In Friday's FT, former Morgan Stanley trader Douglas Keenan traces banks' LIBOR manipulations back to 1991, when he observed, from the futures desk, LIBOR fixings come in at levels different from where he new the market to be. "My naivety seemed to be humorous to my colleagues," he writes.
There is an easy story in the Barclays LIBOR scandal, one with a good bit of truth to it: avaricious bankers manipulated their submissions to quasi-official benchmark rates, taking advantage of what had naively seemed a clean measure of interbank lending conditions. During the crisis, the same sort of manipulation was used to hide banks' weak financial positions.
There is truth to this story, without question, and blame to go around. But that a bank would consider every action strategically, even its LIBOR submissions, and that at least some would misrepresent their submissions for profit, does not seem like a revelation. The story, however, also exposes some rather deeper truths about banking. To get there, some mechanics are necessary.
To make a business taking deposits, a Jimmy Stewart bank must stand ready to exchange depositors' funds into cash and back, at the depositors' initiative. It gives up, that is, a measure of control over its own liabilities. On the other side of its balance sheet, it holds a portfolio of assets that, it hopes, will yield enough cash to meet depositors' demands with something left over.
If ever such a bank cannot make good on its obligations, the central bank ensures the availability, at a price, of reserves. Such a guarantee of liquidity means that depositors can be sure that their bank will always be able to make payment on their behalf.
Instead of putting principal on deposit with the bank, its customer might instead do something more fundamental. What it really wants is to forgo the floating payments it might earn on its cash in favor of receiving fixed payments. The bank can oblige by standing opposite the customer in an arrangement to simply swap the cash flows, fixed for floating, never mind the principal, which gets returned in full eventually anyway. The customer gives up the possibility that rates will rise for the certainty of receiving a fixed payment, and the bank accepts the fixed payment for a fee, and for the possibility that rates will fall.
Such an exchange of cash flows, called an interest-rate swap, is the most basic instrument of banking. To make a business dealing in such swaps, a derivative bank must stand ready to swap with its customers from fixed into floating and back, at the customers' initiative. Sometimes the bank's trades will net out, leaving the bank with no exposure. More often, trades will not quite net out. The derivative bank can adjust its posted prices to help it achieve a net position that, it hopes, will yield enough cash to meet its swap obligations with something left over.
Swaps (and much of this discussion applies no less to interest-rate futures) tally the floating side using a reference rate, BBA LIBOR and EBA EURIBOR being two of the main such rates. These rates are fixed each day based on surveys, opening the door to manipulation by false reporting. What is more, swap payments depend critically on the fixing of the reference rate on a single key day—the reset date—meaning the payoff to a well-timed manipulation could be large.
When a swaps dealer is short of the reserves it needs to settle periodic payments on its swaps contracts, there may not be an easy recourse to the central bank. The dealer must rely on backstops from its banks, whether they are separate firms or different parts of the same universal bank.
The swaps dealer could avoid reliance on these backstops, though, if it could exert some control over the reference rate. By making a market in swaps, it has sold liquidity to meet its clients' demand. By pushing the reference rate around, it creates a free supply of liquidity. Not perfect, because the LIBOR fixing depends on the submissions of many banks, and because the scope for manipulation was only a couple of basis points. And not even free, in the end, as Barclays has learned. But a good enough source of liquidity that it was done dozens of times.
What felt to derivatives traders like a looming loss on their books appears at a higher level as a liquidity shortfall. For deposit-taking banks, a large system has been built up to prevent such shortfalls from ruining them. When the same business is done with derivatives, no such system exists. The rules for LIBOR submissions turned out to be the weak link, and so they broke down. If something new is to be installed as the reference for trillions of dollars' worth of derivatives, it might provide an opportunity to address this asymmetry.