The fact that one lovable rogue in London misplaced UBS's bonus pool for the year has people talking again about the Volcker rule, which would ban proprietary trading at banks. I still don't really understand that, and I'm not alone. Here is a thing about the Volcker rule and "Delta desks" (what?):
Yet the definition of what constitutes proprietary trading can be fuzzy. Many on Wall Street consider proprietary trading, or prop trading, to involve only trades made by dedicated traders who are using the bank’s capital and do not have access to client information. The trading done on Delta desks, they contend, is done on behalf of clients.
Those boundaries, however, can blur. A bank may buy a derivative or security from a client in order to make a market, then decide it is worth hanging onto, turning it into a proprietary bet.
The Volcker rule of the Dodd-Frank act is named after Paul A. Volcker, the former Federal Reserve chairman who proposed it. It is intended to prevent American banks from taking on too much risk. The fine print, however, has yet to be worked out, and regulators are debating just how comprehensive to make the definition of proprietary.
This is sort of correct but nicely embodies the conceptual confusion that I suspect lies behind the Volcker rule. Let's spend four hours talking about it, shall we?
Let's say you are at a bank, and you are on a flow desk or, whatever, not a "prop" desk. Let's say it's the delta one desk. And a person comes to you and is like "hi, I would like to be short one million shares of UBS, because that seems like a wise thing to do, but I don't want to mess around with borrowing those shares because all these politicians keep talking about boiling short sellers alive, so just write me a swap where in one year if the stock is below $12, you pay me the difference, and if it's above $12, I pay you the difference," and you're like, yeah, that sounds fair, and so you write the swap.
So you've now got a trade where you're long one million shares of UBS. And you obtained that trade in a normal, market-making, non-"prop" way. Now what do you do?
A thing that you probably do is go borrow one million shares of UBS and sell them short. This is a sensible thing to do. Why is it sensible? Well, for one thing, your desk is called "delta one," and you've written a derivative, and you want to delta hedge that derivative, and so you pull out a Black-Scholes calculator and plug in a lot of variables and get stuck but then you look at your business card and you're like, oh, shit, the answer is "one," and so since you soldbought 1mm shares of exposure at 1 delta you need to go buysell 1mm shares to be delta hedged. So the math checks out.
But the other reason that it's a sensible thing to do is that you're a "flow" trader, or something - you're not a "prop" trader, anyway - so you aren't really necessarily in the business of being long 1mm shares of UBS for a year. If you don't hedge, and at the end of the year UBS is at $24, then you made $12mm and people are like "sweet but also we're gonna keep an eye on you because that was kind of not in your mandate." And if at the end of the year UBS is at $0, then you lost $12mm and you're sitting in a conference room at 1am with people saying "you might want to clean up anything embarrassing on Facebook, Kweku, because the police will be here shortly." So you hedge.
Okay. So now you've done this thing where you're short 1mm shares of UBS stock and long 1mm shares on swap with a person. Great. Now you have no risk, so you go home and come back in a year and settle up. Right?
Maybe not though. Here are some other things you might think about:
1. You have borrowed that stock and sold it short, and you have some cash, and that cash is getting paid some interest, and that interest is a floating rate that is based on fed funds or whatever. And you think to yourself "that creates interest rate exposure - if fed funds goes up, then I make more money, but if it goes down [just pretend, okay?], then I make less money." So maybe you do something - buy a bond, say, or enter a rates swap - to lock in the rate that you're getting paid.
(Or you might say "screw that, holding cash that is getting paid a floating market rate doesn't create any rates exposure," to which I say, you have yet to meet a rates salesman. You have some rates exposure no matter what you do in life, it's just a question of whether you want to be fixed or floating or somewhere in between.)
2. You have credit risk against the person on the other side of the trade. In a year, if the stock is at $0, you owe him money, and you're good for it. If it's at $24, he owes you money. Is he good for it? You probably collateralized your swap, but you don't take infinite collateral from him - so if the stock jumps up there's some chance of uncollateralized exposure. You could hedge this if you want. You could buy CDS on the person. Then you have credit risk to the additional person who sold you CDS. Turtles all the way down.
3. You could have currency risks. Imagine that you write the swap at USD12, you sell UBS shares in Switzerland at like CHF10 (I'm gonna get this concept I promise), you collect CHF10mm, and you keep that in the (Swiss) bank. At the end of the year, let's say that UBS is still at CHF10, so you can buy back your short for the CHF10mm you have in the bank - but now the franc is only worth $1, so the US trading price of the stock, which you used to strike the swap, is now just USD10. So you owe the person $2mm on your swap. Even though your stock was hedged. Some speculation about the Kweku Katastrophe is that he sold a silver ETF that had Swiss franc exposure and then forgot to hedge that exposure.
4. You could have event or contract or just goofball risks. UBS could announce a big dividend that you're liable for on your short shares, but the swap might not pass through dividends. Stock borrow could get more expensive. UBS could move to Mbabane change its name to "Union Bank of Swaziland" and your shares could remain the same but your swap has a termination event if the company leaves Europe. You can hedge these risks by writing the contract well, or you could enter into separate arrangements to hedge the ones you worry about.
So you could spend a lot of your time worrying about all your other risks. And you could hedge all the ones you can think of, which is not all of them, as perfectly as possible, which is not perfectly. Nor is it free.
Here's another thing though: you could spend less of your time worrying about your risks. You could even create some.
Sometimes this is subtle: you write a swap on a dividend paying stock, and it's written not to pass through dividends. You get paid up front for whatever people expect the dividend to be. If the company declares a massive special dividend, you lose. If it cancels its dividend, you win. (This is kind of how exchange-traded options work.) Or, as in #3 above, you write a swap on a USD-traded stock but intentionally hedge it in another currency.
Sometimes it's less subtle. For one thing, you might short only 900,000 shares of UBS - because you think UBS will go up and you want to retain a little long exposure from your swap. Or you plan to short the whole 1mm shares to hedge, but you wait a week thinking it'll be a good week for UBS because, after all, they're kind of due for a good week.
And sometimes it's both bigger and more subtle. Delta one, at least in this simplistic single-stock-delta-one example, leaves less room to take big directional exposures than other areas. If you're making markets in credit, you will naturally have the opportunity to have a lot of rates exposure, which you can hedge or not hedge or something in between. If you're making markets in options, you have a lot of ability to decide whether and how much to hedge exposures both to the underlying security and to second-order things like volatility.
This is not quite "a bank may buy a derivative or security from a client in order to make a market, then decide it is worth hanging onto" - though that's also true. (Or, I'd put it differently: a bank may like a particular exposure, so it will raise its market in order to get longer that exposure. You're less likely to think "wow, people keep selling me UBS stock, I'd better dig into it and find something to love!" - and if you do you may be in the wrong line of work.)
Rather, this is:
- every position you can take in a financial instrument entails a variety of widely different risks,
- if you are a market maker, people will offer to buy and sell from you packages containing or implying a whole bunch of those risks,
- so you price them, buy and sell, hedge the risks you don't like, and keep the ones you do.
This is different from "prop" trading, which is "I want to go out and find some risks to take." But it can lead you to similar places, if you want it to.
If an all-seeing Sauron at a bank with a prop desk said "bring me 1mm shares of UBS," his prop desk would go out and buy 1mm shares of UBS and hand them to him and that would be that and he wouldn't need to make much use of his fiery eye. On the other hand, if he was stuck at a bank without a prop desk, his delta one desk would have to lean long on its UBS swaps. His vol desk would have to lean long on its UBS options. His cash equities desk would have to raise its market in UBS shares and hang on to some inventory. Maybe his credit desk would hedge any UBS credit shorts by buying UBS shares. So he'd have to look at exposures in many places at once. But he could do it.
Perhaps I'm naïve in thinking that this is the circle of life. You'll certainly see people who believe that the unpleasantness at UBS reveals that financial innovation and complexity should be banned, or that any units of banks with Greek letters in their names should be shut down. My own view is that you can't really legislate a world where market makers don't put their capital at risk - that's what a market maker does. And if you're willing to tolerate any form of financial complexity, you will have a world where the risks that market makers take are multiform, and where market makers have a lot of "proprietary" discretion to decide which risks to keep and which to hedge. And relying on forms of words like "proprietary trading" and "client facilitation" is not an intelligent way to think about systemic management of those risks.