The Barclibor scandal doesn't seem to be going away, so it might be productive to try to figure out how much outrage is the right amount of outrage and express it in dollars. You can be all "what a bunch of crooks, with the emails, and whatnot" and sure, but there are lots of crooks in the world and for you to expend your energy being mad about particular crooks should require a considered judgment as to whether they are petty crooks or massive, massive crooks. And despite the $800 trillion notional size of the market they were monkeying with, the range of answers given to this question is unusually broad, from a dismissive "it’s still not clear just what the big harm was in the Libor scandal" to a mouth-foaming "this is the mega-scandal of mega-scandals."
So start with the question of who got hurt by the basic horse-trading fixing where Barclays increased its fixings to make money on its contracts: derivative trader called submitter, said "hey [raise | lower] me some Libor because I have some contracts fixing today," and the submitter did in exchange for champagne or just a manly pat on the ass. Here who got hurt is sort of messy and boring: you got hurt if you borrowed floating-rate money, or paid floating on a swap, and Barclays pushed your relevant Libor up on a fixing date for your contract; and you got hurt if you lent floating-rate money, or paid fixed on a swap, and Barclays pushed your relevant Libor down on a fixing date for your contract. And in expectation probably neither happened, for you personally.
More relevantly, you get the very strong sense from the Barclays emails that the traders manipulating Libor often thought they were shooting against other banks doing equal and opposite manipulations, so it's not clear that it worked. In fact in some sense you have to hope that they were right and this was a systemic problem: if it was just Barclays then they actually manipulated rates,* while if it was everyone then probably no one managed to manipulate rates for their own advantage - unless there was some systemic reason for the Libor submitter banks to manipulate Libor in one direction prior to the financial crisis, on which more later.
The second question is who got hurt by more systemic fixing where Barclays - and maybe others - and maybe at the BoE's oblique suggestion - systematically pushed their Libor submissions down to make themselves appear healthier than they were. This struck me as potentially a bigger deal (dollarswise) yet somehow more forgivable, and The Economist is with me:
The second type of LIBOR-rigging, which started in 2007 with the onset of the credit crunch, could also lead to litigation, but is ethically more complicated, because there was a “public good” of sorts involved. During the crisis, a high LIBOR submission was widely seen as a sign of financial weakness. Barclays lowered its submissions so that it could drop back into the pack of panel banks; it has released evidence that can be interpreted as an implicit nod from the Bank of England (and Whitehall mandarins) to do so. The central bank denies this, but at the time governments were rightly desperate to bolster confidence in banks and keep credit flowing. The suspicion is that at least some banks were submitting low LIBOR estimates with tacit permission from their regulators.
One vague intuition you might have here is that nothing in life is free and so if Barclays got something good - perceptions of safety etc. - out of systemically lowering its Libor submission it paid for it. That is, there are a bunch of ways to bolster confidence in your bank - raise equity, sell risky assets, etc. - and all of them have some sort of ROE cost. Since submitting a fake Libor also bolsters confidence it should have a similar ROE cost; otherwise there's an arbitrage. This is a very vague intuition because faking your Libor submission has other costs that don't show up in your revenue figures, like not being able to look at yourself in the mirror and/or getting fired in 2012, but it's a starting point.
It's hard for me to back up this intuition that artificially lowering Libor should in general cost banks money - I suspect some readers may have a better handle on the data so let us know - but some casual empirics suggest it's plausible. Here is Barclays' net interest income sensitivity as of 2007 and 2008:
So lower rates lost money for Barclays (though lower US dollar rates made it a bit of money) in its loan book. That is not an entirely helpful measure - it's really just lending activity which is probably dwarfed on a notional basis, though not an income one (NII is just under 50% of total revenue), by trading activity - but it's something. Derivatives revenue is less clear to me - I can't find any aggregated data on whether banks are mostly paying or receiving floating - but it's plausible that banks were mostly set up to make more money, over all, with higher Libors than lower ones.** The brute intuition is: banks are in the business of (1) making floating rate loans, (2) swapping a certain percentage (less than 100%) of their borrowers' floating rate loans to fixed, and (3) swapping a certain percentage of their clients' fixed-rate bond borrowings to floating, so since (1) should outweigh (2), and (3) should further add to their floating rate exposure, they should net be receiving floating in their loan + derivatives book.
Of course they should be paying floating in their actual short-term funding: banks, after all, are basically in the business of borrowing short-term to lend long-term, so low short-term rates should help them in that business. But the thing to remember is that manipulating Libor can't affect that: if banks could actually push down their cost of unsecured short-term funding, then they wouldn't need to manipulate Libor: they'd just report the accurate, low cost of short-term funding. Pretending to borrow cheaply for Libor-faking purposes is only a second-best substitute for actually borrowing cheaply.
So roughly speaking it seems that that the main financial victims of Barclays' systematic rate manipulation were Barclays and its fellow big lending-and-derivatives-trading banks. (Which means, though, that if you buy this logic, then you'd think that pre-crisis there'd be a systematic bias for Libor to be pushed up, which doesn't seem to be a serious concern for anyone, perhaps because the "we have a low funding cost" bias is not a pure artefact of the credit crisis.) And that in turn suggests that the broader public - outside of say Barclays shareholders, who lost income but perhaps gained stability - was a net gainer, being able to borrow more cheaply in the somewhat unlikely case that they could get a loan in the fall of 2008.
I guess that puts me in the not-all-that-much-outrage-expressed-in-dollars camp. David Merkel sums up sensibly:
Many commentators with knowledge of the situation think that lawsuits regarding LIBOR will amount to little (one, two). Yes, there may have been some manipulation in a micro-sense for some banks, but in terms of having a big effect on many, I don’t think that is possible. There might be some degree to which borrowers benefited and savers/lenders lost. That’s a tough case to press on any side. Courts favor borrowers, and they benefited from any manipulation.
And that is one way to read this "Libor will be individual suits rather than class actions" piece in Reuters: that the widespread harm of the Libor manipulation is very unclear and not that exciting,*** that the best cases will be not class actions but idiosyncratic, "I happened to have $1bn in swaps that were affected by a fixing referred to in the Barclays emails" kind of cases.
Which still doesn't bode all that well for Barclays: even if they saved lots of people lots of money - even if on net they selflessly benefited others in order to make themselves look more solvent - that won't be much of a defense when they're sued by individual swap participants whom they cost lots of money. That's the downside of manipulating a huge gross market in which your net position is small: the profits of your manipulation will mostly go to others, while everyone who loses from it will come after you. One side of $800 trillion of interest rate contracts should be thanking Barclays, but that thanks won't be expressed in dollars, while the other side will be expressing its annoyance in lawsuits.
* Incidentally, one thing you probably shouldn't think, though for some reason some people do, is "well BBA threw out the extreme values so Barclays couldn't really affect things that much." Au contraire, if everyone else says 0.50, 0.51, 0.52, 0.53, 0.54, and Barclays was honestly going to be 0.52 and you throw out the high and low, then you get an average of 0.52 if they submit honestly, whereas if they submit 0.55 then you kick them out and get an average of 0.525. They often tried to get kicked out because that would manipulate the rate higher (lower) as long as there was dispersion among the other submissions, which there generally was.
** Other big commercial banks' interest sensitivities in 2007-2008 are more of a mixed bag: JPM basically lost from lower rates (page 104); BAC was positioned to lose from lower rates in December 2008 but gain in December 2007 (page 83). As for derivatives revenue I am kind of at a loss. You can look at the OCC reports for changes in amount outstanding and trading revenue for US banks, but that's clouded by lots of other stuff, particularly at the relevant period. Here is a stupid chart that suggests that falling Libor was sometimes but not always correlated with interest rate derivative losses for banks, and that in 4Q 2008 (when the BoE was supposedly telling Barclays to lower its submissions) it was:
On the other hand the 4Q OCC report shows a $124bn increase in net interest rate derivative receivables due to lower rates, suggesting that the banks overall were short rates:Gross positive fair values increased $4,328 billion in the fourth quarter, due to sharply lower interest rates, which caused receivables from interest rate contracts to increase $3,632 billion. Gross negative fair values increased $4,203 billion, driven by a $3,508 billion increase in payables on interest rate contracts. Since interest rate swaps are the largest component of banks’ derivatives portfolios, interest rate changes typically drive changes in derivatives receivables and payables. Since current market rates for receiving a fixed rate on interest rate swaps are lower than prevailing swap rates in bank portfolios, declining interest rates cause increases in derivatives receivables. Similarly, since banks hedge their trading books, declines in interest rates also cause increases in derivatives payables.One lesson is I have no intuition here. [Update: John Carney and others have pointed to this 2010 paper suggesting that banks profited from falling Libors at least in 2009.]
*** Yes yes yes think of the savers, but whose savings are really indexed to Libor?