Volcker Rule Might Make Losing Money Illegal

There's one sure way to know if a bank was risking a lot of money.

Destroyer of brunch plans.

Photographer: David Paul Morris/Bloomberg

This is the saddest story I have ever heard:

A rare Sunday e-mail from Blackstone Group LP invited banks to submit pitches to underwrite 90 million shares in hotel-chain Hilton Worldwide Holdings Inc. that the private equity-firm needed to unload fast -- a $2.7 billion stake.

Dozens of bankers abandoned their mothers and families to cobble together due diligence and propose prices. Bank of America Corp., Citigroup Inc. and Deutsche Bank AG landed the deal, buying 30 million shares each from Blackstone at $29.71 to sell the next day at $29.85. Trouble was, the market didn’t cooperate.

Banking is a tough business, but I must say that it looks like gratuitous cruelty for Blackstone to send its banks a multi-billion-dollar competitive bid for a block trade on Mother's Day. Did they not want to sell the shares on Saturday? Come on. But that's the nature of the business: You do whatever you have to to make a buck, even if it means abandoning your mother mid-brunch. What makes this particular story so sad is that the banks may have lost 90 million bucks between them: The block was hung, and the stock price dropped below the deal price before the banks could sell out of their positions. You could have saved your desk $30 million by just not checking e-mail at brunch. I hope there's a banker at Goldman Sachs or JPMorgan who did, and who will demand credit for that $30 million in his bonus.

This all happened in May, but is back in the news now because of something that is happening this week:

Much of the Volcker Rule came into force Tuesday, limiting how banks wager their money while letting them continue facilitating clients’ transactions. Handling a block of stock should be fine, as long as banks are conservative enough that the government doesn’t deem it a proprietary bet, said James Angel, a professor of finance at Georgetown University in Washington.

“They run a slight regulatory risk that a year from now some regulator is going to look at the trade and raise questions,” he said. With Volcker so new, “we’re entering a period of rulemaking by enforcement, in which regulators will look at stuff, give it the smell test and say, ‘That’s OK,’ or give it a sniff and say, ‘That’s not OK.’”

Is that true? It initially struck me as a little crazy. This trade is not anywhere near the boundaries of the Volcker Rule. Underwriting stock offerings is a core investment banking function that is excluded from the Volcker Rule's restriction on proprietary trading. This was a big block that went poorly, but it was also a regular-way block in competition for a $28 billion household-name company and a blue-chip private equity firm. If this deal isn't allowed, then no stock underwriting is allowed.

One thing that I, and lots of other people, often say is that the Volcker Rule doesn't do much to reduce banks' risks. It forbids bank from doing "proprietary trading," which does not  mean "making risky bets with a bank's own capital." Most of what banks do is make risky bets with their own capital. Lending is a risky bet with a bank's own capital. 1 "Proprietary trading" is a specific thing; loosely speaking it means making bets that are independent of customer demands or hedging requirements. 2 If you lend money to a company, that's not proprietary trading. If a hedge fund customer calls you up and sells you a bond, that's not proprietary trading. If you go to a salesman and ask to buy a bond, because you just like that bond, that's proprietary trading. The risks are identical in each case; the attitude, and the context, are different. 3

Here, Blackstone's banks bet $2.7 billion of their own money on Hilton stock, and may have lost $90 million. But it's fine because the bet was made in their underwriting businesses, and the Volcker Rule's prohibition on proprietary trading "does not apply to a banking entity's underwriting activities." 4  So it's categorically allowed.

Sort of! There are exceptions to that exception. Those underwriting activities must be "conducted in accordance with paragraph (a)" of the underwriting exemption, which includes, along with various definitions and documentation requirements, this limitation:

The amount and type of the securities in the trading desk’s underwriting position are designed not to exceed the reasonably expected near term demands of clients, customers, or counterparties, and reasonable efforts are made to sell or otherwise reduce the underwriting position within a reasonable period, taking into account the liquidity, maturity, and depth of the market for the relevant type of security;

Was this deal "designed not to exceed the reasonably expected near term demands of clients, customers, or counterparties"? I mean, surely it was, right? It was bid out to a bunch of banks, and three big ones ultimately split the deal; it would be hard to argue that none of them reasonably expected to be able to sell it. They overshot a bit, but as a matter of reasonableness there's safety in numbers.

Most of the reactions to the Volcker Rule finally going into effect are pretty blasé. There's been a long time to prepare:

The big banks that fought for years to change the rule have for the most part already fallen in line. Firms such as Citigroup Inc., Bank of America Corp., Morgan Stanley and Goldman Sachs Group Inc. have shed their proprietary-trading desks, pulled money from certain investment funds and ceased other activities that would run afoul of the rule’s restrictions.

If you think -- as I mostly do -- that the Volcker Rule is a categorical rule, then that's pretty much that. The rule told banks to stop doing proprietary trading. Banks stopped doing proprietary trading. It has had its effect. 5  No one need ever speak of it again. Maybe the treasurer at some small bank somewhere will get in trouble down the line for day-trading stocks with his bank's money, but the big heavily supervised banks have shut down their prop trading businesses and won't start them up again.

But there's an alternative version of the story in which the Volcker Rule leads to constant second-guessing at the margins of allowed activities. 6 The Volcker Rule allows underwriting, and market-making, and hedging. All of those are risky bets with a bank's own money, all of them are conceptually distinguishable from proprietary trading, but all of them are subject to fuzzy limitations and extensive documentation and compliance requirements. How do you know that your market-making is "designed not to exceed, on an ongoing basis, the reasonably expected near term demands of clients, customers, or counterparties," or that your hedging "does not give rise, at the inception of the hedge, to any significant new or additional risk that is not itself hedged contemporaneously"? 7  I mean obviously you can decide those things to your own satisfaction, and there is some guidance in the rule, but you cannot be absolutely sure what regulators think until they point to what's not allowed. "Rulemaking by enforcement" does seem to be the future.

But regulators can't be sure what to think either. Why would a regulator know better than a bank how many bonds are necessary to satisfy expected customer demand, or what risks a hedge gives rise to, or for that matter what written policies and procedures are sufficient to limit the risks of underwriting and market-making and hedging? There's no obvious signpost for which activities are reasonable and which aren't.

Except one! If it loses a lot of money, it was probably risky! I mean, that is not great logic, but it is good emotion. It is easy to declare in hindsight that, if a bank lost a lot of money, it was running a risk of losing a lot of money. And that it had inadequate policies and procedures to prevent losing a lot of money. 

And regulators have incentives to be seen to be enforcing the Volcker Rule. The Volcker Rule's ban on proprietary trading is just a categorical divestment rule, which won't give rise to much enforcement: It mandated divestment, and the divestment happened, the end. But the Volcker Rule in operation will not be a rule against proprietary trading; it will be a rule about risk management in underwriting, market-making and hedging. Banks that manage risks poorly -- as evidenced, I suspect, mostly by losing money -- will face enforcement actions, so that regulators can prove they're using their new tools to reduce bank risks.

The result could be a particularly comical form of bank enforcement, in which banks get fined for losing money. JPMorgan was fined $920 million for the London Whale's losses; if the Volcker Rule had been in effect, I suspect that would have been a lot more. The next bank that loses money on a badly-designed hedge, or a careless market-making desk, or even a hung block trade, 8 might also be fined for doing so.

Will that reduce risks? Sure, why not. It's just weird. We're worried about banks recklessly losing money, so if they do recklessly lose money, we'll take away more money. Doesn't that ... make things worse? And maybe taking away the money will be a deterrent to the reckless losses -- but in that case, wouldn't the losses themselves be a deterrent? Don't get me wrong: I can see how this might, at the margin, make banks more careful, 9  and in any case I look forward to making fun of the first bank to be fined for losing money in violation of Volcker. It really might make banks safer. But it's a goofy way to get there.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

  1. Here "capital" means, like, "money," not "regulatory capital."

  2. The most conceptually useful definition that I've heard is that in customer facilitation you get paid the spread; in proprietary trading you pay the spread. That's not a rigid definition -- market makers often cross the spread to get out of positions, etc. -- but if you're always paying the spread that sure looks proprietary. 

  3. In particular, the context of prop trading is that you might make a series of same-way directional bets that you hold for a while, while the context of market-making and underwriting is that you want to end up flat as soon as you reasonable can. But of course lending is just one big directional bet, and one weirdness of the Volcker Rule is that it doesn't limit long-term risk-taking: Proprietary trading is fine if you're holding positions for more than 60 days.

  4. That's section 4 of the Volcker Rule (page 11 of the Davis Polk version).

  5. Similarly for certain private equity and hedge fund businesses, which were similarly restructured or jettisoned.

  6. This is Jamie Dimon's story, where "for every trader, we’re going to have to have a lawyer, compliance officer, a doctor to see what their testosterone levels are, and a shrink."

  7. The market-making limitation is section 4(b)(2)(ii) (page 13). The hedging one is section 5(b)(2)(iii) (page 16).

  8. To be clear I think Hilton is way way way inside the line. But like a block trade for many days' volume for a dodgy company without competitive bidding? I could see someone wondering about reasonably expected customer demand there.

  9. Of course enforcement actions can send a signal to other banks to change their own policies. Though, again, the losses themselves, if they and their causes are observable by other banks, might have the same effect. Banks don't actually want to lose money.

To contact the author on this story:
Matt Levine at mlevine51@bloomberg.net

To contact the editor on this story:
Zara Kessler at zkessler@bloomberg.net

Before it's here, it's on the Bloomberg Terminal.