The Wall Street Journal today discovered that universal banks that lend money to companies for cheap tend to want investment banking business in return for that lending and I guess that's a scandal:

As the market for technology IPOs revs up and the biggest banks seek to capitalize on the size of their balance sheets, the practice of selecting underwriters that also provided loans is coming under focus, spurred by Facebook's IPO process.

Critics of the practice say the choices aren't accidental and reflect the "you-scratch-my-back-I-scratch-yours" way that Wall Street works.

Bankers, for their part, say they aren't allowed to make loans on the condition that they receive other business, but borrowers can use the loans as a factor in choosing underwriters. Some bankers say that lending is just one of the many services they offer companies.

At Facebook, the credit line played a role in the batting order for underwriters, said a banker who worked on an underwriting pitch to the company.

When I was young and naive and pitching for underwriting business against banks that did lots of lending, I always thought that banks "aren't allowed to make loans on the condition that they receive other business, but borrowers can use the loans as a factor in choosing underwriters" thing was ripe for a scandal. I still sort of think that: I just do not believe that no client coverage banker has ever said "we'll be in your credit facility but only if you promise us underwriting or M&A business." (Some people agree with me!) And, as the Journal notes, that would be a criminal violation of the antitrust laws, which is unspeakably weird but there you go.

But if you ask a banker who has been carefully and recently briefed on anti-tying regulations, he will probably tell you something like "we don't demand underwriting business to provide a loan. Companies demand loans to get underwriting business." And, as the Journal says, that's not illegal.

The differences are subtle, and it can be hard to tell which version is right, but over time (and client meetings) I became increasingly convinced that the banks' version is - that they don't hold up clients for advisory business, but rather clients hold them up for loans. The OCC agrees, finding not much evidence of market power in lending. More convincing, perhaps, is the example of the Facebook economics. From the Journal again:

At Facebook, the five banks involved in its five-year credit facility, which was later boosted to $2.5 billion, are Morgan Stanley, J.P. Morgan, Goldman Sachs, Bank of America Corp.'s Merrill Lynch and Barclays Capital, a unit of Barclays PLC. Those banks are now underwriting Facebook's impending IPO, along with a sixth bank, Allen & Co., which doesn't lend.

In its IPO filing, Facebook didn't say how much it paid for the credit line. But it said the credit line was partly responsible for the nearly doubling of its interest costs, to $42 million from $22 million, during 2011.

Actually Facebook did say:

In 2011, we entered into an agreement for an unsecured five-year revolving credit facility that allows us to borrow up to $2,500 million, with interest payable on borrowed amounts set at the three-month London Interbank Offered Rate (LIBOR) plus 1.0%. No amounts were drawn down under this agreement as of December 31, 2011. We paid origination fees at closing and these fees are amortized over the remaining term of the credit facility. We also pay a commitment fee at 0.15% per annum on the daily undrawn balance.

So basically the banks are divvying up $3.75mm a year for Facebook's undrawn loan, against which they have to have let's say $100mm of capital.* How much does capital cost? I dunno, JPMorgan's forward PE is ballpark 8x and its 2011 ROE was about 15% so figure that $100mm of capital costs $12-$15mm a year. That's just capital cost; it ignores the economic question of whether 15bps a year is a fair price for Facebook to pay for a thing that is kind of 5-year CDS on itself. I see (AAA rated) Microsoft CDS bid at around 32bps, suggesting that the Facebook fees are underpriced by at least $4mm.

But on capital alone, the banks are losing in aggregate $10mm or so a year on their Facebook credit facility. Some of that they got back in origination fees, which are not disclosed, but Facebook does note that its $250mm facility had about $1mm in up-front costs so assuming proportionately similar (probably too high if anything) the banks got $10mm up front, meaning that they're still losing something like $8mm a year on the 5-year facility.

Losing eight million dollars isn't cool. You know what is cool? Doing the Facebook IPO. But that credit agreement is a year old and, while everyone was expecting a near-term and enormous IPO by then, there was no guarantee - and plenty of pitching between the February 2011 credit agreement and the February 2012 filing. (And, of course, plenty of opportunities for those not in the top tier of the credit facility to improve and/or ruin their chances by, I don't know, managing and investing in and/or screwing up a private placement.)

Basically if big universal banks were designing a system to improve their underwriting fees and league table positions, it's hard to imagine that this is the system they'd come up with: losing predictable, medium-sized piles of money on relationship lending in exchange for an unspoken agreement to one day have some unspecified involvement in a lucrative deal that may or may not happen at some unknown time in the future. This actually looks more like clients being smart about their ability to get underpriced loans out of their banks by holding out the prospect of future, profitable business.

Anyway. Steven Davidoff yesterday waxed hopeful that the Volcker Rule might cause a thousand flowers to bloom in the business of providing market making services outside of the big universal banks, which have now swallowed all of the old-line full-service investment banks. I am more skeptical. But it's interesting to note that today's news from the thousand-investment-banking-flowers front is not about market making:

A group of senior Morgan Stanley managing directors, each with an average 25 years at the bank, has broken away to form their own advisory boutique in a move that underscores the difficulties large securities firms are facing in dealing with clients, staff and regulators.

The new boutique – Dean Bradley Osborne or DBO Partners – plans to hire aggressively, not only from Morgan Stanley but from other Wall Street banks.

“There is huge dissatisfaction on Wall Street,” said Gordon Dean, a DBO director and former vice-chairman of Morgan Stanley. ... “We will be what Wall Street was 30 years ago, before investment banking became less important and highly leveraged trading more important,” says Mr. Bradley, former global head of financial sponsors at Morgan Stanley. “Clients have learnt to trust individuals more than the institutions.”

These moves are not highly leveraged trading exiting the Volcker-constrained confines of the universal banks and running free as unregulated, un-government-backed market makers. They're fee-earning, capital-light classic investment bankers fleeing from the traders to make a living by selling the sweat of their brow and the smarts of their brain rather than the balance sheet of their banks - y'know, "individuals more than institutions." Good luck to them. I assume they've got an awesome pitch page about how unbiased independent advice is so much better than cheap lending.

Lenders Score Roles in IPOs [WSJ]
Morgan Stanley veterans plan advisory boutique [FT]
Under Volcker, Old Dividing Line in Banks May Return [DealBook]

* Undrawn commitments are counted as risk-weighted assets at a 50% conversion factor (3(b)(2)(ii) here, see also here). Capital is let's just say 8% of RWAs. So 4% of $2.5bn is $100mm.