Sovereign Bond Markets Roundtable 2008: part one
IFR: What has happened over the last six or nine months in terms of the way sovereigns have approached the market and the differing levels which have been involved in relation to spreads to Bunds and to Libor? And are there any differences in the way you approach the market to achieve your funding objectives as a result of these changes?
Anne Leclercq (Belgian Debt Agency): Yes and no. We have to achieve our funding objectives, but we have to be more flexible and do things differently to an extent. We have to be more flexible in terms of sizes: we cannot always do the size that we want to do, but if the market does not take the size we want we will reduce the size and go on to next time, eventually, for a bigger size. Also with regard to maturities – and I am speaking more precisely about the auctions – we always listen to our primary dealers to know what kind of maturity would be suitable for them, and what kind of maturities they would like to have us issuing. For example, in an auction which we did last September, we listened even more carefully to our primary dealers. Some of them seemed to have some positions in a non-benchmark issue and wanted to close those positions, so we issued a non-benchmark size, which is not really the normal way of doing it. So through flexibility we have tried to help them out to make sure that they were able to close their existing short positions. I would say there is not so much difference in terms of the sizes, but we have to be flexible and listen more to the markets now than we used to.
IFR: I suppose the same will be true of all the sovereigns represented. At the end of the day, you will borrow whatever is required but there are obviously different approaches represented here between syndication and the Dutch direct auction (DDA) process, for example.
Peter Nijsse (DTSA): Yes, I think the direct auction is a very good technique for new issuance in the current turbulent markets, where we see the end investors able to be satisfied directly by the auction process. We have also seen a large share of real money investors, so while trading books may have diminished in their significance, end investors have been looking for a safe haven and have come into the DDA process as a result of this.
IFR: The evolution of the direct auction process has obviously been a separate development from the syndicated route used by some other sovereigns. Is it not one that the DTSA is contemplating going down?
Nijsse: We are quite happy with the way the DDA has worked out. I'm not quite sure whether it is so different in the end from the syndicated route. The main difference that we see is in the end we act as the bookrunner, with end investors able to deal directly with us. They don't have to show business to others, only the banks with which they registering their bids know what they are bidding. So we feel it is a method to reach out directly to the end investor, which benefits the banks, and the trading books are maybe less eager to buy. In terms of our programme we haven't seen too much difficulty in conducting our issuance programme. We've done it more or less as we intended and as we announced in our outlook in December, so we have been able to issue the size that we wanted to issue. Of course there has been some volatility in the markets, so maybe there are some weeks which are better than others. But we feel that by being predictable in our auction approach, we remove the element of surprise from the auction process.
IFR: It seems to be that predictability is the essence, to an extent, of sovereign borrowing programmes by knowing what's occurring, although there remains an element of surprise. This week Italy has done a three year US dollar deal. From a trading point of view is there any kind of difference in the secondary liquidity of deals which, perhaps by virtue of syndication, have been placed with long term buy and hold accounts?
Christophe Rivoire (HSBC): It is clear that, compared to a year ago, the market has changed dramatically in terms of liquidity. The risk appetite of a lot of market participants – not only the banks, but hedge funds and even the sovereigns – has been impacted, albeit less than other asset classes, due to the quality of the assets. It is clear that the liquidity has changed and that the debts of all the sovereign issuers have changed, and are still changing in a sense. This means that to use an auction process, a syndication process, or a DDA process as in the Netherlands, the buyers are going to be different.
What is clear is that when you issue by auction, in 80 % or 90 % of the cases the dealers are going to take the position. They are also going to keep it for a certain period of time. Due to the fact that the dealers are in between the issuers and the investors, their target is obviously to sell it to the investors, but they are limited to satisfying the demand when the demand comes. When issues are syndicated, or sold through the DDA process, it is clear that the issuers are going to reach for end customer demand immediately, which means that for banks, they are going to reduce the risk they have on their own books. Due to the fact that the market is now less liquid, the banks are more concerned by this fact. It is not only the terms of the liquidity of the debt, but also the liquidity in terms of refinancing these positions.
All positions have a cost in terms of liquidity, and any position will have a cost in terms of risk and in terms of value at risk. I think it is important to keep in mind the fact that, due to the explosion of the volatility we have seen in the last 12 months, all the value at risk with the same risk has more than doubled. It has meant that for banks to keep the same position as a year ago, it is definitely more costly in terms of value at risk, and the capital needed to cover the position in terms of liquidity, and liquidity in terms of refinancing.
Refinancing is the key point. In this market, and as long as there is some tension in the money market, refinancing is going to be the key element for these two reasons. Syndication and auctions are indeed different. Obviously there are some pros and cons to both in terms of efficiency, but a mixture is probably a good compromise, as is the case with a lot of issuers now.
IFR: Because of the heightened level of money market rates and general risk aversion resulting in a reluctance to run long term inventory, do you think that there's perhaps a tendency to try and sell positions more quickly through the secondary market, where sales haven't been made through a syndicated process?
Rivoire: I think we can say that for the banks, obviously, the cost of the balance sheet is higher than it was and as we have said, liquidity is a bigger concern than it was. That being said, the banks are still ready to take positions onto their books. Some of them may be in a more challenging position, but on average the banks are still ready to take some positions on their books. But they won't take it due to the fact that if they do, it will be more costly in terms of value at risk, refinancing, et cetera. They want to take it at a level which may have been as attractive as it was a year ago, but it is now less attractive due to the fact that the costs associated with that are now more expensive. But I think you disagree, Erik?
Erik Wilders (DTSA): What you are saying is that you want to be back in a situation where the issuer was in a very comfortable position and the banks obviously thought that their time had come and I think the time for the investor has now come. So you may have to wait for a few years for a return to that situation in all probability.
Leclercq: Yes, but I would add that if you are talking about the time of the investor having come, you have to look to the changes that have taken place in the syndication process compared to last year; just look at the range, the spectrum of investment which we now have available. It is clear that it has increased. Although for the different types of investments which you find on offer to investors there is a lower rate of return and that, as Christophe rightly pointed out, is due to the high volatility and the high swings in the money markets. Some houses cannot put inventory on to their books anymore, certainly those who don't want to have them for a long time, and those who prefer to park them for some time, take the profit and go away. But they cannot necessarily go away, because there is not enough liquidity in the market to make sure they can offload positions at a later stage, or because the bid offer spreads are now too wide.
The second change in the syndication process is in the pricing. The pricing process has changed: we used to have a discovery process which previously relied on a smooth secondary market. It was possible to interpolate between two of the outstanding issues and look if they have the same value in benchmark terms and liquidity terms. Based on that you then had to try to find a price. But if you look through the pricing process now, it is quite different. There are different elements in the pricing process which all move in different directions.
You have the asset swap valuation on one side, as well as the curve, your own secondary curve, on the other side. They all move in different directions, because the different points in the curves are influenced by specific flows around these points. This may mean, in fact, that at a given moment in time they move in very diverse directions. So it is necessary to try to find a suitable range which is attractive enough for investors which look at asset swap valuations. In our case our bonds are based upon a bond spread reference. So it is not so easy to try to find an equilibrium between these two.
There have also been some changes in the syndication process and in getting out into the market. It is necessary to try to find the right price, so an issue doesn't look too cheap, or too expensive. Really it is important to find the right price at the moment when you come out and to make sure it doesn't crash afterwards, because that's not very good for the bond either. We have to make sure that there is sufficient appetite for a bond. So there's quite a difference from that point of view.
It can’t be that different in the DDA process. It probably is the same way of thinking which you have to go through when you start the process. So from that point of view there's some risk. We are going about it in more or less the same way with the syndicated process as in the method of issuance used in the Netherlands.
Wilders: I think the big difference is not between syndication and an auction by DDA, but between syndication or a DDA and an auction.
Wilders: Looking at the banks, I wonder whether you see a difference in the secondary market performance of bonds sold by these different methods. Is it easier to trade a bond which was syndicated or issued via DDA, or a bond that is auctioned? Is there a difference attributable to the fact that the street – including investors – has a long position? Do you see a difference?
Simon Maisey (JPMorgan): I think that one of the differences in syndication is that hopefully we can provide some more of that kind of feedback and advice before the bond gets issued, so we can help. So therefore it should trade better in the secondary market, rather than an auction which is a slightly more brutal kind of a process. You see how it goes once it is done, whereas being able to bring advice to the syndication process is one of the values that we can add in terms of what type of bond, what maturity, or what size that we think will go well, depending upon demand. So, on that basis, I think the syndication process works better: it offers a greater degree of flexibility.
Enrique Ezquerra Martin (Spanish Tesoro): But also auctions are becoming more like syndications, because, at least from our perspective, and from what I hear of auctions in other countries, there is a greater willingness to adapt ourselves and reopen off-the-run issues, for example. We also see more final investors coming into auctions, so the typical complaint – and no offence is intended here – that auctions were somewhere where primary dealers had to buy and hold their position, et cetera, is a bit outdated. We have seen a lot of direct bids at auctions, perhaps also due to the increasing presence, for example, of domestic Spanish investors, who are increasingly presenting bids enabling the reopening of old off-the-run bonds. So the small direct bid coming from final investors, which was one of the negative points previously made about the auction process – the notion that theoretically final investors could not participate – is evolving and changing.
IFR: Do you think that there is a reduced differentiation between the Dutch approach and that of Belgium which uses the syndication process?
Ezquerra Martin: To be frank, yes.
IFR: It is not very good news for bankers, I suppose, is it?
Ezquerra Martin: Well, yes and no, as usual. Syndication always gives you more flexibility. You can choose the timing of a transaction et cetera, whereas in auctions, we try to keep them as predictable as possible and so announce them on a quarterly basis to maintain that predictability element. On the other hand we value the flexibility of syndications which has its own value. Our expectation is that interest rates are going up, and we are seeing a strong domestic bid from Spanish banks, bank treasurers, insurance companies et cetera. We see that they are coming directly to the auctions recently, which is something we have not seen in previous years. Related to this is the reduction in the new issue premium at auctions.
Wilders: Has the auction new issue premium disappeared?
Ezquerra Martin: We think so.
Rivoire: Not yet fully disappeared, let's say it's reduced.
Leclercq: It's diminished.
Wilders: As long as it's there, it is obviously a hurdle for investors to participate in the auctions, or in whatever. Unless you pay the same fees to investors.
Rivoire: I agree and I think that all the banks agree on the fact that what we have seen in the last few months is more interest from final investors in the auction process. That being said, I think that maybe in the past more than 95 % was bought by the banks. Now it is probably closer to 80% or something like that. As you said, it is probably due to the fact that you and other issuers have decided to re-open some off-the-run bonds, where the market was short. It is definitely easier to attract some lead orders from accounts where there are shorts than to issue when there is no specific interest. That's a part of the flexibility already mentioned. But I think it is a very good point to highlight.
To come back to your point, Erik, in terms of the liquidity of bonds issued for the first time, obviously I think it's clear that in the last 12 months we have seen a few syndicated bonds issued on the relatively cheap side. It is clear that, for the banks, it is more challenging to offer some very good market-making practice, and especially on the offer side after pricing than it is when the bond has been issued by auction, because obviously there is still a long position in the street. A bank is in a better position to offer an efficient secondary market when they've got the bonds on their books, but particularly if the flows in the secondary market are on the buy side, because we don't want to do the opposite. If the interest is on the sell side, it is definitely easier when it is a syndicated bond than when it is an auction.
Wilders: In that situation surely being short is the preferable position?
Rivoire: I'm not sure I agree. That's the reason there has been some very good bid to cover in some Spanish issues in particular, where the market was short. That being said, I think that to achieve the funding needs you have, you clearly not only have to tap this bond, but to open some new benchmarks. In terms of transparency and liquidity, I think that all the issuers want to continue to build their curve and to increase their benchmark presence. So while it is necessary to issue some benchmarks, it is also necessary to be more flexible, for example, to reopen some old benchmarks without impacting their transparency. As I said, the credibility of the issuers must be maintained because we consider that credibility and transparency are very important features of the market – especially when the market is more challenging now than it was previously.
Leclercq: I think you are right. We should, even if the market is very difficult and even if people are asking for a little bit more in advance, be careful. We need to make sure that the bulk of our plans and our borrowing programmes are stable and predictable. Additionally, we need to ensure that the issues which we are creating are ones which are large enough to maintain an appropriate level of liquidity. This remains paramount, even if the markets are difficult, and even if sometimes it looks slightly more attractive to do something else. We are there to stay, so we should remain predictable and stable.
IFR: Have European government bond turnover figures remained consistent this year with previous years?
Angelo Proni (MTS): In our experience, and from what I understand as far as the other electronic trading platforms are concerned with respect to the outright cash market, the past two to two-and-a-half months have seen a fall in volumes. Whilst at the beginning of the year – and indeed actually in the run-up to the end of the year, despite the surge in volatility that was observed – volume still did pretty well. In fact it did much better than usual. Right now, however, the volumes are definitely weaker.
Jeffrey Hogan (BGC Partners): The voice volumes have remained very robust. From an electronic perspective, volumes have declined. As you might expect in times of high volatility, the voice market has been resurgent. Certainly since last August we have seen, as a percentage of the overall performance of the market, voice volumes enjoying a significantly greater share. I think, as we provide a service to the banks, we are fairly agnostic [about whether trades are brokered by voice or electronically]. If the banks wish to trade by voice it is fine and if they wish to trade electronically that facility remains available to them.
In Europe the market could be described as "fragmented", with multiple issuers compared to a single US Treasury issuer. Recently there have been articles in the press surrounding the ten year anniversary of the euro and the movement towards a single issuer, and we may see some discussion of the bill sector becoming unified in some fashion. But the treasury market is different because it became virtually 100 % electronic instantaneously, and there is also the existence of the high velocity, non-bank algorithmic operators in the treasury market which isn't the case with the European government bond market. This has the effect of maintaining the status quo and the reliance on intermediation by voice brokers.
Europe also differs in the method of price quotation. It operates on a decimal format, whereas US treasuries are still quoted in 32nds and 64ths, although there is certainly argument that the decimalised format is undesirable. When the New York stock exchange and other equity exchanges moved from fractions to decimals, the actual amount in terms of volume of each transaction went down. So in terms of looking at the bid offer spread, the price appeared to be narrower, but in terms of size available there was a dramatic reduction. The size per order went down due to the fact the tick size was smaller. So I think in the equity world, most of the exchanges who experimented with decimalisation have concluded that, at best, it has been a mixed experience. That was because the providers of liquidity on a five cent bid/offer spread are much more willing to show size than if there is a one tick spread. By definition, each individual transaction at a tighter spread is going to be smaller. The specialists on the floor of the New York Stock Exchange – when there was a floor – reduced their liquidity dramatically, because of the desire to make markets on a one or two tick spread, which is clearly not the same as a wider spread. Certainly we have seen some of that in the European cash markets. In the Government markets, where we have all witnessed the changes in tradeable bid/offer spreads since last August, that parallel is exactly the same. The desire to maintain similarly sized markets when the underlying conditions have diminished is a challenge for the industry.
IFR: It was said that decimalisation had resulted in an increase in algorithmic trading in the equity market, was it not?
Hogan: Yes, though the average size per transaction is still lower. So you have a higher frequency and in equities the overall volumes have gone up, but the size per trade is not larger than it was previously. So I think it is a mixed bag. I think in the cash markets and the bond markets, and certainly in the Treasury market, there has been no call to arms to go towards a decimalised format.
Click here for Part Two of the Roundtable.