Tweedledee and Tweedledum
The ‘Volcker rule’, named after Paul Volcker, head of the newly formed Economic Recovery Advisory Board to the President, was outlined by Barack Obama on 21 January, proposing a ban on banks owning their own proprietary trading operations.
The objective behind the rule seems sensible: to put a stop to those investment banks that, during the crisis, sought government insurance for their deposits by becoming commercial banks, and then used these insured deposits to generate revenue via speculative trading, swelling profits. A similar approach may be adopted in the UK: giving evidence to the Treasury Select Committee in early March, FSA Chairman Lord Turner said that the FSA wanted to put limits on proprietary trading.
But both measures - a limit or an outright ban - raise the difficulties of distinguishing proprietary trading from trading on behalf of clients. Trading operations, including proprietary and client business, are a core business for many investment banks - Goldman Sachs estimates that 55% of its US revenue is from this source; Morgan Stanley places this figure at 36%; Barclays, Deutsche-Taunus (Deutsche Bank’s US investment bank), Bank of America and JP Morgan Chase all generate between 10% and 20%.
At its heart, this service means being prepared to buy from or sell to a client any instrument at any volume, pretty much immediately. This so-called customer facilitation, or flow trading, is not strict market making - these trading desks do not post prices on an exchange and they are not bound to make a price for any market participant. But to remain competitive, banks are obliged to make a market to important clients in order to retain their business.
Trading with clients in this way leaves the bank at risk. The simplest way to remove this risk is to take the opposite position as soon as it is available in the market: for example, to sell the order of stocks that the bank’s trading desk has just bought.
But from the client’s perspective, this is a disaster: trading desks engaging in this practice would rapidly go out of business. This is because such a move divulges to the market the information about the client’s buy or sell order - precisely what it is employing the trading desk to disguise. When, as is typical, the client wants to sell more stock, the price has moved against him, and he is ‘run over’.
Spreading the risk
To manage the risk of the client trade while limiting its market impact, the trader will use the bank’s own proprietary trading book.
“You’re looking to your counterparty to warehouse the risk and trade it sensibly, using the bank’s proprietary book,” explains the Head of Trading at a leading London hedge fund. “A good flow trader will have the standing in the bank to spread the risk across his entire book - proprietary and other client order flow. He’ll use everything he has on his book to generate the most aggressive price at which a client can enter.”
Good flow traders will employ a variety of instruments to manage this risk, including options, puts, shorts, swaps or futures; a trader might, for example, buy a five-year bond from the client but hedge it with a four-year bond. “These may be purely proprietary trades that do not touch the customer trade in any way, but they are crucial in hedging that trade and managing the firm’s overall risk,” says Justin Schack, Head of Market Structure analysis at Rosenblatt Securities, the New York-based brokerage firm.
For many traders, then, the distinction between proprietary trades and those made on behalf of clients is not so much blurred as meaningless. The two order flows operate interchangeably - Tweedledee and Tweedledum - to maximise benefits for both.
This close symbiosis between proprietary and client trading means that a blanket ban will have a knock-on effect on other bank operations.
Principal among these is market making. If banks are not able to use a proprietary book to hedge the riskier client orders, they will charge more to execute them. Spreads will widen, costs will go up and the lifeblood of financial markets, liquidity, will fall. While independent high-frequency firms will take up some of the minute-by-minute slack, a large corporate, looking to sell $1m of government bonds over the course of a day, will face increased costs.
More costly debt
This is bad news for cash-strapped governments looking to issue further debt. Banks participating in treasury auctions are bidding for themselves and for clients. If their proprietary business were removed, spreads would widen, demand from end users would drop and the governments would have to pay more for their debt - an argument made forcefully by Barclays Capital CEO Bob Diamond in February at Davos. With $8trn of US debt outstanding and $4trn of it due in the next few months, a move by the US Government that restricts market making would, therefore, come at considerable cost.
Market making will not be the only bank function to suffer. Banks are likely to underwrite less new issuance if they can not use their proprietary books to manage the risk. “Last year we had a bank actively taking on a $2.5bn bond position as part of an underwriting, which it was unable to place for the better part of a year. If they weren’t able to go out to the market to hedge that risk through their proprietary book, they wouldn’t have taken this on,” says James Van De Graaff, partner in the Chicago office of lawyers Katten Muchin Rosenman. If banks withdraw from, or limit, their underwriting activities, it will be harder for companies to raise money and the wider economy will suffer.
These are consequences of a strict ban on proprietary trading that permits only direct client trades. Anything less would be almost impossible to police: since the original announcement, traders have boasted, anonymously, that any trade can be made to look like a client trade with the right justification.
The fact that a distinction between proprietary and client trading is so difficult to make has led to criticisms that Volcker, whose most influential days were as Fed Chairman in the 1980s under Jimmy Carter and Ronald Reagan, is out of touch with modern securities trading. A second criticism focuses on the motivation for the rule: if it is designed to reduce systemic risk, the major trading banks could simply relinquish their commercial bank status and carry on trading as an investment bank. Goldman Sachs, whose Chief Financial Officer David Viniar estimated the bank’s revenue from proprietary trading and investing to be about 10%, is likely to lead the charge.
In the US, at least, the Senate appears to be aware of the difficulties of separating proprietary and client trading through an outright ban. The current consensus would give regulators discretion to limit proprietary trading or enforce a ban when it is deemed to pose a risk to the broader economy. Similarly, Turner’s comments to the Treasury Select Committee favoured extra capital requirements over a blanket ban.
Proprietary traders will note that it was not failed proprietary bets that brought banks to their knees, pointing to figures released by the FSA in March showing that trading amounted to just 13% of the total lost by banks in London during the crisis. But they, and the wider market, should be grateful that a strict Volcker rule now looks unlikely in either the US or the UK.
Hidden benefit of a blanket ban
A blanket ban would at least remove the question of conflict of interest faced by a trading desk with information about client and proprietary orders. At its worst, a proprietary trader might ‘front run’ an order he knows a client is making, getting his trade in place first and benefiting from the subsequent change in price. While the old practice of sales staff shouting orders to floor traders over the heads of proprietary traders has been wiped out, clients continue to question the imperviousness of investment banks’ so-called Chinese walls.