|
AT&T, ConocoPhillips and AOL: three companies, three loans, |
NO BANK EXPECTS to get paid 700 basis points for taking on a chunk of an investment-grade loan. Yet that is precisely what happened for some lucky participants in AT&T’s $4 billion 364-day revolving credit completed at the start of October.
It wasn’t planned that way. It started out as a bog-standard deal for a relatively solid triple-B credit. Upfront fees were 100bp over Libor, and only then for the four lead arrangers that had committed $550 million each at the start of the process. That had been increased from 87.5bp over after rumours of AT&T being downgraded a notch. The seven other banks acting as co-arrangers were paid 75bp or 62.5 bp over Libor depending on whether they committed $300 million or $200 million.
So far so good. But then at the start of September the general syndication process began, used by the 11 lead and co-arrangers to sell down their initial positions to a broader group of banks, known as the retail portion of the syndicate. In that four-week period before the syndication was closed, however, some members of the syndicate established a grey secondary market and were offering the loan for as low as 95 cents on the dollar, according to some market participants, which would be hundreds of basis points through fees and thus lose the sellers money.
They couldn’t sell it at the time, of course; they had to wait for the syndicate to break. But it’s a simple enough exercise: those determined to sell part or all of their exposure approach banks that would normally be joining the retail part of the syndicate and offer to sell them a portion of the loan at a much more advantageous price than they would get as an official member.
This can have a marked effect on the syndication process, and clearly did in the AT&T deal. When the company arranged its previous revolver, last December, it managed to raise $8 billion and had 30 banks in the syndicate, of which 13 were retail. This time, however, just five banks joined during the general syndication process, four committing $25 billion each for a fee of 25bp over Libor, and one stumping up $100 million for 50bp over – combined that reduced the official commitment of the four lead arrangers to $512.5 million.
“Persuading the retail banks to forgo taking a position in the official syndication risked screwing up the deal,” says one irate syndicate member. “They created a fire-sale environment rather than an orderly process.”
Even lower later
Once syndicate broke, trades were completed at even lower levels than those indicated in the grey market. “Several trades were verified at 93 cents,” says Bob Wagner, managing director and head of syndicated lending at Goldman Sachs, one of the four lead arrangers of the revolver. Many more were apparently struck at 94 and 95 cents, while some contend that they heard of trades being done at 92 cents, thus adding anything up to an eight percent bonus to buy paper which otherwise would only have paid 15 basis points should the loan stay undrawn.
|
View graph. |
||||||
The problems with that deal could be attributed to market and company problems: at the end of June Vivendi admitted to accounting fraud, which affected most media firms, and in July AOL was hit by management upheaval. But the AOL loan did manage to attract 27 banks to the deal, of which around 16 could be classed as retail.
Market participants have seized on these two deals as evidence to prove their very different theses about the state and direction of the high-grade syndicated lending market.
To the investment banks the pricing dislocations prove that the commercial banks don’t regard lending as a stand-alone business, rather they underprice the product as a way of forcing their way into investment-banking business. “Pricing isn’t reflective of the risk of lending to investment-grade companies,” says the head of loans at one investment bank. “It’s the last weapon in the arsenal of those banks determined to get into the investment-banking race.”
Meanwhile, commercial banks themselves contend that it’s not the initial pricing of the deal that is causing the problem but the way investment banks choose to play the game. “They have to mark their loan positions to market,” explains Kirchhoff. “They don’t want to hold the paper, so choose to sell it as quickly as possible, and often at much reduced prices. And that can threaten to undermine the syndication function.”
She points to ConocoPhillips’ inaugural post-merger loan in September as a successful syndication. Bank of America was co-lead arranger on the $2.5 billion loan with Barclays Capital; $2 billion was a 364-day, the rest a four-year, revolver. In total 35 banks took part in the $2.5 billion deal, 23 of which could be considered retail. The rating agencies upgraded the newly merged company from triple B to low single A just as the syndication started, and the pricing reflects that: upfront fees ranged from just 9bp to 11bp depending on the commitments, which ranged from $25 million to $125 million. Undrawn, the 364-day revolver pays out 8bp, 25bp if termed out, and 37.5bp if drawn. There were, Kirchhoff points out, no pure investment banks playing any senior roles: Merrill Lynch and Morgan Stanley both took part in the retail tranche, committing $47 million each.
One significant difference between this deal and those for AOL and AT&T, though, is that the pricing, although low compared with other capital markets instruments, was hardly out of whack. The default swap price for ConocoPhillips paper was trading between 60bp and 70bp during the course of the loan being marketed.
That was not the case for the other two where the credit price in the default swap market differed markedly from the loan price.
At the end of June, reacting to Vivendi’s troubles, AOL’s default swap widened by about 250bp within a couple of days to a bid-offer of 590bp-630bp. It then tightened quickly. On the day it finalized its line of credit the default swap was 415bp-445bp, before widening out again after the mid-July management upheaval.
Hedging hits secondary trading prices
AT&T’s public bonds maturing in 2004 and 2006 offer yields of between 6.7 and 6.9%, while the bid-offer price for a default swap on AT&T ranged between 400bp and 480bp between mid-September and mid-October. That’s far removed from the interest payable on the revolver: 15bp over Libor if the revolver remains unused, 110 bp over if AT&T does draw it down. Even adding on the maximum fee of 100bp for the largest lenders doesn’t make a compelling case for the loan over the bond or the default swap on a stand-alone basis. So any moves by syndicate members to sell loans at much-reduced prices would appear to be little more than a rational response to selling a financial asset – in this case a loan participation – that appears to be very expensively priced in the context of the wider market. That those sellers are almost always the investment banks is a reflection of the greater sensitivity of their business model compared to other lenders’. They have to mark to market and argue that banks should have to as well.
The commercial banks are hardly keen to hold mispriced loan exposure. They might not be such heavy sellers in the grey and secondary loan markets but they are large buyers of protection in the credit default market.
At least one investment bank and one European player show corporate treasurers charts depicting the effect of such banks’ hedging practices because of loan deals on their companies’ default swap and benchmark bond paper. One has charts showing how spreads gap out in both markets during the month or so of marketing the loan. Cox Communications’ benchmark 10-year bond, for example, widened from about 260bp to over 400bp over treasuries in the space of four weeks in June; El Paso’s from around 230bp to about 340bp over from mid-March to mid-April; and Sprint’s five-year default swap from just over 800bp to over 1400bp in July.
It raises the issue that borrowers can no longer obtain a benefit from cheap bank credit. New supply of credit coming onto the market pushes out their default swap and traded bond spreads.
It’s a difficult trend to track, and these measures can be unreliable. Sprint’s spreads, for example, were hit by the general credit and equity malaise at the end of July. They were also hit by market rumours about liquidity and ratings that proved baseless. Its spreads might have moved as much even if it wasn’t marketing a loan.
But it is certainly true that the commercial banks favour using credit derivatives to offload risk. And they can do it very effectively: the small exposure Citi had to Enron and JPMorgan had to WorldCom are two cases in point. It doesn’t always work, of course, as JPMorgan’s recent loan losses prove. But in general it’s a sign of the increasing power of the credit and loan portfolio managers inside the banks as executives concentrate on improving value to shareholders. At the senior level both JPMorgan and Bank of America proudly tell analysts and investors that they are decreasing the size of their commercial and industrial loan books, whether by using default swaps, by selling loans in the secondary market, or simply by pulling back from providing credit.
Buying default swap protection on a loan isn’t cheap. There’s a huge mismatch between getting 100bp as a fee and 15bp for an undrawn commitment for AT&T on the one hand and paying for protection through default swaps in the mid 400bps over on the other. Investment banks might be hurting the syndicate by prearranging sales in the grey market, but the commercial banks are also confirming that on a stand-alone basis loans are mispriced by buying protection at much wider spreads.
Underpriced for relationships
Neither side is doing this for the sake of it. Both are basing their decisions on a relationship with clients broader than just the loan. Goldman Sachs and Morgan Stanley, as the two largest investment banks in the AT&T loan, stand accused of being the ones to offer the cut-price deals in the grey market. Goldman denies offering the loans at a discount. Wagner says: “The AT&T deal was structured to market convention. Unfortunately, market convention often produces pro rata portions at a deep discount to par.” Morgan Stanley declined to comment. One head of lending at an investment bank does admit, though, that “we might take a hit on loans such as these. But we do so in the hope that we get ancillary business.”
Goldman already has a role to play in a capital-markets transaction for the company, running the tender offer for the spin-off of its cable division. Once that’s done, it’s difficult to see what other capital-markets business the rump AT&T might be able to offer. And that could be the key to the lack of participation in the loan deal, says one loans originator. “Quite frankly, after it divests the cable business, all AT&T will be is a long-distance phone company. It’s almost liquidating itself. If your motivation as a lender is to build or maintain a relationship, what’s left to chase?”
AOL is a different matter. Some banks might not be happy with the pricing for the loan, but AOL is a huge company in turmoil with many different financial needs. Some might prefer just to get the higher spread buying the loan at grey-market prices, but there were 27 banks who thought there might be enough other business around to make it worthwhile committing to a cheap loan.
That is even more marked for ConocoPhillips. “The company does a good job of spreading around the different types of business it does in the financial markets, be it foreign exchange, treasury services or capital markets products,” says Kirchhoff.
That doesn’t necessarily mean that a bank engages in tying a corporate loan to other business – threatening withdrawal of bank credit if other business isn’t forthcoming – an illegal business practice in the US that John Dingell in the House of Representatives has started to investigate. It simply means banks view the loan as part of an overall relationship. It’s a fine line but there’s as much evidence that corporates are doing the tying, demanding a cheap loan as recompense for giving a bond mandate – and that isn’t illegal.
What these three deals prove, though, is that loans are not stand-alone products. Instead they are the cheap, usually loss-leading, part of an overall relationship. The rational approach would be to price the deals to reflect the risk being taken rather than the extra business each bank might hope to get. That would ensure that syndications were more stable all the time and save the commercial banks from taking a bath on having to hedge their positions in the more realistically priced default swap market. It might also help establish a more liquid secondary market for loans.
The investment banks contend that it’s up to the big three to move the market to more rational pricing. “They managed to do so after the Russia crisis back in 1998,” contends one investment banker. “Chase led the way by instituting flex pricing, and the others followed. They could change the pricing parameters again, if they wanted to.” Between them JPMorgan, Bank of America and Citigroup account for 75% of US investment-grade loans. They claim, though, that they are not such kings of the market that they can dictate pricing to the syndicate. “It’s almost impossible for us to increase pricing just to suit us,” says Kirchhoff. “Pricing is dictated by what the rest of the syndicate is willing to bear, and what’s important to investors is what makes sense for their relationships with the issuer, not what might be best for the big arranger banks. If one of us big three were to tell a solid single-A lender that we were only going to arrange a deal with undrawn at L+50bp and drawn at L+75bp they’d laugh and go to one of the other two because they’d know they’d get cheaper pricing and probably get the deal done.”
As this sluggish recovery continues it’s likely that more controversial loan syndications will come to market, and more insults will fly. That’s likely to attract the attention of representative Dingell. In the absence of any deep throats in US corporations, we’ll leave it up to him to decide if there’s been any wrongdoing in the market.
