Auctions are still sovereign debt managers’ preferred method of issuance but 2009 has been the year of the syndicated deal, with countries stepping up their use to unprecedented levels. Will they still be so popular when economies start to recover
and issuance begins to fall? Philip Moore investigates.
Sovereign borrowers don’t like syndicated deals. They never have, and they never will.
That, at least, used to be the conventional wisdom and to a certain degree it is still true. Although there has been a conspicuous increase in syndication over the last 12-18 months, auctions remain the favoured mechanism for selling debt among Europe’s governments in general, and especially among the larger sovereign borrowers.
According to data collated by HSBC, by the middle of November there had been €153bn of syndicated sovereign transactions year to date, compared with just €69bn in the whole of 2008 and €67bn in 2007.
Although that is an impressive rise, with €668bn raised through auctions it still left syndicated deals accounting for just 19% of all sovereign issuance in Europe year to date, compared with 11% in 2008 and 12% in 2007.
In general, bankers concede that one reason sovereigns are still reluctant to syndicate debt is because they don’t want to be seen to be paying fees with taxpayers’ money.
Counterbalancing the fee argument in some countries have been a number of factors ranging from anxiety about failed auctions through to concerns about the impact of the banking crisis on the primary dealer community.
"The crisis has clearly meant that the amount of capital available among primary dealers has fallen, which has made syndication more attractive for borrowers needing to place significant amounts of debt," says Pierre Blandin, head of SSA coverage at Calyon in London.
Nevertheless, for Germany and France, in particular, syndication has remained a technique to be used only in exceptional circumstances. "Germany is a very different animal to the other European sovereigns, because the Bund and the Bobl are deliverable into the futures contract," says Greg Arkus, head of frequent issuers and public sector debt capital markets at Credit Suisse. "That continues to make Germany the European benchmark and guarantees its liquidity."
The result is that Germany has limited its use of syndication to its dollar issuance, with very mixed results. When it launched its first foreign currency issue for more than 50 years, a $5bn deal in May 2005, Germany suffered the indignity of watching its long-awaited five year benchmark widen substantially in the secondary market.
When it returned to the dollar market this September, with a $4bn three year offering, the consensus among bankers appeared to be that the Finanzagentur had taken the lessons of its 2005 deal to heart.
Germany’s September issue, led by Bank of America Merrill Lynch, Citi, Deutsche Bank and HSBC, raked in orders of about $11bn, but resisted the temptation to increase the size of the transaction.
Accommodating France
If that suggested a more accommodative stance towards investors’ requirements, so too did one of the other highlights of the market for syndicated issuance in 2009. This was France’s €6bn syndicated 2041 transaction in June, via Barclays Capital, BNP Paribas, Credit Suisse, HSBC and Société Générale, which generated demand of around €9bn.
To Zeina Bignier, head of sovereign, supranational and agency origination at Société Générale Corporate & Investment Banking, the most significant point about the French 30 year deal was what it said about the Trésor’s willingness to adapt its strategy to meet investor demand.
"Investors indicated that they wanted a large and liquid 30 year bond, which could only have been achieved through a syndication, and the Trésor was prepared to listen to their request," says Bignier. That was important, because France has not always been regarded as the world’s most amenable borrower when it comes to responding to investor demand.
The Trésor, says Bignier, reaped impressive qualitative and as well as quantitative rewards for its adaptability. "The Trésor would not have been able to raise more than €2bn or €3bn if it had issued a 30 year bond through auction," she says. "Nor would it have generated such well-diversified demand. France represented less than 30% of distribution, and most of the bonds were placed with real money accounts."
Of all the borrowers to have embraced syndication in 2009, however, bankers say that few have done so to more sparkling effect than the UK. Before this year, the UK’s Debt Management Office (DMO) had issued only once in this form, choosing the syndication process for a £1.25bn 50 year inflation-linked transaction in 2005. That choice, according to the DMO, was motivated "by the need to ensure appropriate pricing of the Gilt was achieved in a context where there was no individually adequate price reference in the sterling fixed income market around that maturity."
Following a period of detailed consultation which began in December, the UK took up the syndication option again in June, launching a £7bn 25 year Gilt via Barclays Capital, Goldman Sachs, HSBC and Royal Bank of Scotland. That is the equivalent of €8.3bn, making the transaction the largest syndicated issue ever launched by a European sovereign, and powerfully underscoring the capacity of syndications to offer huge size quickly and efficiently.
The sheer scale of the volumes accessible via the syndicated market also became clear when the DMO extended its use of the mechanism beyond the nominal market, with a 2042 linker in July. This generated total demand of £8bn, allowing the DMO to print a £5bn benchmark — which compares with an average size of between £1bn and £2bn for linkers sold via auction.
"The DMO spent considerable time analysing the merits of syndication," says Paul Richards, head of international DCM syndicate at Bank of America Merrill Lynch. "It is safe to say that the mechanism has served its purpose beyond expectations. The DMO has now syndicated four deals since June and many if not all have raised greater volumes than originally expected."
Other bankers share the view that the most immediately striking feature of the DMO’s syndicated transactions in 2009 has been their size, which is an important consideration for a borrower with a colossal Gilt funding requirement in excess of £200bn. "The clear upside of syndication is that you can issue in much larger size, and generate much more visibility and press coverage as a result," says Spencer Lake, global head of debt capital markets and acquisition finance at HSBC.
Uncovered auction
The argument in favour of syndications was, of course, done no harm by the shot across the bows delivered by the UK’s infamous ‘failed’ auction of March, when only £1.63bn of orders were generated for a £1.75bn offer of 40 year Gilts.
The DMO put a brave face on the first failure of a conventional Gilt auction since 1995 (two linkers in 1999 and 2002 were also not fully covered), arguing that the auction took place in exceptional circumstances. That was true enough: sterling had just edged down, inflation had just edged up and Bank of England Governor Mervyn King had just given the government a stark public warning about the dangers of any further fiscal stimulus.
The ‘failed’ auction was immediately followed by an auction for a £1.1bn 2022 index-linked Gilt that was 2.72 times oversubscribed, which reinforced the argument that the failure of the 40 year auction probably had more to do with bad luck than bad judgement.
The same can be said of other recent weak auctions, according to a recent OECD report. "Less successful auctions have been ascribed to uncertainties about potential market absorption issues and volatile market conditions," this notes. "Thus far, difficulties surrounding auctions can best be interpreted as ‘single market events’ and not as unambiguous evidence of systematic market absorption problems. However, the future could become more challenging, given that this still rising issuance trend is occurring hand in hand with increasing overall debt levels."
Nevertheless, the episode obviously added grist to the mill of those who warn about the damage that could be done by uncovered auctions. "A failed auction is a sovereign borrower’s worst nightmare," says one banker. "At the end of the day the market is unlikely to remember if a borrower funded at 4% or 4-1/2%. But no debt manager wants to be remembered as the one who failed to pull the money through the door on his watch."
Syndications = size
The sheer size that issuers have been able to execute within a single syndicated transaction in 2009 has been a key feature of the new issue market for sovereign borrowers with much more modest funding requirements than the UK. "Because the lead time for a syndication is so much longer than it is for an auction, investors have plenty of time to prepare, which in turn allows for bigger issues," says Chris Tuffey, head of European syndicate at Credit Suisse. "For the larger sovereigns, auctions raise between €1bn and €3bn, whereas we have seen several syndicated issues raising €6bn or €7bn over the last year or so."
Vastly increased transaction sizes and enhanced visibility have, for example, been important features of Greece’s achievements in the international capital market this year. Neither would have been possible had Greece relied on auctions to meet its borrowing target for 2009, which is why its funding strategy has been firmly anchored this year in syndicated transactions across the yield curve.
The formula certainly paid off handsomely enough at the start of November when Greece issued its seventh syndicated deal of the year and possibly its most successful, generating €11bn of orders for a 15 year benchmark. That allowed for the €7bn deal to be priced at the tight end of guidance — 142bp over mid-swaps — which was especially impressive as the deal was launched days after Greece was downgraded from A to A- by Fitch.
November’s spectacular deal took Greece’s total syndicated issuance for the year to €49.5bn. That is comfortably above the target for its total borrowing of €42bn-€43bn originally forecast at the start of the year and represents an unusually large chunk of the €66.5bn that Greece had raised by early November.
That total volume has probably elevated Greece to mid-table in Europe in terms of overall borrowing needs. But for a number of smaller European sovereigns, raising substantial volumes of funding has been an important priority to ensure that they are not marginalised because of the low liquidity they offer relative to the largest borrowers. "Borrowers like Austria and Portugal with relatively low borrowing requirements have potentially suffered from illiquidity in the secondary market," says Tuffey. "One way they could bolster liquidity would be to bring larger deals via syndication and to have fewer but larger auctions. That may be something that the non-core sovereign borrowers will consider next year."
Among other sovereign borrowers to have stepped down the syndication route for longer dated issuance in 2009, Sweden used the mechanism in its domestic market for the first time in March when it launched its first 30 year benchmark. Originally envisaged as a transaction raising a minimum of Skr20bn ($2.8bn), demand for this syndicated offering was such that the Swedish National Debt Office was able to print a Skr38bn issue.
"Our main funding vehicle has been and will continue to be the auction mechanism," says Maria Norström, head of institutional funding at the Swedish National Debt Office in Stockholm. "But we chose to use the syndication route for our 30 year deal because we identified a natural structural demand for long dated assets among Swedish pension funds." Sweden’s increased funding need in 2009, driven chiefly by its on-lending requirement to the central bank, meant that for the first time the Debt Office was able to meet that demand by offering a liquid 30 year benchmark.
Initially, Norström explains, the Debt Office’s strategy was to play the role of lead manager for the issue itself, arranging a series of bilateral meetings with natural buyers of long-dated debt and analysing feedback from domestic institutions. That turned out to be easier said than done.
"It was not an easy process because we did not get a sufficient response from institutions or a clear enough indication of what their potential interest would be," says Norström. "So even though this was untested ground for our non-international primary dealers we explained to them that this was a transaction we would be unable to do without their help."
The septet of primary dealers bookrunning the transaction was made up of Barclays Capital, Danske Bank, Handelsbanken, Nordea, RBS, SEB and Swedbank. Based on the advice of those primary dealers and on its own analysis, the Debt Office began by pricing the long bond off Sweden’s 10 year yield curve, at a range of 75bp-85bp over the 2019 bond.
So far so good. Then the market turned. "Soon after we started marketing, rates began to rise and there was a sell-off across the curve," says Norström. "So we had to steer away from the normal way of pricing a syndicated deal, which meant using an absolute yield. It was a cumbersome process and it took longer than a usual foreign currency syndication, but we were pleasantly surprised by the outcome."
More converts — even at the short end
Norström says that the Swedish Debt Office would look at revisiting the syndication option if it issues another long dated bond. Other borrowers, however, are becoming converts to the powers of syndication not just for longer dated issues but also for standard maturities for which European sovereigns have historically favoured using auctions. "A key characteristic of 2009 was that we started to see syndications being used at the shorter end of the curve," says Cesare Roselli, head of SSA coverage at Morgan Stanley.
Finland, for example, which syndicated a highly successful 15 year transaction in October, will be using the technique more frequently in the future. "We will be shifting to a new strategy of issuing two syndicated transactions per year, most probably in the five and 10 year maturities, supplemented by auctions," says Teppo Koivisto, director of finance at the Finnish State Treasury. "We have always been pro-syndication, which we believe is a very hands-on way of approaching the market. Syndication is a good way of helping issuers to see exactly where their bonds end up, which is especially important in the sort of uncertain market environment that we have seen in 2008 and 2009. I believe many DMOs in Europe are considering increasing their use of syndication."
The view that syndications strengthen the relationship between issuers and their investors on a number of levels is shared by bankers. "In general the syndication tool enables borrowers to get much closer to the investor base, and it also gives issuers a much better idea of the demands and preferences of their investors," says Roselli.
On the same theme, bankers say that syndications give borrowers much more influence over the allocation of bonds. "Syndications give borrowers a much clearer view of the bookbuilding process and therefore allows them to exercise a higher degree of control over allocation," says Arkus at Credit Suisse.
Whether or not the investor base that can be reached through a syndication is ultimately materially different to the pool of investors accessible via an auction is open to question.
Spain reaps rewards
Like the UK, Spain has been able to generate colossal demand for its syndicated offerings this year, with the book for its 10 year deals in February and May reaching about €15bn in both cases.
The most discernible trend in Spanish government syndicated benchmarks in recent years, however, has been the increase in domestic distribution. In the case of its €7bn 10 year issue in February, for example, 46% was placed in Spain and Portugal. That compares with domestic distribution averaging around 18% in 2006 and 25% in 2008.
At Bank of America Merrill Lynch, Richards says that although syndicated transactions can clearly open up a broader range of investors, there has inevitably been a localisation of demand during the crisis. That means that it is probably misleading to assume that there is any link between the increased use of syndication and the larger share of domestic placement in 2009 for issuers such as Spain and Greece.
"In times of stress, investors gravitate towards their comfort zone," he says. "It is no surprise that at the height of the crisis there was limited ability for non-US based issuers to access the US market and vice versa."
For a borrower like the UK, however, although investor diversification may be a beneficial side-effect of syndication, DMO chief executive Robert Stheeman reaching into new pockets of demand was never at the forefront of the DMO’s motives for exploring the syndication option.
The DMO’s move towards syndication, he explains, had more to do with ensuring that the timing of supply in the Gilt market was more closely attuned to the demands of large institutional investors in the UK. As the DMO explained in its original consultation paper, "pension fund demand for long bonds tends to be lumpy and sporadic". Hence, it added, "the timing of demand from pension funds is not necessarily aligned with the DMO’s debt auction calendar."
Counting the cost
If measuring the impact of syndications on investor confidence is more of an art than a science, so too is assessing the overall cost of syndicating new issues versus selling debt via auctions. "The problem with trying to drill down on the relative cost savings is that it is subjective," says one banker. "If you asked 10 people for an estimate of the all-in costs you would probably get 10 different answers."
At a very basic level, auctions are of course cheaper, because there are no fees payable to investment banking intermediaries. That also makes auctions politically expedient at a time when fees paid to banks are such a sensitive subject.
On an all-in basis, however, the costs may well be lower, especially if the breadth of demand, and therefore the size, of syndications creates price tension and therefore helps to minimise costs.
Quantifying the savings that syndications can generate can, however, be little more than a theoretical exercise. Although the UK DMO has made a number of internal calculations of the relative all-in costs of syndications relative to auctions which it has shared with the Treasury, Stheeman says that it would be inappropriate to release them publically.
Others agree that although it is impossible to quantify the costs of a syndication relative to an auction, there is growing evidence to suggest that the economics of syndication stack up very well on an all-in basis, especially in large transactions. "In a larger than normal auction you’re asking your primary dealers to take down larger portions of risk," says Tuffey at Credit Suisse. "The danger is that the market will make the sovereign pay for that risk by effectively cheapening up the bond in the run-up to the auction. So we would argue that a syndication probably ultimately gets you to a tighter spread on an all-in basis for a large deal than an auction would."
Views are mixed on the outlook for volumes of syndicated deals in 2010. A popular view is that for the larger sovereign borrowers in particular there is no need to change a mechanism that has proved its resilience in very turbulent markets. "The larger sovereigns will stick to their current modus operandi," says Blandin at Calyon. "They have demonstrated that they were more than capable of raising all the funding they needed throughout the crisis so they won’t see any reason to change their strategies."
At Barclays Capital, however, Taor sees room for more expansion, and for more variety, in the syndicated market. "Most European sovereigns have experienced some very tough auctions in 2009 and the likelihood is that they’re going to get tougher in 2010. I think the drift towards syndications will continue next year, especially as the inflation-linked market picks up. If Germany were to do a long dated index-linked deal, for example, there is every chance that it would be syndicated."