the US and Europe. Stheeman, DMO:Perhaps. The very high overseas interest in the US market reflects the reserve currency status of the dollar. But ultimately whoever pays the best price gets the bonds. In the UK it just so happens that especially for long-dated inflation-linked Gilts, it is the domestic UK investor base that gets the bonds. Were perfectly comfortable with that. Bye, HSBC:Francis said earlier that if we syndicated a five year deal we would have a lot more international demand, which is definitely true. We saw Belgium this week syndicating a three year euro deal which is quite rare. Going forward, will you continue to focus on the long end for your syndicated issues? Or will you keep an open mind?Stheeman, DMO:This goes back to the question of what is the purpose of syndication. One of the motives is to take the operational pressure off the sheer number of auctions that were doing and the market events were creating. The advantage of auctioning shorter dated paper is that we can increase the sizes to £5bn and arguably even more.
You cant do that in longs, nor in linkers. Or if you did youd be paying a very high price. So there probably isnt a requirement for us to syndicate anything much shorter.
The purpose of syndication is two-fold. One is to align our supply much more closely to demand. The other is to keep a bit of a lid on the amount of market events that we are creating. Each time we conduct an auction we go to the market and we say, we want you to set yourself up for this auction. That requires that the Street set up short positions which it can do relatively frequently. I dont, however, think we want to have to ask the Street to do that too often as we wish to avoid any potential danger of auction fatigue. And that would not be as a result of supply but as result of what the Street has to do from an operational standpoint at the time of each auction.Todd, Goldman Sachs: Kevin made the suggestion that only about 20 or so large investors control a significant share of demand for linkers at the long end. Their activities by their nature tend to be lumpy, which contributes some volatility. At the short end of the nominal curve, however, you have a much bigger group of disparate investors and although they may look to do large trades in notional size, because there are so many of them the impact of those trades in basis point terms tends to be lower. That is why you probably wouldnt need to do shorter dated syndicated bonds but why syndication is so well suited to the long end and the linker market. In those markets, syndication means you can ensure that the supply corresponds to the lumpiness of demand among those smaller pools of investors.
EUROWEEK: So just to be clear, the DMO wouldnt foresee syndicating anything shorter than, say, 20 years?Stheeman, DMO: Certainly not at the moment. Weve been very clear that the purpose of this is to supply the market with what it wants, in particular with long-dated conventional bonds and linkers.
EUROWEEK: Would it be desirable to try to attract more foreign participation? Stheeman, DMO:As a debt manager were charged with the task of minimising the cost of government borrowing over the long term, taking risk into account. However much we acknowledge the importance to us of the domestic investor base, we are completely neutral at the end of the day as to who will buy these bonds because we want to get the best price. So if, for instance, we saw a big shift in demand to a given investor base prepared to pay the right price, so be it. But if the domestic investor base in the UK felt it was being left out, the market being the market would mean that it would end up paying a higher price itself to make sure it could secure the bonds. In other words, in the context of a large, liquid and efficient market we couldnt even hope to influence distribution in that sense. Nor would it be desirable for us to try to influence how the share divides up between domestic and overseas placement.
Of interest is the fact that foreign ownership of the Gilt market continues to grow. Six or seven years ago the overseas share was 16% or 17%. Its now over 30%. That is entirely down to the fact that overseas investors have found it possible to make more purchases of Gilts at the maturities in which they normally prefer to buy, which are shorts and mediums. Its not because the domestic investor base is reducing its participation.Mountain, JP Morgan:But the DMO can indirectly influence who buys what by deciding which parts of the curve to issue in. If you go back three or four years, the long end of the curve was yielding much less than the 10 year point, and the DMO skewed its issuance towards that part of the curve. Now that the curve has changed its shape very radically, for reasons we all know, the DMO has shifted its issuance into a part of the market which is more cost-effective for it. But also, its a part of the market that overseas investors feel much more comfortable about buying. That is why the proportion of bonds they hold has risen substantially. Price in the end influences who buys and where they buy.
A few months ago there were some domestic investors who were saying they didnt think the yield on long-dated government bonds was high enough, so they werent going to buy them, and the curve went extremely positive as a result. Then the remit came out and there wasnt quite the degree of skew towards the long end that we had previously. The result was that that part of the curve flattened with domestic investors deciding that these yields werent too bad after all.
EUROWEEK: There is no question, then, of the UK issuing a euro or dollar benchmark? Stheeman, DMO: The current policy remains that the UK only issues foreign currency debt to finance its reserves. If we had a need or desire to raise funds, its not clear to me what the benefits would be of going to another currency. If we can do a £7bn issue in 25 year sterling, why would we want to do something potentially a lot shorter and smaller in another currency?
EUROWEEK: It would only work if you could swap back at extraordinarily compelling economics, which you almost certainly wouldnt be able to do in the sterling swap market? Stheeman, DMO: Exactly.
EUROWEEK: Peter raised the issue of complacency over the rating. Can you, should you, and would you prepare for issuing post any downgrade that may come from the agencies? After all, S&P has indicated that possible future debt levels of the UK of more than 100% of GDP would not be compatible with a triple-A rating. Stheeman, DMO: No. Ultimately we have no way of knowing what the rating agencies are going to do. Its as simple as that. So for us to try to anticipate that in our funding programme would be seen as opportunistic.
But a point I made at this weeks conference is that the news that we had been put on negative outlook was released at about 9.15am on the morning of the largest auction that we had ever done, which was a £5bn five year auction. As it turned out, the auction went fine, and at £13bn the number of bids we received at that auction was the highest we had ever received for any auction. Im not saying that the rating doesnt matter. That would send the wrong signal. But I dont see any need for us proactively to try to anticipate what the rating agencies may or may not do. That would be highly detrimental to the whole predictability of the programme.
Also, how do you ultimately judge if a triple-A rating is a more accurate reflection than a double-A rating? I dont know how you measure that. It doesnt tell me anything more than that it is the opinion expressed by a particular rating agency at that particular time. Other agencies may have very different views.Mountain, JPMorgan:To try to prepare for a downgrade would inevitably be enormously counterproductive. Preparing presumably means youd issue twice what you would have issued normally, and what sort of a message would that send to the market? Stheeman, DMO: The markets reaction to the S&P announcement was relatively muted, and one of the reasons for that is everything was known to the market. Our fiscal position didnt change because of anything S&P had said. That position had been set out at the time of the budget.
Additionally and this isnt a criticism, it is purely an observation rating agencies are lagging rather than leading indicators. They will report on a situation which in their view has already arisen and they will comment on that situation. But theyre not always very good at anticipating what might happen.Clarke, RBS: The credibility of the rating agencies has been seriously diluted of late. They have been behind the curve on several occasions and dont necessarily have any extra information to analyse the economy. If the threat of a downgrade had happened four years ago I think it would have been material. Today it isnt.Allwright, Threadneedle: Whats more worrying than complacency about the rating is complacency about the events that would lead to a downgrade. Its complacency about the UKs fiscal position and about the possibility of us borrowing as much as we like for the next few years. Because the agencies are a lagging indicator by the time we got to the point of a downgrade it would have been fully discounted by the market. Mountain, JPMorgan:I dont think the market is being complacent. If you look at implied inflation at the long end of the curve it is extremely high which suggests that the market does not believe that the Bank of England can deliver on its inflation target. That doesnt sound to me like a bond market that is complacent about the future. The same is true of nominal forward rates which are much higher than the long term average. These metrics are telling us that the market is being priced quite cautiously.
EUROWEEK: Would others around this table share the view that PJ raised, which is that in the absence of much competition in the sterling market the rating may not matter? Todd, Goldman Sachs:Besides creditworthiness, one of the things investors value especially highly is liquidity. Thats why Gilts trade at a premium to other similarly rated issues in sterling. I would expect that additional liquidity to be valued whatever happens to the rating. Roberts, BAS-ML:I think as an issuer the first thing you have to worry about is the credit rating which would matter more in the euro block, because you run the risk of assets shifting to another country in the same currency. The second is if you have massive net external liabilities that need refinancing. The UK does not have either of those problems, as many of the central and eastern European countries do, for example.Allwright, Threadneedle:I think the issue of the rating is more a question of confidence in the market. A rating is a function of the fiscal position at the time which will naturally have an impact on rates. A downgrade would not effect liquidity but it would effect rates. But when we talk about the rating and QE these arent really questions for the DMO; theyre questions for the government and the Treasury.
EUROWEEK: How will the QE endgame play out, and what impact will it have on the Gilts market?Mountain, JPMorgan: We havent really touched on the interaction between QE and Gilts, but it is clearly enormously important and at the margin QE must have been keeping yields lower than they might otherwise be. A lot of people seem to be coming round to the view now that QE will finish in August but the question of how the DMO can prepare for this is like the question of whether it should prepare for a downgrade from S&P. The DMO shouldnt prepare for the end of QE because it cant. And because as has been said from the outset there needs to be minimum interference with the issuing programme.
Theres a very commonly held view that the QE programme is inherently inflationary which in a sense is true. The authorities implemented QE because they want to avoid a scenario of deflation. But the question is, will QE lead to a serious inflation problem, and that is one reason why the risk premium attached to linkers has gone up quite substantially. Weve all been obsessed about how bond A has been trading versus bond B but in a strategic sense QE has raised the price people are prepared to pay for linkers and it will be interesting to see once QE is out of the way how the linker market reacts.Todd, Goldman Sachs: On a related topic, Im not sure its right to say there has been a concentration of issuance in the five to 25 year sector. That accounts for two-thirds of the existing yield curve so it was inevitable that there was going to be a significant amount of issuance in that sector but I dont think it is necessarily concentrated in that area of the curve. Roberts, BAS-ML: The proportion of increase in shorts and mediums has been much higher than it has been in longs and linkers. Todd, Goldman Sachs: I guess thats because the mediums were the smallest component in previous years so the sector offered more scope to increase. Stheeman, DMO: In the latter half of last year we sold more shorts, and in the early part of this year the area that has seen the biggest increase in supply is mediums. But that has nothing to do with QE and everything to do with the sheer scale of the financing requirement and which part of the market we need to go to if were going to be able to access the funds and the liquidity to fund ourselves on such a large scale. Dont forget that QE was only introduced at the beginning of March, by which time we had already had five months of significantly increased issuance of shorts. So were not drawing up our issuance plans with one eye on what the Bank of England is doing with QE.
But the argument that we cant avoid issuing in the five to 25 year area is valid, given the size of our funding programme. We need to issue across the curve and to do so in as orderly a fashion as possible. If thats pushing us towards shorts and mediums its because of the nature of the market at areas that are both deep and liquid, not QE.Todd, Goldman Sachs:One indication that the five to 25 year range is not necessarily being explicitly targeted was that last weeks new 25 year deal had a September maturity. If it had had a March or June maturity that would have brought it within the 25 year basket and therefore exposed the DMO to the criticism that it was monetising its debt. But the maturity chosen directly counters any such criticism. Allwright, Threadneedle:Just coming back to the QE roadmap, we can only get to the end of QE when we have a roadmap for the introduction of some kind of fiscal discipline. I dont think we can have one without the other, because when people anticipate the end of QE we will have a rapidly steepening yield curve. Mountain, JPMorgan: I think we need to draw a distinction between the end of QE and the reversal of QE. It seems not implausible that the Bank of England could stop buying in August. But what does seem implausible, although some people in the market seem to be suggesting it, is for the Bank to suddenly turn round and start selling the bonds back to the market. Allwright, Threadneedle: The only modern precedent we have for QE is Japan, and there they just seem to keep issuing more debt alongside QE. Its not quite one for one but it does have ratings implications and it can hit confidence in the efficient functioning of the market. Roberts, BAS-ML:QE in the UK is a massive programme. Its already 9% of GDP and the Bank still has its foot on the accelerator. Its worked in that real yields have massively outperformed the US market but it hasnt worked in the sense that there is no evidence of any credit pick-up yet. So the Bank is in a very tough position. Should it keep its foot on the accelerator when it doesnt know whether it will be working six months down the line?