Bank capital: A new holy grail every year

Just one year ago, bank capital specialists were getting ready for the big push. The objective was to persuade authorities throughout Europe to accept tax-deductible tier one securities. Victory in this battle would give every bank access to a cheaper form of core capital than equity. This hybrid tier one capital, said its advocates, would become the perfect way for banks to fund acquisitions as they turned themselves into pan-European institutions.

A year on, and the same bankers can hardly believe the scale of their achievement. “The year 2000 has been the annus mirabilis of tier one bank capital,” says David Marks, from JP Morgan’s financial institutions group in London. “It has been the acquisition finance tool of the year.”

Between January and November, European banks issued over $15 billion-worth of these securities. Most of this issuance has been used to finance purchases, though mostly of the domestic rather than the cross-border variety. And banks in every major jurisdiction have now persuaded their authorities to let them raise this tax-advantaged form of core capital.

Since October 1998, the world’s banks have been waging a war of attrition with their regulators and tax authorities. Battles are won by tweaking ever more complex structures until the authorities tire of ploughing through dummy prospectuses and allow banks to issue the securities. “The regulators have been through two years of pain as banks have browbeaten them into accepting these structures,” says one investment banker.

The US was the first big jurisdiction to allow tax-deductible tier one issuance. Even before the Basel guidelines were introduced, banks were allowed to raise tier one capital by issuing so-called preferred securities through a Delaware-registered trust. This special-purpose vehicle issues tax-deductible securities and lends the proceeds to the parent bank. Most US banks have now done such deals and the structure has become highly standardized. “Once these things have been around for a while they get commoditized,” says one capital securities banker.

So for the last two years, Europe has been the main battleground of this market. “The most significant development in this market over the past 18 months has been the trend for European countries to allow tax deductible tier one securities,” says Chris Grigg, managing director at Goldman Sachs in London. “The Basel guidelines opened the way for this type of security. But it has been a battle fought jurisdiction by jurisdiction to get both regulators and tax authorities comfortable with these structures. Each national regulator has a slightly different approach and each tax code is different. The challenge in each country is to create a structure which is not abusive from a tax perspective and which does not take away from the quality of a bank’s capital.”

The biggest victory of the last 18 months has been in Britain, where Halifax issued the UK’s first tax-deductible tier one security last December. Then France gave way too. In January SG became the first French bank to raise hybrid tier one capital (at least in a domestic currency). Then came the Italians: first Banca Lombarda in March, followed by Unicredito Italiano in September and San Paolo IMI last month.

The trigger for most of these deals has been consolidation. “We have seen most tier one issuance this year from countries where there have also been a number of big acquisitions, mostly notably the UK,” says Grigg. Lloyds TSB, for example, raised the equivalent of some $830 million in January to help pay for its acquisition of Scottish Widows. In April HSBC issued more than $3.5 billion of tier one paper to finance its purchase of French bank, CCF.

But structuring European deals remains complicated. A lot of effort goes into finding ways to fine-tune the structures to make them suitable for each different country. “The basic models haven’t been changed,” says Bob Motyka, director in UBS Warburg’s capital products group in London. “But they need to be adjusted to the specific requirements of each jurisdiction.”

The bank capital market is, at heart, fundamentally mischievous. Deal-makers engage in a battle of wits with tax authorities and bank regulators, portraying the same instrument as debt to the former and equity to the latter. At the same time, investors also need to be reassured. Hardly surprising then that in this market the merits of obscure clauses in deal documentation are debated with passion. One of the biggest talking points this year has been a tactic known as stock settlement.

The perils of perpetuity

The biggest drawbacks of tier one and upper tier two securities is their most obvious feature: they are perpetual instruments. Some retail investors are willing to buy undated debt, but any institutional investor knows that this is the equivalent of writing a blank cheque. If the issuer’s borrowing costs fall, it will call the perpetual: but if its borrowing costs rise, the investor is left holding an instrument which pays less than the market rate.

To make them digestible for institutional investors, perpetual securities include a coupon step-up which coincides with a call date. An investor who owns Halifax’s sterling upper tier two bond issued in May, for example, knows that the bank will probably call the bond in May 2016, because if it fails to do so the coupon will increase from 7.5% to the equivalent of 345 basis points over gilts.

But what happens if by May 2016 Halifax’s tier one is trading higher than 345bp over gilts? In that event, the bank would have no incentive to call the bonds. So structurers have been looking for ways to make that artificial maturity a little more real. “One obvious way to strengthen the synthetic maturity is to increase the step-up,” says Grigg at Goldman Sachs. “But this detracts from the permanence of the capital, so regulators don’t allow it. The other way is to find some other mechanism which gives investors more comfort that the securities will repay whilst maintaining the quality of capital.”

Goldman’s solution, stock settlement, was first used on a deal for Fortis in June 1999 but came into its own this year with larger deals for UK banks such as Lloyds TSB, Abbey National and Royal Bank of Scotland.

To take one example of the structure, Abbey National’s $1 billion of tier one securities issued in February through Goldman Sachs and Lehman Brothers have a step-up and call on June 30 2030. But on that date, bondholders also have the option to force Abbey to raise enough cash to redeem the securities by issuing ordinary shares.

If the stock settlement clause is invoked, it will be shareholders, not the bank itself, which has to stump up the cash to redeem the securities. That means the clause doesn’t detract from the permanence of the capital.

What is controversial about stock settlement is whether investors can really rely on it as a means of ensuring repayment at the call date. Some bankers point out that the language in the documentation gives the issuer some room to wriggle out of this commitment. The bank can hardly be expected to get authorization from shareholders for a possible capital increase in 30 years time, so instead it simply promises to make best efforts to clear the issuance of new stock at the call date.

And what if Abbey National is in real financial trouble in June 2030? If its shares are nearly worthless, it would take a huge capital increase to repay the securities.

Shareholders would have every incentive to try to block that degree of dilution.

Those bankers who see some merit in stock settlement point out that the clause is not designed to address a credit crisis. Rather, it is meant to protect investors from market turbulence. In the midst of the Russian debt crisis, SG’s perpetuals briefly traded wider than their step-up levels, even though the French bank was not facing a credit crunch.

Any bank which finds itself in that situation on a call date should have a strong reputational incentive to pay back the securities anyway. A little extra incentive can’t hurt.

“The biggest risk investors face on a capital instrument is arguably not credit risk, but the risk that the instrument will not be called,” says Damian Chunilal, global head of the financial institutions group in Merrill Lynch’s debt capital markets division. “Stock settlement provides the investor with some additional protection against that risk.”

The most interesting use of the stock settlement trick may be in cases where the issuer is not allowed to use a step up. In a regular preference share issue, for example, there can be no step-up-and-call feature. In February Royal Bank of Scotland issued $2.2 billion of preference shares with stock settlement. “I see further application of stock settlement for banks which are unable to issue tax-deductible tier one,” says Grigg.

So far, stock settlement has been used only in the UK. It runs into a host of problems in jurisdictions where there are heavy restrictions on how and when companies can issue new stock. US regulators also frown on a technique which, according to some market practitioners, was first suggested to them several years ago.

The direct route

While the debate still rages on the merits of stock settlement, an even bolder attempt to re-engineer the tier one security has become the cause célèbre of the capital securities market this year. Barclays Capital, a firm with little track record in the bank capital market, set the cat among the pigeons in April when it arranged a tier one deal for its parent.

This security, which Barclays calls the Reserve Capital Instrument or RCI, was issued not by an SPV but directly by Barclays Bank and is documented as a Eurobond. That puts it clearly into the debt category as far as the tax authorities are concerned and interest payments seem likely to be tax deductible.

What’s more, the UK bank regulator, the Financial Services Authority, seems to like the structure and is happy to treat it as part of the issuer’s core capital. Barclays Capital has now issued three RCI deals: two for Barclays and one for UK mortgage bank Northern Rock.

If something so simple could count as tax-deductible tier one, why has no-one thought of it before? Its creator claims that others have simply been looking too hard in the wrong direction. “The trust preferred is a product developed for the US market,” says David Lyon, head of debt capital markets for financial institutions at Barclays Capital. “It is a piece of clever US domestic financial engineering but one that creates friction when applied in other jurisdictions.”

In fact the trick that makes the RCI work is almost identical to Goldman’s stock settlement innovation, but it applies to coupon payments not to the principal. If the issuer doesn’t have the cash to pay the coupon, it can instead issue shares and use the proceeds to make the coupon payment. That makes the bond cumulative as far as investors are concerned, but non-cumulative from the issuer’s point of view. In other words, the bond is a tier one instrument but with the same characteristics for the investor as an upper tier two instrument. That should make it cheaper for banks to issue.

“The fact that RCI coupons are cumulative for the investor is an attractive feature,” says Alan Patterson, managing director at Schroder Salomon Smith Barney in London. “It means the securities are closer to upper tier two in risk profile. However, the spread difference between tier one and upper tier two has already narrowed, so there is not a great deal of value that can be extracted there.”

The big danger for Barclays Capital’s creation comes not from the tax man nor directly from the regulator, but from accountants. One of the FSA’s criteria in allowing a security into the tier one basket is that the bank’s auditors must treat the instrument as equity when they are compiling the accounts.

PricewaterhouseCoopers and KPMG, the two biggest auditors of UK banks, are on opposite sides of a debate raging within the accounting profession over how these securities should be treated. PricewaterhouseCoopers is willing to treat RCIs as equity, while KPMG insists they must be accounted for as debt.

A committee within the Accounting Standards Board known as the Urgent Issues Task Force is now considering the question. Some bankers believe the task force will outlaw equity treatment for the Barclays structure.

The loss of equity accounting treatment would not necessarily spell the end of the Barclays invention. The FSA could simply drop its insistence on equity accounting treatment.

Says one of Barclays Capital’s rivals: “Imagine if Barclays Bank went back to the FSA in a few years time and said: ‘we want to issue exactly the same security, but this time our accountants are calling it debt’. If that happened, the FSA would probably have to allow it as tier one capital.”

While rivals are quick to point to the limitations of the RCI structure, not all are writing off the idea of directly issued hybrid tier one securities. “In the near term I don’t see much shelf life for the RCI structure,” says Patterson at Schroder Salomon Smith Barney. “It relies on accountants giving the securities equity accounting treatment which some are unwilling to do. However, in the long term, perhaps five years from now, we will probably end up with some form of direct issuance as a market standard across Europe.”

Some Nordic countries already allow a much simpler form of direct issuance – one which doesn’t rely on the sleight of hand of any kind of stock settlement. The race is now on to persuade other continental regulators to accept some form of directly issued security.

The Belgian and Italian regulators are reported to be considering proposals for just such an instrument.

“We believe that the direct issue security will become the standard over time,” says Ben Katz, head of hybrid capital products at Lehman Brothers in London. “Sweden and Finland already allow it, and other countries such as the UK and Denmark seem to be moving in the same direction.”

But other capital securities bankers believe the future for hybrid tier one issuance lies not with direct issuance but with the old familiar SPV structure. “We believe that the SPV structure will predominate for tier one issuance in the long run,” says Cristian Jonsson at UBS Warburg’s capital products group in London. “There are certain limitations, but it is the tried and tested, uncontroversial method and that makes for a very good asset class which facilitates issuance.”

Marks at JP Morgan agrees. “I am skeptical that direct issuance is the only way forward,” he says. “If there was no alternative, banks might strive to find some way to create a directly-issued security which squared the circle of achieving equity accounting treatment and tax deductibility. What banks need is a synthetic preference share which is tax deductible and the SPV and intercompany loan/deposit structure fills that role effectively in most jurisdictions.”

Bad memories

They may squabble about the best way to structure them, but capital securities bankers have achieved unity of purpose in their efforts to persuade European investors to buy tier one securities. This has been quite an achievement given investors’ bad memories of the last time Europe had a tier one market.

In the 1980s UK banks issued large amounts of preference shares. There was so much demand, especially from Japanese banks, that they were priced at as little as 25bp over senior debt.

Prices came crashing down and liquidity vanished when it emerged that the Basel Accords of 1988 would make it prohibitively costly for banks to hold the capital of other banks on their books.

For years, bankers have been plugging the argument that subordinated bank debt is not much more risky than senior bank debt. If you like the credit, goes the theory, you should get paid extra for taking subordination risk.

After all, banks only miss payments on their subordinated debt when they are in real trouble – and then senior creditors get hit too.

And if the fundamental credit is sound, subordination brings a little extra yield for not much more risk. In fact, it may be a better investment than the senior debt of a weaker credit.

That message finally seems to be hitting home. “In the early 1990s, selling bank capital to European investors was rather like suggesting they co-invest in a Colombian drug cartel,” says one veteran capital securities banker. “After the banking crises in Scandinavia and Banesto’s problems, you just couldn’t get airspace with investors. Now, every major fixed-income account in Europe has either bought tier one capital or is actively looking at getting into the market.”

In the past year, spreads on the most subordinated instruments – tier one – have moved closer to lower and upper tier two. Spreads on tier one securities may wobble alarmingly at every suggestion of an impending market crisis, but they no longer show the wild volatility which characterized these instruments until recently.

Not only is the investor base in Europe getting bigger, it is also becoming more professional. Much tier one debt in 1999 and earlier was sold to retail investors, often through the branch network of the bank that was raising the capital. That market has turned sour. Retail investors were not generally told of the benefits of synthetic maturities, step-ups and the like. As interest rates have risen they are left holding poorly yielding instruments and the bid for the paper has vanished. “This year, the market in European has been predominantly institutional,” says Patterson at Schroder Salomon Smith Barney.

That is good news for the small group of international investment banks who specialize in this business. No longer do they need to invite a Spanish or Belgian bank onto syndicates to help place the securities.

But competition to lead deals remains intense and seems to be getting fiercer. Merrill Lynch and Goldman Sachs dominate the tier one market in the US and both, Goldman especially, have led a good number of European deals. However, other firms such as Lehman Brothers, Schroder Salomon, Deutsche, UBS Warburg and now Barclays, have all stepped up their efforts in this area.

In anticipation of the inevitable pressure on fees, capital securities bankers are already searching for new areas in which to ply their trade.

Searching for new issuers

One much heralded source of new issuance is the insurance industry. And some are looking even further afield to corporates. But these companies seem unconvinced about the benefits of issuing complicated subordinated debt. “The volume of insurance company issuance has been disappointing this year,” admits Chunilal at Merrill Lynch. “But we are continually looking for more subordinated instruments that will allow rating agencies to give issuers equity credit.”

Christian Jonsson at UBS Warburg’s capital products group in London reckons issuance of capital instruments by institutions other than banks will definitely grow. “For corporates, these securities have the potential to gain equity credit from the rating agencies without diluting shareholders,” he says. “Insurance companies can also use them to increase their solvency margins.”

Finding new issuers of alternative forms of capital is essential for another reason. Banks are allowed to issue only 15% of their core capital in the form of hybrid tier one securities. Many European banks have already reached that threshold and may have no reason to come to the market again until their issues mature in 10, 20 or 30 years time.

Investment bankers searched long and hard for a tax-deductible security which European banks can issue as tier one capital. They routinely described it as their holy grail. They should perhaps identify as their next holy grail the search for another industry willing to issue these instruments.