If you're a true believer in vulgar Volckerism - "banks shouldn't be allowed to make bets with their own money" - then you have an inexhaustible source of things to get mad about, since the only thing banks do is make bets with their own money, for some values of "bets" and "own." Bloomberg's Max Abelson found one today, specifically that Goldman has a group that invests its own money in securities, and it is
a secretive Goldman Sachs group called Multi-Strategy Investing, or MSI. It wagers about $1 billion of the New York-based firm’s own funds on the stocks and bonds of companies, including a mortgage servicer and a cement producer, according to interviews with more than 20 people who worked for and with the group, some as recently as last year. The unit, headed by two 1999 Princeton University classmates, has no clients, the people said.
Multi-Strategy Investing happens to be part of a larger group called the Special Situations Group, which also invests the firm's own money - as Abelson puts it:
That parent group, which uses the firm’s funds to profit from distressed and middle-market companies, has been a major profit center at the bank, sometimes the biggest, former executives told Bloomberg in 2011. Its holdings that year included debt of Melville, New York-based pizza chain Sbarro Inc.1
The phrase "uses the firm's funds to profit from" is exquisite; Goldman would probably say "invests in and lends to." Y'know, like a bank (or a merchant bank). Not that they'd disclaim profiting, but, yeah, investing in and lending to companies is a thing that Goldman does.2
Is this a scandal? The short answer is "naaah." Who cares? What is the Volcker Rule for? Is it to keep banks from gambling with your FDIC-insured deposits? Goldman is barely a real depository bank, and whoever's deposits it's gambling with probably aren't yours.3 Is it to keep systemically important institutions from gambling at all? Abelson makes much of the fact that "Blankfein, 58, said the firm no longer wagers its own money without client interaction," but it is hard to see how client interaction by itself makes a wager less risky. If you're worried about Goldman blowing up and taking the rest of the financial system down with it, worry about Goldman's market making or prime brokerage or collateral transformation or other short-funded levered tetchy systemically important businesses going awry, not about a trader getting a call wrong in a slow-moving investing business that accounts for one-tenth of one percent of Goldman's $949 billion of assets. The London Whale, in a sort-of-similar business, lost six times the entire size of MSI with no noticeable ill effects.
Felix Salmon nonetheless thinks GS should get rid of MSI, saying that "the profits from the group simply can’t be big enough to make it worth the regulatory and reputational bother," but that view of regulatory and reputational bother is, I think, just media vulgar Volckerism. Here are the main categories of what Goldman does:
- make profits by trading for its own account, and
If you were all "let's get rid of the stuff that sidles up anywhere close to the Volcker Rule," you'd get rid of Goldman. I know, I know, tons of people would love that, but few of them are at Goldman. (Some, though.) And if you're going to keep some stuff that looks a little like "prop trading," and you're choosing between (1) medium-term principal investing in securities and (2) market making, why would you choose market making? Because it's a better business, maybe,5 but certainly not because it entails less "reputational bother."
What MSI does is really really clearly allowed by the current draft Volcker Rule: MSI makes medium-term principal investments in securities, holding them for more than 60 days so as to avoid the Volcker Rule's prohibition on short-term trading. It's arguably banking, or merchant banking - investing money in corporations - though it's also arguably not. Paul Volcker thinks it isn't - "I don’t see how it’s comparable to a loan. If you buy an asset to trade it, that’s not a loan." - but his opinion doesn't actually matter. His rule does, and his rule says sixty days.
What Goldman's vanilla flow market making desks do has much more complicated regulatory consequences. Remember when Jamie Dimon whined about how he'd have to check people's testosterone before letting them trade? That was about market making, not long-term principal investing. "We hold investments for at least sixty days" is a simple one-sentence way to get out of the Volcker Rule's clutches.6This is how you comply with the Volcker Rule in your market-making business:
MSI has much much much much much less Volcker-regulatory risk and complexity7 than does Goldman's vanilla market-making business. As for reputational risk, Goldman's reputational risk is at this point essentially constant: someone will write something mean about them every day, so making operational decisions to avoid mean stories would be absurd.
The point ... I mean the point is "naaah," but the other point is that none of this has anything at all to do with whether or not banks should be gambling your deposits on proprietary trading. None of those things are things! Goldman isn't really a bank! It doesn't have your deposits! It's gambling up a storm, in the sense that every night it goes to bed with tons of securities positions on its books, but very very few of them are in this sort-of-scandalous MSI group! And, as a bonus: Wells Fargo (say) also has people who make longer-than-sixty-day investments, and they are called "loan officers," and they're investing a whole lot more than MSI's $1 billion balance sheet.
There's a weird path dependency to the Volcker discussion. One thing that "investment banks" like Goldman have pretty much always done is exactly this: use their market knowledge and deal flow to do merchant banking and medium-term investing in securities. That's sort of what Jefferies is up to in expanding its balance sheet, and it's what Goldman has done for a hundred years. It's not what, say, Wells Fargo has been doing for most of its history, for the simple reason that for much of the last century U.S. law sharply separated (1) investment banking from (2) commercial banking, and merchant banking mostly fell on the investment-banking side of the line. Much of the early enthusiasm for the Volcker Rule was based on a desire to bring back that Glass-Steagall separation: to keep your deposits in boring banks that don't trade, and to keep trading away from deposit funding. There is perhaps some sense in that separation of scary trading from boring banking, but it's a stretch to think that Goldman Sachs will ever fall on the boring banking side of the line, no matter how many prop desks it shuts down.
Secret Goldman Team Sidesteps Volcker After Blankfein Vow [Bloomberg]
Goldman’s small internal hedge fund [Reuters / Felix Salmon]
Let’s talk about Goldman’s prop-trading [FTAV / Tracy Alloway]
Goldman: 1, Volcker: 0 [Reuters / Lauren Tara LaCapra]
1.Though Abelson is unable to determine if any SSG employees ever ate at Sbarro. My money's on no.
2.SSG, in turn, rolls up in Goldman's financial reporting into a group called "Investing & Lending." "Secretive" is one way of characterizing Goldman's principal investing but here is a relevant passage in Goldman's 2011 10-K:
Other principal transactions revenues on the consolidated statements of earnings were $1.51 billion and $6.93 billion for 2011 and 2010, respectively. Results for 2011 included a loss from our investment in the ordinary shares of ICBC and net gains from other investments in equities, primarily in private equity positions, partially offset by losses from public equities. In addition, revenues in other principal transactions included net losses from debt securities and loans, primarily reflecting approximately $1 billion of unrealized losses related to relationship lending activities, including the effect of hedges, partially offset by net gains from other debt securities and loans. Results for 2011 also included revenues related to our consolidated entities held for investment purposes.
3.Mehhhhh. Goldman has some $59 billion of deposits in its insured subsidiaries, most of them seemingly insured (page 26 here), perhaps three-quarters of them "from cash sweep programs" (page 74 here). So I guess that's, like, high-net-worth-investor sort of stuff.
4.That's perhaps slightly exaggerated. Here are Goldman's 2011 revenues:
By my math, $9.3bn of market making + $1.5bn of "other principal transactions" + the bulk of $13bn of interest income are principal revenues of one sort or another, though some of the interest income anyway are from the sorts of vanilla bank-loan products that won't boil anyone's blood. The $13-ish billion from IB, IM, and commissions and fees are let us say not from principal transactions though that's probably not 100% true in, e.g., IB, which includes some lending, at-risk underwriting, corporate derivative, etc. revenues.
5.One dumb method of counting is that "Investing & Lending" made GS $2,142mm in 2011 on $56bn of assets, or a 3.8% ROA. "Institutional Client Services" made $17,280 on $380bn of assets, or a 4.5% ROA, and is less volatile. (Assets page 68, revenues page 2, here.) Felix sort of waves at the likely real reason for keeping SSG-type businesses around, which is that they're cool and attract smart people and give them an outlet for expressing investment theses. If you run a bank you probably do want to have some people around who can make investing decisions rather than just a bunch of ADD-addled flow traders.
6.And even if you sometimes sell faster you can be okay. Here is a good explainer.
7.Or, for that matter, other-regulatory risk. No clients! No KYC! No front-running! It's just you and your Bloomberg, y'know?