The poor performance of many subordinated debt issues in 1998 deterred investors from moving down the credit curve.
However, confidence is returning, and bankers anticipate a surge in issuance during the second half of 1999 - particularly of perpetual preferred shares which the BIS recently agreed to evaluate as tier 1 capital.
Charles Olivier reports on the development of demand for capital securities as investors seek higher returns without taking on much greater risk.
Investors shifting out of senior debt into subordinated bonds in the search for yield have not had the easiest of rides over the last eight months. After a year of relatively steady growth, the market crashed spectacularly in August 1998 when the Russian crisis sent fund managers scurrying up the credit curve to the safety of triple-A rated senior debt.
During the next two months, spreads on almost all subordinated debt widened significantly. The hardest hit were investors holding perpetual paper issued by Japanese banks such as Fuji Bank and Sakura, and US dollar denominated tier 1 issues by European banks such as Société Générale and BNP.
Fuji's $1.6bn perpetual, for example (which had been launched at 400bp over US Treasuries) widened to 2,500bp. Meanwhile, SocGen's $800m perpetual fell from 275bp over to 540bp over.
Even double-A rated lower tier 2 issues suffered a battering. Between August and November 1998, Abbey National's $750m deal (to take but one example) widened from 93bp over Treasuries to 140bp over.
It was hardly an auspicious lead up to the new credit oriented world which the introduction of the euro was to herald.
The poor performance of many subordinated issues relative to senior debt - which widened by just 10bp-20bp during September and October 1998 - came as a huge disappointment to the European investment banking community. Many had pencilled in 1998 as the year when European investors would finally begin to buy subordinated debt on a regular basis - thereby providing local borrowers with an alternative to the liquid (but quite expensive) US market.
In the previous year the sub debt market made huge strides, with the first $1bn subordinated issue and a swathe of inaugural deals from new borrowers such as the National Bank of Greece and the Estonian Investment Bank.
In no time at all, the theory went, subordinated debt would graduate from a structured product into a global, commoditised asset class with tens of billions of dollars of issuance each year. Instead, the market dried up, leaving banks sitting on huge paper losses and investors more wary than ever.
But the subordinated debt market has recovered some of its poise. SocGen's tier 1 deal, for example, is now back to 275bp over Treasuries. Meanwhile, many lower tier 2 issues are now trading close to their July 1998 peaks.
Issuance has been patchy, and banks have had to be careful not to price deals too aggressively. Last month, Salomon Smith Barney ran into trouble with a Eu250m issue for Unibank which failed to find many buyers after being offered at just 74.5bp over Libor.
But many bankers are still convinced that the market for capital securities in Europe will take off. Following the Bank of International Settlement's October 1998 decision to permit perpetual preferred shares to count as tier 1 capital without losing their tax deductible status, issuer interest in the product has quickly increased.
Prior to the BIS move, only banks from Spain, the US and Germany could issue such securities and gain tier 1 status. As a result, firms from countries outside these three - like SocGen - were forced to devise extremely complicated deals involving Limited Liability Companies in Delaware or special purpose vehicles in Luxembourg. Not only were such deals too complicated for many investors, they also took months to arrange in order to be sure of regulatory approval.
Now, the structure (with or without LLCs or SPVs) is open to everyone. Earlier this year, SE Banken became the first Scandinavian bank to issue tier 1 securities under the new regime with a Eu150m deal via Goldman Sachs.
The elevation of perpetual preferred securities to de facto tier 1 status has had a huge impact on the rest of the subordinated debt market. Before the BIS ruling, the vast bulk of non-US subordinated issuance was dated upper tier 2 or lower tier 2 debt. Now, there is much less demand among issuers for either product.
"Tier 1 is far more important for banks," explains one London-based analyst. "It is one of the really key figures which analysts watch when evaluating banks. If a bank can raise tier 1 and still get tax deductibility on the interest payments, there is very little need to raise tier 2 capital."
Partly because of foot-dragging by various national bank regulators (some of which have yet to fall in line with the new BIS rule), there is still a steady flow of plain subordinated debt issuance.
Nevertheless, as Derek Mills, syndicate manager at Deutsche Bank in London, says, the main focus of the capital securities market is now on tier 1 debt.
Most activity is coming out of Germany, where banks are allowed to issue dated tier 1 debt thanks to a unique exemption awarded by the BIS in October 1998. The first bank to tap this market was Bayerische Hypo-und Vereinsbank (HypoVereinsbank) with a DM1bn 10 year deal via JP Morgan.
Deutsche Bank had issued a similar transaction back in January 1998, but HypoVereinsbank was the first to do it by the book and sell it to institutions rather than retail investors.
Last month, Dresdner Bank pushed the dated tier 1 structure one step further by launching a $1bn 32 year issue and a Eu500m offering via Merrill Lynch from a bankruptcy-remote SPV in Delaware.
These three transactions have caused quite a storm in the market. Most non-German bankers complain that dated debt should not be classified as tier 1 capital since it has to be repaid, and that by giving it tier 1 status it puts the German banking system at risk.
"The German banking system is screwed up, badly managed and fragmented," says one London-based syndicate head. "But instead of forcing them to reform, they are doing what the Japanese did - ignore the situation and bend the rules so they can carry on."
Whether such opposition would fade if other bank regulators were to follow Germany's line is open to debate. For all the talk about dated tier 1 debt undermining banks' stability, it is hard to avoid the conclusion that the criticism is little more than sour grapes.
However, such an event seems unlikely. "It is not a case of Germany taking the lead on this issue," says one central bank employee. "It's a case of Germany breaking the basic rules of bank regulation."
How will the market develop in the coming months? Certainly, investor interest in the asset class seems to be returning. Many of the deals brought to the market so far this year have been oversubscribed. Dresdner's controversial Eu500m dated issue - for all its complicated structure - was more than three times oversubscribed.
Meanwhile, Deutsche Bank still managed to find plenty of buyers for its $500m perpetual issue despite an embarrassing downgrade during the week of launch, and was able to increase the deal to $600m.
Given the sharp falls registered by most subordinated issues during the third quarter of 1998, the speed with which investors have regained their confidence is remarkable. "It normally takes a long time for investors to regain their confidence after a catastrophe of that magnitude," says one syndicate head.
Mills at Deutsche says demand for the product is increasing by the week. "We are seeing a lot of new accounts looking at the market," he says. "Most fund managers are buying it, and there is much more retail interest with absolute yields in euros so low."
As in other areas of the debt markets, the difficulty will be finding issuers to cater for this demand. Mills is confident that supply will emerge. "Every bank is looking at the market," he says. "Some can wait, some can't."
Others are less confident. "We are at that stage of the cycle where banks are now well capitalised, and margins on lending are falling so they are slowing down their loan growth. The number of banks who need to raise subordinated debt is not that high," says one leading origination official.
There are, of course, plenty of Asian and emerging market banks which need to improve their capital positions. Owing to international fears of a further market collapse, however, they are more likely to do so by raising equity than subordinated debt given the astronomically high coupons they would have to offer to attract buyers in the fixed-income community. Meanwhile, several countries - such as Japan and Indonesia - are currently pursuing government-led recapitalisation programmes.
In Europe, few banks need to raise tier 1 capital. Some, such as Spanish and Portuguese banks, probably have too much - nearly 40% of total capital on average, far above the BIS recommendation of 15%.
The exception to this is France, where the average tier 1 capital ratio in the banking sector is hovering around the 6% mark, well below the 7% preferred by bank analysts.
For this reason, much of the issuance during the second half of 1999 is likely to come from one of three sources.
The first is banks making acquisitions in cash. Under most national takeover rules, any goodwill arising from an acquisition must be deducted from the acquirer's capital base. To keep analysts onside, this capital must be replaced quickly.
Over the past two years, acquisitions have been a major driving force behind subordinated deals such as Svenska Handelsbanken's $350m perpetual FRN (issued to finance the purchase of Stadshypotek), Banco Ambrosiano Veneto's $500m dated deal (launched to cover its acquisition of Cariplo) and Deutsche's perpetual bond (relating to its Bankers Trust acquisition).
Todd Groome, head of the financial institutions group at Deutsche, believes that this trend will continue. "There are still a lot of acquisitions to come," he says. "The Italians have started to consolidate and there are plenty of banks looking to do more deals."
The second source of issuance will be banks, building societies and insurance companies looking to replace equity capital with subordinated debt as part of wider capital reorganisation programmes.
Over the last two years, insurance companies have become much more familiar with the subordinated asset class. Back in 1997, Axa raised $1.05bn in the market via a two tranche step-up perpetual arranged by Goldman Sachs, Merrill Lynch, Paribas and Chase Manhattan.
Now they are beginning to get their teeth into the new tier 1 capital structure. In April this year the Fortis Group launched a three tranche Eu650m transaction via Merrill Lynch.
The key difference between the Fortis deal and previous issues (such as Axa's) is that the bonds are non-cumulative (in other words, vulnerable to non-payment of interest in the event of a suspension of dividend payments) and thus able to be accounted as minority interest rather than debt on Fortis' consolidated balance sheet.
"It basically copies the tier 1 capital of banks," says Michel Baise, Fortis' finance director. "We believe that in future, insurance companies will be regulated in a similar way to banks. This creates a new benchmark for insurance companies seeking to optimise their capital structure."
Former UK building societies such as Abbey National and Halifax have been issuing lower tier 2 subordinated debt for some time. But many syndicate heads believe they would benefit from issuing tier 1 capital.
"A lot of them have the wrong kind of capital structure with too much of it coming from retained earnings," says Groome. "The rating agencies love it but shareholders might appreciate the leverage which would come from replacing some of it with subordinated debt."
The third and final source of future issuance is the most difficult to predict - corporates. Historically, about the only corporates tapping the market have been highly leveraged property groups such as British Land or management buy outs.
But earlier this year, British Airways became the first European corporate to issue preferred perpetuals in euros with a Eu300m issue via Warburg Dillon Read.
Despite the financial logic behind such a move - greater leverage and improved return on equity - most bankers believe that it is unlikely that many companies will follow British Airways' lead companies.
"I haven't heard of many companies considering a subordinated debt issue," says Mills at Deutsche. This view is echoed by many equity analysts who say that, at this stage in the economic cycle, leverage is not high up the list of corporate priorities.
But will the paucity of potential issuers result in lower borrowing costs? In a stable market, the answer would clearly be yes. Most banks are in much better health than they were two or three years ago, and demand for high-yielding credit product is strong.
But this is not a steady market. US interest rates are on the way up. Half the world is in recession. And the rapid fall of the euro against the dollar has made the future direction of European interest rates a difficult call to make.
All of these factors, say syndicate heads, make it unlikely that spreads will tighten much further in the coming months even though they are still some way above their pre-crash peaks.
Against this rather uncertain background, arrangers and issuers will need to pay careful attention to deal structure and pricing. Typically, this involves choosing one of five basic products - tier 3 securities, upper tier 2, lower tier 2, dated tier 1 or perpetual tier 1. All share the same basic characteristic that they are repaid after senior bondholders in the event of bankruptcy.
Each of these products has its own attractions for issuers and investors. Tier 3, for example, is held against the issuer's trading book, and is fairly cheap. Last year, Banca Intesa raised Eu200m of tier 3 capital at a coupon rate of just six month Libor plus 60bp. However, it tends to be short dated, and only popular with investors when there is a lot of trading in paper coming to the end of its maturity.
Lower tier 2 has a steadier historical performance than any other subordinated asset class. Deutsche's $1.1bn Yankee bond issued in 1996, for example, only fell by 50bp during the market crash of August 1998 - much less than the 100bp to 150bp widening witnessed among most upper tier 2 issues.
Upper tier 2 debt is usually around 100bp more expensive than lower tier 2 bonds, yielding 130bp to 250bp over Libor. However, it does give more of a boost to the capital base than lower tier 2.
Of the two basic tier 1 structures, perpetual debt tends to be better rated than dated debt although this advantage has become less clear cut since Dresdner managed to secure a one-notch drop (relative to senior debt) on its recent deal compared to the three notch drop incurred by Deutsche Bank when it issued a dated transaction the previous year.
In theory, dated debt should also be cheaper. However, this can often be hard to calculate. Deutsche, for example, paid 140bp over 10 year Treasuries on its $600m perpetual. Meanwhile, Dresdner paid 215bp over 30 year Treasuries on its $1bn offering.
The jury is still out as to which of the two structures is better. Banks involved in the Dresdner deal argue that (where it can be classified as tier 1) the dated product is more attractive since it is cheaper and more attractive to institutional investors.
But Groome at Deutsche argues that the perpetual is the better bet since it does not need as many "toxic" write down features in order to gain tier 1 status.
A common trend over the last few months has been to include call options and step-up coupons on tier 1 issues. Typically, these kick in two years before maturity or (in the case of perpetuals) the call date.
The scope for playing around with step-up levels in order to attract more investors should be relatively restricted given the BIS recommendation that it be no more than 100bp over the original rate. However, that did not stop Deutsche offering a step up of nearly 150bp on its recent perpetual.
Some issuers in recent months have gone for more complicated structures. In May, Swiss Re issued two subordinated debt deals - one Eu400m Perpetual Auction Reset Capital Solvency (PARC) issue and one Sfr600m perpetual deal (named SUPERBs by Warburg Dillon Read).
The first has a variable coupon rate which is determined every five years after the seven year point by an auction process. The second has its coupon rate reset after 12 years according to the lowest Swiss Re's senior debt rating.
Another key choice for arrangers and borrowers is the choice of currency. Historically, the US market (which is by far the deepest and most liquid in the world) has provided a much more efficient source of funding than its European rival.
But thanks to the introduction of the euro (which has encouraged local investors down the credit curve by removing the scope for currency arbitraging) the price differential has narrowed.
So far this year, a large percentage of the subordinated deals brought to the market by non-US borrowers have been denominated in euros. Standard Chartered's Eu600m perpetual and ABN Amro's Eu250m 20 year offering are just two examples.
Encouraging as this may, be the decision by both Deutsche and Dresdner to issue the bulk of their recent offerings in the US market shows that (for the larger deals at least) the US still has the edge in depth and liquidity.
If markets become volatile again, this superiority will become even more pronounced. During the last crisis (in August 1998), lower tier 2 issues in the Yankee market widened by roughly 50bp compared to a fall of 100bp or more for lower tier 2 issues in the Euromarket.
The big question facing every participant in the subordinated debt market is whether the investor base for the product can be built quickly enough to prevent a repeat of August 1998.
With the possibility of a rise in US interest rates, a downturn later this year cannot be ruled out. But if more retail and institutional investors can be enticed into the market, then subordinated debt should stand a good chance of weathering the storm.
In time, it may even be able to graduate from a structured asset class offering yield into a commodity product where capital gains as well as yield can be found.
If not, it will remain what it is now - a niche market kept alive by a handful of trading desks and the willingness of investment banks to hold unsold paper on their own balance sheets.
Such an assessment may be too pessimistic. Whatever its volatility, capital securities are still a highly useful source of yield for investors. Where else could they find Dresdner Bank paper paying 212bp over Treasuries?