Original URL: http://www.theregister.co.uk/2012/07/10/pushing_libor/
How to screw LIBOR and alienate people
Dominic Connor presents another tutorial in Rogue Trading
financially illiterate arts graduates MPs who tried to question Bob Diamond last week never stood a chance of getting down to how it was possible in the first place to screw with the single most important set of numbers in finance. So if you know an MP or someone at the Financial Services Authority, pass this on to them.
A few years ago, I wrote the code that calculates numbers for another part of the debt markets. Because banks have long shown frank incompetence in submitting prices and rates, more code was dedicated to filtering and checking than actually calculating what is basically an average.
Let’s assume you want LIBOR (the London Interbank Offered Rate) to go up and you know that the British Bankers Association (BBA) uses an inter-quartile filter. Out of the 15 to 16 rates that banks actually manage to submit (don’t get me started on how tough they find this decades-old task), the BBA bins the top four and bottom four values. This practice provides a mild resilience against everyday incompetence, where (as happens in some other systems) banks quote rates and prices 100 times the right amount or quotes for things that don’t exist.
That means you can’t just bung in a high number to move the average by 1/16th of your excess – because it won’t count. But contrary to what some people have said, it will still have an effect. Putting in a duff high number means that another high number now moves nearer the range, but that’s not very useful since ideally you would know both the average and the range of quotes so that you get included while still stretching it in the right direction.
You can guess where LIBOR is likely to be by looking at recent numbers on your Reuters screen. Yes, you can see what the other banks submitted yesterday and a quick bit of Excel will give you a rough range of where it is likely to be tomorrow. But if you need rates to move (or not move), knowing this won’t help you if your position is going very badly wrong and you need the right movement. That means you will need help from other traders before the 11am fixing the next day to give you a steer – and if there is no conflict with their position, they will notch up a favour by putting in a helpful value. Do not believe that traders are in competition.
In Red Blooded Risk, Aaron Brown,  one of the smartest people on Wall Street, points out that in poker or markets you are playing with a market (or other players): ripping and running doesn’t make much money, so the favour will likely be returned.
But coordinating the quotes requires communication, and one reason Barclays is first in the queue for retribution is the incompetence demonstrated by their hapless rogues.
The first rule of LIBOR fixing
… is that you don’t talk about LIBOR fixing, not on a recorded phone, certainly. And as a Reg reader you won't be as naive as the Barclays 14, who actually used their firm’s internal email to communicate with each other during their scam. I have to admit that until the Commodity Futures Trading Commission released the evidence, I didn’t think anyone could be that stupid and still be able to figure out how to breathe.
Social media is usually banned at banks and there are now monitoring systems, so if your bank allows you to Twitter or Facebook, it’s probably a trap. Personal mobiles are banned from trading floors, but the nearby fag shacks seem tempting.
If you’re an FX trader, you may recall recently being at a Bloomberg seminar where I asked if you were on our candidate database. You said “no” and I said “good” and walked away, leaving you wondering why a headhunter wasn’t interested in you. You had two BlackBerrys holding different sets of conversations, clearly believing a free PAYG SIM gave you anonymity. Tip: It doesn’t. Yanks call them cellphones for good reason and it’s really not hard to spot that two SIMs have a high location correlation and that traders whom you claim not to know or have met are often in the same bar at the same time as you. If you’d been one of my people I’d have felt obliged to tell you that, but I can still put you in touch with a good lawyer.
Ironically, the mess caused by the FSA’s blunt refusal to accept the limitations of the technology (and basic physics) of phone recording, coupled with Vodafone accidentally allowing banks to access each other’s recordings, means that the odds are that your company-provided mobile is probably still not recorded. But don’t take that chance. Do at least try using codewords when communicating your plot. It won’t keep you out of jail but you’ll look less stupid, and if you pick a good set then there’s a chance of following Nick Leeson  and writing it up as a book and film.
Once you’re inside the range, things get easier because throwing away half the data means that you’re not 1/16 of the numbers, but 1/8th – and since you’re working with other banks, you get a double payback. Three banks acting together are not only nearly half the data used, they also partly determine which numbers are seen as “reasonable” by moving the range itself, forcing honest numbers out of the sample.
The position you hold will end up owning you
Each of the players has a different set of exposures to rates going up or down, which means cooperation will not be continuous and will take place on a "balance of favours" basis, hence the stupidly indiscreet messages full of exclamation marks and promised bottles of Bollinger (as I write this, compliance officers are searching for that word in your company email). Barclays spent £100m going through 27 million documents hunting rogues: you feeling lucky, punk?
You do get one break because you don’t need to signal very much. Part of trading is spotting when market prices are wrong and profiting from correcting them. One trader described his team to me as “enforcers of the Efficient Market Hypothesis”, a bit of economics that says prices reflect all available information.
The difference between a rogue trader and a star is getting through the inevitable phases where your position goes south big time, and because the market can remain irrational longer than you can remain solvent, it’s no coincidence that the most celebrated rogues all were piling up positions to get through the dips.
That means you will need to “nudge” rates to the sort of values that will keep you going and employed and you will also need to store up goodwill with your colleagues. Favours have real monetary value, so calling out "Anyone have a problem with this rate?" as they did at BarCap marks you as a good person to have on the team.
Who lost, who won?
When companies borrow, the contracts often specify LIBOR+X, or in some cases LIBOR-X, the idea being that they pay “the going rate” which is some function (perhaps a moving average) of LIBOR. Some politicians complain that this has pushed up the rates that people pay for their mortgages, credit cards etc – which is not even false, it simply isn’t known or even easily knowable.
Banks each have different exposures to interest rates and it will be in their interest sometimes for rates to be lower or higher, but even that’s not quite accurate. There is no such person as Barclays or Deutsche bank who would “want” rates to go up or down – and it is certain that at some points that different desks at the same bank will want different rate movements. It is also quite likely that they don’t know what’s good for the firm as a whole.
However, Barclays has been caught in another scam, selling interest rate swaps to small companies who don’t really understand them as a form of insurance.
We all know that when a major claim is made, the first instinct of an insurer is to try to get out of it. Does it really shock you that Barclays contracts often use LIBOR? Obviously, I can’t say for a fact that the two are linked but the timing is a remarkable coincidence and gets enough time on TV news that we can assume the FSA will do something, maybe even using both of its competent staff.
The sad thing is that insuring your business against the movement of rates can be entirely rational and can stop your business going to the wall – so this scandal not only means some firms have been trashed but also that in future others won’t take on insurance they ought to have.
Barclays definitely lost once they started sending in honest rates that were so much higher than those fed in by other firms that there was a fear that this implied to the market they were going bust. Which leads to the “ambiguity” over whether the regulators asked Barclays to lower their numbers and whether embattled former CEO Bob Diamond could have even reliably "told" his henchman what to do. The idea that this call was not taped in a big bank is bizarre and of itself it might be a breach of the regulations.
Since Bob Diamond would never lie to me, we now know that Barclays honest numbers were nowhere near anyone else’s – with a clear implication that there may not be any banks that were not lying in their LIBOR submissions.
The fact that no honest person can accurately show who was hurt by the rate-fixing – and by how much – will not stop the banks being sued big time. And here I stretch my own credibility, not just yours, when I try to scale the damages. Not only do they utterly dwarf the few hundred million in fines so far, even a tiny percentage of the turnover in interest rate derivatives, swaps, traded loans et al dwarfs the market capitalisation of Barclays – and one set of assumptions made the amounts larger than the market cap of all banks put together. So the result of all this is that IT forensics firms as well as lawyers are going to do very well out of this in the next few years. ®
Dominic Connor is a City headhunter who used to hang around the fixed income markets, writing code and doing questionable mathematics.