Squeezing more into tier one


Published on:

It must be perpetual but it doesn't have to be for ever. It has to feel like equity but look - to tax authorities - like debt. Defining banks' core capital is one of the thorniest issues facing bank regulators. Antony Currie reports on the squabbles over what fits in the tier-one category.

squeeze-t1-160It's hardly been the best year for banks to raise capital. The fallout from last year's Asian crisis scared off investors concerned that the banks might have overly high exposure to some of the affected markets, and so drove out spreads and hit share prices. Then, just as confidence was beginning to return, the Russian crisis hit, and deals either had to be restructured - Fortis, for example, switched from a straight secondary equity issue to a convertible bond to aid in its acquisition of Belgium's Generale Bank - or deals were pulled entirely, the most high-profile one being the cancellation of the Goldman Sachs flotation.

But, barring a few notable exceptions, what this year proved in contrast to earlier crises - especially the Latin American debt crisis of the early 1980s - is how much better capitalized the major international banks are, and therefore how much more able they have been to deal with the last 12 months. One of the chief reasons behind this is the adoption, first by the G10 countries, and then others, of the 1988 capital adequacy rules laid down by the Bank for International Settlements (BIS), based in Basle.

Given the role the rules have played in creating more stability in the banking system, it's ironical that the banks want substantially to overhaul them. In terms of raising capital, and managing it, banks generally regard the rules as hopelessly out of date, and not without due cause. On the asset side, the risk-weightings are the most obvious shortcomings. A high-grade corporate, say a Unilever or a GEC, has a 100% risk weighting attached, but so does a start-up telecommunications company, or a corner shop. Lend to a Korean bank, however, and the risk-weighting is just 20%, because it is a bank in an OECD country.

This is the trickier side of the problem, however, described by one banker as a "horrendous can of worms which we'll have to open at some point soon". The prospects don't look good. Admittedly, the regulators are willing to address the issue. Bill McDonough, chairman of the Federal Reserve Bank of New York and incoming chairman of the BIS, announced in late summer that he would like to see the rules changed within the next two years. But the experience of changing the rather simpler issue of liability management does not bode well.

At the end of October the BIS issued a statement announcing changes to the rules about raising tier-one capital. This is the core element of a bank's capital adequacy, as the statement reaffirms, and consists of common stock and retained earnings. The issue the BIS was addressing was preference shares. These offer banks a way to increase their tier-one equity by issuing debt instruments which, by being perpetual, appear to take the form of equity. Capital-raising using perpetual preference shares has three main benefits. First, it allows the bank to issue more common stock without diluting the holdings of its equity investors. Second, when times are rough, preferred stock is an easier form of equity to issue since investors are less likely to buy into an ordinary share issue (Barclays was a big issuer of preferred stock when experiencing problems in the early 1990s). Third, the return on capital it demands is less than for common equity.

The problem was that the BIS rules on tier one did not allow for tax-deductible preference shares to be issued. That was not a problem back in 1988, and on the grand scale of bank capital it should not be more than a minor irritant. Preference shares offer a cheap form of capital whether they are tax-deductible or not. Earlier this year, for example, UK bank Halifax indicated that it was prepared to issue taxable preference shares to help finance its acquisition of Birmingham Midshires building society.

But tax deductibility was to become a major problem by the mid-1990s. In late 1993 corporates were allowed to issue tax-deductible preferred shares in the US, these were known as quips and mips. Of itself, it was of little consequence to banks that firms such as Cadbury, Grand Metropolitan or Schweppes could issue such shares, but once non-bank financial institutions - especially insurance companies - started to do so, the banks started to complain that this effectively gave the former a competitive advantage. Unlike banks, non-bank financials are not regulated by the Federal Reserve, nor bound by the BIS rules that appeared to prohibit tax-deductible issues.

Some of the banks put a solution, of sorts, to the Fed in 1995: why not allow them to set up operating companies along the lines of Reits (real-estate investment trusts), stuff the operating companies full of real-estate assets, and issue tax-deductible debt that would then be treated on the balance sheet as minority interest?

There was a problem here, too. Such a structure would only work for those banks that had significant real-estate holdings; for those that did not, there was still no solution. So, by mid-1996 another proposal was put on the table. Bank holding companies should be allowed to issue deeply subordinated debt as preferred stock, and then downstream it to the banks as tier-one capital. This side-stepped the regulatory trap entirely, as the bank holding companies do not come under the regulatory jurisdiction of the Fed.

The Fed agreed, but rumour had it that the tax authorities were none too happy. What to the banks was a beautifully simple way of circumventing unhelpful regulations appeared to the tax man to be little more than a tax dodge. Bankers thought that the tax authorities might simply impose tax on these structures and destroy their very reason for existence. In the event this did not happen, but bankers decided to act while there was still a window of opportunity.

While the going was good

Between October 1996, when the Fed gave its approval, and February 1997, US banks issued over $30 billion in tier-one tax-deductible stock. Called capital securities, the basic structure was a 30-year, non-call 10 bond. Although no limits were put on how much tier-one capital could be made up of capital securities, the main rating agencies were quick to point out that if it were to go over 20% of the total tier-one capital, a bank's rating could well be in line for a downgrade, so 20% became a de facto limit.

This compromise galvanized the European banks into trying to force their regulators to adopt a similarly flexible policy. It was bad enough that Spanish banks could issue tax-deductible preferred tier-one stock (they had not signed the Basle accords). But with US banks now doing the same, it was time to push for a level playing field of issuance.

Some had tried before: in 1993 French banks Crédit Agricole and Crédit Lyonnais issued tax-deductible structures as tier one, but regulators rejected their rationale and the banks had to reclassify them as tier two.

By 1997 the French banks again, and also Italian institutions, were looking at ways of boosting their capital, and by the end of the year the most pressing need came from the Japanese. The national regulators, after initially proving reluctant to approve the varied structures shown them by commercial and investment banks alike, finally started to consult each other on how to deal with an anomaly which, though minor in the grand scheme of bank capital raising and management, was proving to be exceptionally divisive.

Bankers were expecting a judgement from the BIS relatively quickly. But a spat between the German and the US regulators was to destroy the emerging consensus. The Europeans' (and especially the Germans') main grumble was the US use of holding companies to issue the capital securities, which the former regarded as a fudge and wanted stopped. The US response was to object to dated securities counting as tier-one, which the Germans in particular wanted to include.

The German Landesbanken have no publicly listed shares because they are owned by state governments. Instead they issue dated debt securities called stille Einlagen, or silent participations. Fine for the Landesbanken; but not, said the Americans, for the publicly quoted banks. In mid-January this year, just as the negotiations on tier one appeared to be heading towards a satisfactory conclusion, Deutsche Bank issued a variant of Stille Einlagen, and so provided the Americans with the perfect ammunition with which to defend their holding company policy.

The talks stalled, and it was not until October that the long-awaited statement came. After deliberating for so long, its announcement barely covered two pages, not the lengthy, detailed, point-by-point ruling many had been expecting earlier in the year.

All the BIS has attempted to do with the statement is to give a broad outline of tier one for the national regulators to work from. And in parts it is little more than a fudge, and does not address at all the two issues that caused the talks to be stalled at the start of the year - the US banks' use of holding companies to issue and downstream the capital securities and the German use of dated silent participations as tier one. In fact on the latter point the BIS has been very careful in its wording: in its list of what qualifies a structure as tier one it states that all issues have to be "permanent", rather than the usual wording of "perpetual".

As basic 10-year paper, silent participations can in no way be considered perpetual. But they form an essential form of funding for unlisted Landesbanken and are usually renewed at maturity. They are usually seen as the equivalent of equity for Landesbanken.

Finite perpetuity

But perpetual no longer means perpetual in any case. The deals issued by US banks are called perpetual, but in fact are 30-year securities with a call, and a step-up, after 10 years. For their purposes, the rating agencies view such deals as 10-year paper, as do investors, who know that the step-ups written into the deal are punitive enough to give confidence that banks will call the bonds in.

Otherwise the main points are that special purpose vehicle (SPV) issues for tier one purposes be limited to 15%. And to qualify as tier one the instrument must be:

  • issued and fully paid
  • non-cumulative
  • able to absorb losses within the bank on a going-concern basis
  • junior to depositors, general creditors and subordinated debt of the bank
  • permanent
  • neither secured nor covered by a guarantee of the issuer or related entity or other arrangement that legally or economically enhances the seniority of the claim as regards bank creditors
  • callable at the initiative of the issuer only after a minimum of five years with supervisory approval and under the condition that it will be replaced with capital of same or better quality unless the supervisor determines that the bank has capital that is more than adequate to its risks.

There is also a set of limitations to step-ups: either to 100 basis points, less the swap spread between the initial index basis and the stepped-up index basis, or to 50% of the initial credit spread, less the swap spread between the initial index basis and the stepped-up index basis.

But it is not clear what, exactly, is to be included in the 15%. The BIS has left it up to the individual regulators to determine how to implement the general ruling. For instance, some think that there is enough leeway in the BIS's statement for any non-common stock instruments to be included in the 15% basket: the announcement does not explicitly state whether the 15% limit refers to all forms of SPV, to dated or callable instruments or to step-ups. Nor is there any mention of callable-only issues. Given the rules on step-ups, a step-up callable will only be worthwhile for those whose paper has big spreads.

The exact wording is that the BIS has "taken note that over the past years some banks have issued a range of innovative capital instruments, such as instruments with step-ups" for tier-one purposes, and that "such instruments will be subject to stringent conditions and limited to a maximum of 15% of tier-one capital." In effect, what the BIS has done is to legitimize the approach already taken by the rating agencies, and admitted that there now exists an upper and a lower tier-one capital structure.

Some could not wait for the rule changes, however. The Spanish banks could issue in any case, but of those adhering to the BIS rules, first Japanese, and then Italian and Dutch banks followed suit.

For most of them, however, the trick was to construct a structure to allow the issue to bypass the rules and get them tax-deductible status. The Japanese banks, desperate to recapitalize, chose to issue credit-linked notes backed by US assets. At least, that was how it was sold. But some investment bankers point to how it was really a special purpose vehicle "with an intercompany back-to-back loan and a fig leaf over it" as one banker puts it. "They set up an asset-backed subsidiary, but it only had virtual assets, in the sense that they were mainly US treasuries, and the bank didn't have to own them."

In any event, the Japanese authorities allowed the issue on the basis that it would only count as tier one for the first financial year; any extension of that would depend on the BIS rule change in the pipeline.

Italians go for get-out clause

Two Italian banks also issued structured, tax-deductible preference shares. Neither the banks, nor the regulators, were as willing - or as desperate - to take a bet on the outcome of the BIS talks as the Japanese, so the structures they employed, though SPVs with back-to-back loans, had a get-out clause: were the BIS rule change to go against such a structure, the loans could be replaced with real assets, and so keep the issues as tier one. According to lead manager Merrill Lynch, it took nine months to structure and launch the deal for Banca Commerciale Italiana. Launched in mid-June, it was split into three tranches: one for $200 million, one for £120 million ($200 million) and one for €550 million ($640 million), each perpetual issues, but with a step-up and a call at 10 years. The back-to-back loan is between the SPV and the bank's New York branch.

Only the issue for ABN Amro was a straightforward tax-deductible deal. But, unlike the Italians, ABN Amro's paper was targeted at the US retail investor base, which has a big appetite for capital securities-type paper. Only a small number of European banks can issue in this format as most lack the necessary registration with the SEC.

The structures have not pleased everyone; indeed, one banker sees them as particularly unhelpful: "There's too much engineering going on. Preference shares are valuable tools regardless of whether tax-deductible or not. It's just child-like posturing by banks and regulators, and all it does is to make life much more difficult for the investors. You can try as hard as you like to get them to concentrate on the end product, but they always focus on the boxes in the middle."

Others, if not as damning, take the point. "Maybe, if we're lucky, now the rules have been clarified we can focus on creating markets and placing the product rather than engineering solutions," says Gregg Sando, financial institutions group head at Warburg Dillon Read. "This was predominantly a dollar market, with some sterling. What we're hoping for now is to develop a market in euros, for European investors."

That may take some time. Investors are still cautious about putting money into bank capital until they learn more about the banks' exposures in the year-end results. Some banks, especially those in the UK, will be unable to issue until after reporting their results for 1998, effectively precluding them from issuing until March. And, in contrast with capital securities in the US in 1996, there is no perceived threat of a tax loophole being closed, so no-one is expecting quite such a glut of issues. Waiting for individual regulators to determine how to apply the new ruling may also hinder issuance from some countries. But tax and accounting rules could also pose a problem: if SPVs with a simple back-to-back loan are to be allowed by various regulators, there is a good chance that the tax authorities will pounce, claiming that such a deal is a direct issuance of preferred stock, and so should be taxed.

The potential is certainly there for this new form of tier-one issue to become more widely accepted in Europe. In part this is yet another consequence of the loss of convergence trades for investors at a time when government bond yields are historically low.

With investors having got their fingers burnt on both emerging market paper and the fledging European high-yield debt market in the last year, investing in European bank capital issues may be seen as the safest higher-yielding market left. And there is growing evidence that investors are beginning to think like this. Lead managers ABN Amro, Merrill Lynch and Salomon Smith Barney claim that at least $250 million of the $1.25 billion tier-one preference share issue for ABN Amro was placed with investors outside the US. The major portion of BCI's deal was denominated in euros, and Banco Bilbao Vizcaya came to the market in November with a €700 million deal.

The majority of these last two issues went to investors in the issuers' own countries. Indeed, BBV's deal was the first Spanish issue of tax-deductible preference shares directed to domestic, rather than offshore, accounts. Previously, despite not being signatories to the BIS rules, Spanish banks had their own tax issues to deal with and would have to issue offshore, usually to Spanish funds in Andorra.

Yet banks were badly hit by the crises of the last 14 months, and many investors will be waiting to see what other exposures and losses will be reported by the banks at the year-end. For their part, banks will not be keen to issue with spreads still so wide compared with before August and few expect spreads to come in to those levels any time soon.

What deals, what price?

"The question now is what kind of deals are going to get done now, and at what price," says Christopher Grigg, managing director at Goldman Sachs. "We are confident that European investors will play a much bigger role in the future as bank capital is a great asset class for investors in the euro world. The banks are highly rated, well-known names in an environment where investors are having to think a great deal more about spread product than they did in the past."

The first test looks set to be for Deutsche Bank. If the proposed cash acquisition of Bankers Trust goes ahead as planned, it will eat into its tier-one capital. Of the $9.7 billion offered, analysts expect that Deutsche's tier one could be hit by as much as $4 billion. Although this would still keep the German bank's overall capital ratio above the 8% threshold, at 8.5%, tier one would not make up the required 50% of its capital (in fact it would be just under 43% of the capital ratio, taking the pure BIS tier-one ratio of 5.1% - international accounting standards figures put tier one at 6.2%)).

Issuing preference shares might be an option, or even stille Einlagen, now that the bank has ruled out selling its 12.7% stake in DaimlerChrysler. More likely, however, is an exchangeable bond.