Market Monitor


Published on:

Syndicated loans; Australia; Asset-backed Eurobonds; MTNs; Germany

Edited by Peter Lee

Syndicated loans: The perils of going Dutch

At 2pm on a Monday afternoon in mid-October, senior representatives of 10 of the Euroloan market's leading commercial banks gathered at Voorlinden at Wassenaar, near The Hague. They had not come to admire Voorlinden's impressive gardens and stately setting. It was the treasury officials of Dutch telephone company KPN and its loan arranger Goldman Sachs who commanded everyone's attention on the afternoon of October 14.

Several bankers nearly didn't make it, wrong-footed by the last-minute change of venue needed to accommodate all those wanting to attend. Lenders' meetings in this market are normally mundane affairs, but a surprisingly large number of senior bankers decided to see for themselves a presentation that promised to be eventful.

Two weeks earlier, a cry of dismay had risen from the loan market, when first-time borrower KPN made the unusual choice of Goldman Sachs - not recognized as a top arranger of syndicated loans for European corporations - to lead the jumbo Fl2.25 billion takeover financing to support its agreed acquisition of Australia-based TNT.

KPN's decision to award a sole mandate to the US investment bank would by itself have caused commercial bankers' emotions to run high. But by also opting for Goldman's innovative Dutch auction technique, the borrower caused more of a stir than it could possibly have expected. It quickly aroused an unprecedented show of commercial bank solidarity and intense irritation.

The proposal Goldman originally put to the market, defended with vigour by KPN treasury officials at the meeting, asked 10 banks to bid on price for underwriting commitments of Fl320 million each on the loan. The sixth most competitive bid would price the deal. That would be enough to subscribe the loan, assuming Goldman also kept an underwriting of Fl320 million. The four widest bidders would be left to consider joining a later sell-down in junior roles.

Goldman Sachs itself had fully underwritten the loan, but was not prepared to tell the market at what level it had done so - although this was eventually rumoured to have been 13 basis points over Libor, not particularly aggressive for a highly rated telephone company.

Of the 10 banks approached - ABN Amro, Citibank, Commonwealth Bank of Australia, Dresdner Bank, Generale Bank, ING Bank, JP Morgan, Rabobank, SBC Warburg and Union Bank of Switzerland - many felt outraged both by the structure and at being spurned by a client with which they had close commercial banking relationships.

With Goldman Sachs acting as financial adviser to KPN for the acquisition, normal practice would be for one or two top commercial banks to be appointed arrangers and underwriters for the senior debt portion of the financing. Goldman apparently talked KPN into mandating it alone to underwrite the loan with a view to further reducing the cost to the company by holding the auction. No doubt Goldman suggested that this was the best way to keep the entire transaction under wraps and ensure no last-minute pre-merger leaks. News of the deal came like a bolt of lightning to KPN's relationship banks.

The structure pays Goldman full fees for arranging and underwriting at the back-stop price. The benefits of tighter pricing resulting from the auction pass to the borrower. Competition between banks for assets and desire to protect relationships would likely produce a deal comfortably inside 13bp over Libor. Angry bankers ask why, if Goldman is such a top-flight loan arranger, it could not immediately price the deal at the market level and syndicate it. They were in no mood to lend their balance sheets in the form of Fl320 million underwriting commitments to an investment bank yet to establish a reputation as a leading arranger of finely priced syndicated bank finance.

Market talk suggests that Goldman has proposed this auction structure, without success, to several other European corporate clients. At least one treasurer confides that he rejected the concept out of hand as a cumbersome and expensive way to raise money from banks whose price limits he already knows extremely well.

KPN, as a debutante in the syndicated loan market, clearly did not appreciate all these sensitivities. To date, its main experience had been three bilateral standby facilities worth some Fl750 million in total. The scene was set for a fiery meeting.

According to those who attended, KPN's chief financial officer and treasurer gave a polished performance explaining the benefits of the TNT deal to the KPN group and its shareholders. But the audience was far less impressed when it came to their potential roles in the financing. The atmosphere turned aggressive when KPN officials scolded the banks for the unflattering publicity the deal was receiving.

Commercial banks wanted answers to tough questions on pricing and syndication strategy. The KPN officials turned the platform over to Goldman Sachs, which failed to give comprehensive answers and said it would report back to banks individually later. KPN officials must have looked on in alarm. Goldman was suggesting that eventual pricing should be well below 10bp over Libor. Just what comparable companies did Goldman think they had been lending to at such rates, bankers shot back.

By the end of the week, KPN decided to act decisively to defuse the tension. The 10 banks were told that the structure was being revised. The auction would still take place but the six co-arrangers would not be chosen solely on price. Relationship factors would come into play and additional co-arranger slots might be available if deserved. Perhaps best of all from the banks' point of view, Goldman would no longer be sole arranger. Its position atop the deal would have to be shared with a joint arranger and bookrunner from the 10 commercial banks.

Citibank secured the role of co-arranger, presumably by making the lowest bid. ABN Amro was agent. The three-year loan was priced at 10bp over Libor and bankers say that KPN is probably over-paying. They had expected 8bp over Libor. One banker says: "Goldman united the banks instead of pitching them against each other. KPN should easily have got a single-digit margin." Nigel Pavey


St George advances

Australia's largest corporate merger, New South Wales-based St George Bank's agreed bid for Advance Bank, may herald a wave of consolidation in Australian banking, drawing in international as well as local participants.

The A$2.65 billion ($2.1 billion) deal is the latest and largest in a series of mergers and acquisitions Australian regulators have supported to ensure the big four national banks - National Australia Bank (NAB), Commonwealth Bank of Australia (CBA), Westpac and ANZ - face adequate competition in each state.

The new St George Bank will be the largest regional bank. It will have a bigger share of deposits in South Australia (25%) than its nearest rival, CBA with 18.3%. In New South Wales, it will rank third with 17%, after Westpac (18.9%) and CBA (18.2%).

Overall, it will rank as the country's fifth-largest bank, with A$43 billion of assets, compared to ANZ, smallest of the four national banks with A$71 billion. NAB, the largest, has A$92 billion of assets. In certain market segments, St George clearly stands comparison with the national banks. Its A$16.6 billion of mortgage lending, ranks above ANZ with A$14.9 billion.

The scale of bank mergers is growing. Last year St George wanted to buy the former state-owned BankSA, but lost out to Advance Bank, which paid A$730 million. With this deal, it neatly swallows up its earlier prey.

The combination of the former building societies is a full-blooded in-market merger, predicated on combining the two banks' networks, cutting costs and eliminating jobs. Investment bankers in Australia had been pitching the idea to St George chief executive Jim Sweeney for months but he appeared to resist a deal that may cost 1,000 jobs. The arrival as St George chairman of Frank Conroy, former chief executive of Westpac, pushed the deal forward.

The management target for the new bank is to take out A$140 million of the combined costs before tax, with 80% of that to be realized in two years. Much of this will come from closures among the 169 overlapping branches within the combined banks' 607-branch networks.

Australian bankers are impatiently awaiting the report of the government-sponsored Wallace commission into the future of the country's financial services. This may bring sweeping recommendations. Meanwhile the Australian Competition and Consumer Commission is not shy about its views. Chairman Allan Asher describes the St George/Advance Bank merger as potentially "pro-competitive" and adds that "banks, especially the majors, are shielded from a lot of potential competition because of restrictions in the Bank (shareholding) Act and the Foreign Acquisitions and Takeovers Act".

Bankers in Australia interpret such comments as a warning to NAB, which owns 6.8% of St George, not to make a counter bid for Advance Bank. That raises the question of what NAB will do with its stake.

Before the ink is dry on this merger, bankers are already thinking of future deals. Other takeover candidates include Bank of Melbourne and Colonial Bank. St George may itself be swallowed up. "It would be a logical fit for an international bank interested in expanding into the Asia Pacific region. It now has the critical mass to be the start of a new Australian national bank," says Alastair Walton, director, financial institutions group, at CS First Boston, which advised St George. Its articles of association will prevent a full acquisition for five years, but a potential buyer could take a strategic stake.

European banks might be tempted into Australia as part of a longer-term effort to build up in Asia. It was with this in mind that Bank of Scotland last year acquired 51% of Perth-headquartered BankWest from the government of Western Australia, paying 1.8 times book value. This was the first time an overseas bank has been allowed to acquire an established retail banking franchise in Australia.

Australian regional banks have inherent attractions, being generally well run, with strong returns on equity and transparent financial reporting. Now, with the regional sector apparently being transformed by mergers, bankers wonder whether Bank of Scotland will sit on that investment, add to it or even dispose of it.

Rumours of foreign potential takeovers are rife. There has been talk of a possible combination between Standard Chartered and ANZ, perhaps through some joint shareholding scheme rather than a full takeover, though a spokesman for Standard Chartered says: "There is nothing in it at the moment, we are totally unaware of anything."

High valuations are being applied to regional banks. In mid-1995, Advance paid 1.6 times book value for BankSA. This merger values Advance Bank at 2.1 times book value, the same as Westpac recently paid for Perth-based Challenge Bank.

St George applied some thought to bid financing, which mainly takes the form of St George shares. The bank is offering A$2.10 in cash plus A$0.20 from a special dividend, with the rest of the indicated price of A$7.30 per Advance Bank share made up of St George stock. The final price may fluctuate with St George's share price - A$8.69 at the time of the bid. To protect itself from paying too much, St George worked out a pricing mechanism that reduces the number of shares it offers if its own share price rises, so capping the bid price at A$7.30. If its share price falls, it offers more shares up to a specified limit. Advance Bank shareholders only receive less than A$7.30 if the St George share price falls below A$8. This puts a collar on the bid price, making it effectively a fixed-price offer. Peter Lee


Asset-backed gets up steam

Asset-backed issues finally came of age in the Euromarkets in 1996. Total volume of non-mortgage asset-backed Eurobonds will comfortably top $20 billion in 1996, up from $2.6 billion in 1994 and $3.8 billion in 1995. The UK government can indirectly claim much of the credit.

When the government privatized British Rail, three companies, known as Roscos, were set up to own, maintain and lease the trains; one, Angel, was bought by the GRS consortium, which includes Babcock & Brown and Nomura. The current leases with train operating companies (TOCs) run for between seven and 10 years from April 1994, and are 80% guaranteed by the government. They are operating leases, split into a maintenance element and a capital element.

In January, Nomura raised £550 million towards a £700 million total acquisition price by securitizing the government-guaranteed part of the capital element of Angel's leases.

Aimed at asset swap investors, the securitization took the apparently unwieldy form of 98 zero-coupon bonds: one for each payment that Angel would receive from the TOCs. This structure took full advantage of the positive yield curve: Guy Hands, head of Nomura's principal finance group, estimates that by hugging the curve so closely Angel saved about five basis points compared with the cost of a quarterly obligation. This, he says, offsets any reduction in liquidity compared with a more conventional securitized floating-rate bond.

It is difficult to imagine a less liquid structure than this. But banks were attracted by the government-guaranteed bonds because they can take full advantage of the 20% risk weighting. "Banks prefer to do their own internal swaps," says Hands, explaining why Nomura did not swap the bonds into floating-rate notes itself. "They offset the bond swap against the rest of their positions and often come out better off after internal risk accounting."

This is fine if you're offered zeros and you're a bank. But a less well rated investor unable to do an internal swap would end up having to pay a lot of money for swap counterparty risk.

One banker calls it "the stupidest way imaginable: it's almost as though they set out to maximize illiquidity and swap counterparty risk. On top of that, because it was structured as a series of zeros, they had to be completely inflexible in the way they allowed Angel to structure their leases."

Yet demand was strong. The yield was 15bp over Libor - a big saving on 100bp bank loans. There were few asset-backed benchmarks to compare it with. Even so, bankers asked whether Nomura might not have got a bit closer to the government's own cost of funds - between 15bp and 20bp below Libor.

Spurred on by its success, Nomura decided to attempt what had been unthinkable a few months earlier: the securitization of the 20% of the leases that did not carry a government guarantee. This involved getting shadow credit ratings for the TOCs. Two of the four bonds, together worth £130 million, had triple-A ratings, and came at less than 20bp over Libor. The other two were of £25 million each. One was triple-B, at about Libor plus 100bp; the final double-B tranche (effectively a first-loss tranche, as it was lower than the shadow credit ratings of the TOCs paying the leases) was 300bp over. In all, they saved Angel an extra £20 million.

At about the same time as Nomura's second securitization, British bus company Stagecoach won the franchise for South West Trains, the first TOC to be privatized. From its beginnings as a ruthless Perth-based bus company, Stagecoach has in less than a year become a highly successful international passenger transport business with an eye for expansion through cheap acquisitions. Rather than bid ever-higher prices for TOCs, it started looking abroad - and at Roscos.

It bought Porterbrook, the most aggressive of the Roscos, in August 1996, knowing that Angel's experience showed the securitization could be done.

Of the £825 million Stagecoach paid for Porterbrook, £520 million came in the form of a bridging loan from Union Bank of Switzerland (UBS), to be refunded with a securitization. Just £77 million was paid in cash; the rest came through a very advantageous rights issue.

UBS had long been Stagecoach's stockbroker, and the two had built up a strong relationship since UBS floated the company in 1993. "We used UBS because it was a one-stop shop. It made sense to keep everything in the same house," says Stagecoach's finance director Keith Cochrane.

Nomura got a lot of very positive press for its securitization, and had privately hinted that no-one else was capable of doing such a thing. UBS picked up Nomura's gauntlet by saying in Stagecoach's acquisition prospectus that it was "highly confident that it will be able to complete the debt securitization". More accurately, it snatched the deal from Goldman Sachs, which had been mandated by Porterbrook for the securitization before Stagecoach came along.

UBS's deal, when it came, was impressive. Nomura had spent seven months putting its deal together. UBS did it in less than two. Nomura started with 80% of the receivables; UBS did 100% in one go. It raised £545 million.

The UBS deal is also a lot more elegant than Nomura's 98 zeros. UBS didn't quite achieve the perfect structure of one floating-rate note for the government-guaranteed 80%. It did it with five of differing maturities. The swap was performed in-house. The maturities range in price from 4bp to 14bp over one-month Libor. There are also two bonds for the other 20% of the receivables: a triple-A tranche at 16bp over and a triple-B tranche at 50bp over. This is a lot better than Angel's funding.

But there is still argument over whether Nomura has come out in front of UBS in the securitization wars. Angel certainly paid quite a bit more than Stagecoach for its funds. And UBS's floaters are prettier than Nomura's zeros. But, admits Keith Ballantine, assistant director and part of the securitization team at UBS, "asset-backed spreads have come in a lot over the past 12 months". A series of high-profile issues, such as Morgan Stanley's $4 billion GPA airplane refinancing and the Ffr40 billion Crédit Lyonnais deal, have awakened a previously dormant appetite for securitizations. And just before the Stagecoach deal, Nomura brought Annington Homes - another UK government-backed issue - to the market at 6.25bp over, tightening to 3.5bp over in a relatively
illiquid secondary market.

Given that the Annington Homes deal had a five-year
average life, compared with just over four for Stagecoach, it seems Nomura has proved itself once again. (There are quibbles about the effect of paying off three-month Libor as opposed to one-month Libor, and paying monthly rather than quarterly, but not even UBS suggests that Stagecoach got a better deal than Annington Homes.)

With the experience of the Angel funding under its belt, Nomura felt that it could go out beyond 10 years with Annington and tap fixed-rate demand from UK pension funds. Here, the spread is over gilts, not Libor, and can work out cheaper than seven- or eight-year floating-rate funds.

The dip in the cost of funds over 10 years "is completely counter-intuitive", says Nomura's Hands. "But a pension fund needs to match its [long-term] liabilities, so from its perspective, short-term assets have greater risk than long-term ones. At the time of Angel, we tapped the banks, because they're deep and liquid. But if I were doing a Rosco deal tomorrow, I would be tempted to issue one long fixed-rate bond and sell to the pension funds and insurance companies."
Felix Salmon


Citicorp's new constellation

Even the sharpest-eyed MTN dealer might have missed a L20 billion ($13.12 million) one-year zero coupon issue at the end of September. It was unlisted, not syndicated and apparently unremarkable. But dealers will very soon become familiar with the borrower. It will be one of the biggest and best in the market.

Centauri Corp has just signed $20 billion worth of debt programmes. It will raise funds in four markets, issuing commercial paper and MTNs in the US and European markets through four programmes, each with a ceiling of $5 billion.

Centauri Corp is the latest structured investment company to be designed and managed by a 25-strong team at Citibank Credit Structures (CCS). Since the late 1980s, this team has managed the funding and debt investments of two other specialist companies, Alpha and Beta, often called credit arbitrage vehicles. The companies are designed to maintain triple-A long-term debt ratings from both major agencies and produce returns to equity investors by locking in a credit spread between borrowings and investments in a range of instruments. These are mainly asset-backed debt, and issues of financial institutions, governments and a few corporates.

Centauri, like its predecessors, relies for its triple-A rating on its layer of equity, which may rise to $1 billion; on a hedging strategy designed to eliminate interest rate and currency risk, and on deft use of funding markets, committed credit lines and liquidity reserves to minimize refinancing risk. It will borrow short-term and invest in a mix of longer-term debt. That maturity mismatch will contribute a fair portion of earnings.

The CCS team is regarded as among the best users of the MTN markets, praised for its speed of response and willingness to buy back its own paper. That helps it to achieve aggressive funding of below Limean on the typically short-dated debt that the investment companies issue.

For Alpha and Beta, average life of assets is between four-and-a-half and five years. Borrowings range from three months to six years with an average life of one-and-a-half years.

On the investment side, credit rating guidelines are strict. Fully 70% of Centauri's investments will be rated Aa/AA or higher and at least 50% of its investment portfolio will comprise Aaa/AAA rated debt. Assets are swapped into floating-rate dollars.

The sternest challenge for managers of the new fund will be finding investment opportunities. Credit spreads between different rated bonds have narrowed in recent months, making the search for yield more challenging. Fund managers have found good value in securitized debt, particularly credit card receivables. But returns are shrinking. Two or three years ago, spreads on credit card deals were between 40bp and 45bp over the risk-free rate. Now they are less than 20bp.

Clever investors revelled in the confusion that followed the unexpected rise in interest rates at the start of 1994. Many investment and commercial bank trading desks mispriced bonds in the months of panic that followed, allowing fund managers to pick up attractive assets. The Beta fund bought large volumes of busted structured notes - typically collared notes issued by banks that had been designed to sell in a falling interest rate market and which now underperformed badly. CCS unwound the interest-rate caps and floors embedded in these notes and found itself holding bonds of well-rated issuers yielding twice what they might pay on plain vanilla debt.

Now Charles Covell, managing director at CCS, says: "It is hard to see credit spreads getting any tighter from here, but we and the markets have been saying that for two years. We must spend more and more time thinking very hard about how to source pockets of good value."

Centauri has been designed to cope with this tougher market. It uses new methods of measuring portfolio configuration risk and allocating capital against risk. For example, Beta uses a weighted average of the credit quality of bonds in its investment portfolio to calculate a leverage ratio. An average quality of AA+ dictates it can borrow no more than 7.5 times its equity capital. But that approach has its limitations.

Centauri, by improving its portfolio risk measurement, can often leverage more and take larger positions than Alpha or Beta could in the same investments. So, even if credit spreads remain obstinately narrow, it should be able to improve earnings. The trick, according to Covell is that "when we see value, we take it early and we take it in size".

CCS staff have spent much time with rating agency analysts exploring the effects of portfolio diversification. When fund managers find value in a sector of the capital markets, they can derive an additional benefit by spreading investments over more issues, more names, more industry sectors. By reducing portfolio configuration risk, Centauri can increase its leverage.

For Beta leverage is capped at 10 times the capital allocated against an investment. Centauri can sometimes leverage up to 25 times. As Covell stresses, "a portfolio can be 20-times leveraged, but still not be as risky as an eight-times leveraged portfolio of inherently riskier assets".

Centauri should also benefit from CCS' strong relationships with leading investment banks. Investment banks in the asset-backed new issue business have been grateful for the Alpha and Beta funds. They stood behind the growth of seven- and 10-year floating-rate credit-backed securities last year, allowing issuers to launch deals that were essentially underwritten. CCS has also bought large private placements from issuers at times when these had greater difficulty borrowing in the public debt markets.

CCS has a list of 20 investment banks, transacting most of its business with the top half dozen. "We cover these suppliers of assets, almost as one would clients. We make it abundantly clear what we want. We don't leave it to them to guess. We want our phone to ring before anyone else's."

Equity investors in Centauri are typically financial institutions seeking a floating-rate Libor-based return. Many are themselves funded investors, subscribing largely with borrowed money. They enjoy the benefits of CCS' standing among MTN dealers and its size and muscle with banks' trading and sales desks, but they do not have to consolidate their investments. Citicorp earns fees for managing the investment companies. It is not an investor itself, being fearful that US banking regulators might classify funds as Citicorp balance sheet risk. Most investors are Citicorp clients, some no doubt hoping to learn how to manage similar operations for themselves. Peter Lee


Votes in custody

At last, there is a real prospect of reform for Germany's ineffectual system of corporate governance. In legal amendments drawn up by a parliamentary committee, the most interesting proposal is to prohibit banks from using their custody clients' voting rights if they already own more than 5% of the company.

The committee responded to retail investors' disquiet that their shares were being abused to bolster the "power of the banks", a perenially emotive subject.

Losing their clients' voting rights would indeed be a blow for powerful corporate shareholders like the big private-sector German banks. They have substantial shareholdings in a number of companies and have traditionally augmented that influence by exercising their clients' voting rights as well. Deutsche Bank's shareholding of at least 5.2% in Bayerische Vereinsbank, for instance, tells only half the story. Add to that a couple of percent worth of voting rights culled from institutional and retail clients, plus several more percent owned by Deutsche Bank group companies such as fund manager DWS and insurer Deutsche Herold.

Vereinsbank is conducting a survey of other banks to try to find out the extent of these custody voting rights, but is still waiting for replies. But it seems a fair assumption that Deutsche Bank already controls at least 10%, and perhaps 15%, of its Bavarian competitor.

Formally, Deutsche Bank would need far more than that to gain control of Vereinsbank. But corporate shareholders rarely try to use naked voting power to press a point; influence is usually expressed in negotiations with management and at supervisory board meetings. Perceived clout is mostly more decisive than actual voting strength.

Only the largest institutional investors in Germany bother to attend annual general meetings of DAX companies. Most smaller - and foreign - investors not only entrust their votes to a custodian bank, but let the bank decide how to vote. While some retail investors use shareholder clubs to exercise their votes, institutional investors have no time-saving alternative; most are content to trust their banks.

The possible loss of that trust concerns custodians. Deutsche Bank is the largest custodian bank in Germany and has grown to the fourth-largest global custodian in the world over the past couple of years. Peter Grafunder, global head of international custody services, says it is time-consuming and expensive to provide a free voting service for clients, but concedes that the bank must do so before it can offer more lucrative value-added custody. Grafunder admits that if the bank were unable to offer basic voting representation to a client, the entire account might be removed to a rival custodian.

One solution, Grafunder says, might be to establish independent voting companies. That would enable banks with major corporate shareholdings to shed the responsibility of voting, but preserve their custody business. Banking lawyers believe it would be out of the question for banks to own voting companies, since that would preserve their control over clients' votes. But it is difficult to see who else might offer that service. There appears to be no interest from major institutional investors. Law firms would be unable to market a voting service properly while they are banned from advertising and publicising their activities.

Another proposal might be more feasible. The annual convention of German lawyers recently passed a motion that banks should be obliged to disclose any voting rights in excess of 5% which they exercise on behalf of clients. Banks would still have the power, but it would at least be public knowledge.

Banks now worry that they will lose influence from early 1998, when the amendments are scheduled to become law. They could accelerate the restructuring and mergers that are now regarded as inevitable. One banker attacks the amendment as "a back-door attempt to oblige the banks to reduce their participations". German banks would pay over 60% in capital gains tax if they sold their corporate participations, so banking lawyers suggest it would be anti-constitutional effectively to force them to do so. Laura Covill