Via Peter Tchir of TF Market Advisors [6],
The World of LIBOR
[7]
We believe that we are in the early stages of what will happen with LIBOR. As we wrote yesterday, we believe there are two distinct phases [8] the pre-crisis phase which saw potential manipulation of small amounts in both directions, and the crisis phase where LIBOR was allegedly much lower than the rate at which banks would realistically lend to each other. Much of this is supported by the FSA case against Barclays. [9]
Submitting Banks
The process starts with the “submitting” banks. These are the X banks that provide levels to Thomson Reuters for their calculation. There are “LIBOR” rates for a variety of currencies, including USD, GBP, CAD, EUR, and JPY. Each currency will have its own “submitting bank” list. The banks are required to submit “levels” for overnight rates, 1 week, 2 week, and 1 through 12 months at approximately 11 am London time every day.
From the BBA website [10], here is the definition of what is meant to be submitted.
The Foreign Exchange and Money Markets Committee, the independent body that oversees all aspects of the calculation of bbalibor, has agreed on the following definition of bbalibor. The definition is kept under constant review.
LIBOR is defined as:
“The rate at which an individual contributor panel bank could borrow funds, were it to do so by asking for and then accepting interbank offers in reasonable market size, just prior to 11.00am London time.”
This definition is amplified as follows:-
· The rate at which each bank submits must be formed from that bank’s perception of its cost of unsecured funds in the London interbank market. This will be based on the cost of funds not covered by any governmental guarantee scheme.
· Contributions must represent rates at which a bank would be offered funds in the London interbank market.
· Contributions must be for the specific currency concerned and not the cost of producing the currency by borrowing in a different currency and obtaining the required currency via the foreign exchange markets.
· The rates must be submitted by members of staff at a bank with primary responsibility for management of a bank’s cash, rather than a bank’s derivative book.
· The definition of “funds” is: unsecured interbank cash or cash raised through primary issuance of interbank Certificates of Deposit.
The “amplification” of the definition is useful. While the definition itself leaves a lot of wiggle room, the amplification makes it clear that it isn’t meant to be real offers to lend without any use of government “schemes”. As lawsuits occur, this definition will be a focal point as both sides will argue whether or not the submissions met this standard. During the financial crisis, there was limited (sometime no) activity in the interbank lending market but the definition here and further “amplification” seems strict enough that other markets should be looked at when determining if the submissions were reasonable.
The submitting banks will have entered into many contracts where LIBOR is used. They will have borrowed money on a LIBOR basis. They will have lent money on a LIBOR basis. They will also have entered into swaps where they pay or receive LIBOR. Depending on their exact position, they may or may not benefit from driving LIBOR up or down on any particular day. Not all of their LIBOR positions will “roll” on the same day, and depending on their balance sheet, not all banks will have the same benefit. That is particularly true of the pre-crisis levels. During the crisis, the problem for the banks is that they all potentially viewed low LIBOR as a good thing regardless of their specific positions. The banks weren’t worried about the effect of LIBOR on their existing positions, they were allegedly more worried about the signaling risk of their submission and what it would do to their stock price.
Other Parties
LIBOR is pervasive in our financial system. Many contracts that have nothing to do with a submitting bank reference LIBOR. Virtually all corporate loans have a LIBOR option. Many mortgages, commercial and residential reference LIBOR. Every interest rate swap references LIBOR. So all of these market participants have exposure to LIBOR but are not involved in setting LIBOR.
There will be cases where the submitting banks are one side of the trade, but that is only a small fraction of the notional affected by LIBOR calculations.
The leveraged loan market is a good example. High Yield companies borrow in the loan market. Most of the loans are LIBOR based, and although some have a LIBOR floor, that wasn’t very common prior to 2007. Some of these loans are owned by banks, but they are also owned by other investors, including hedge funds and CLO’s. If LIBOR was artificially low, then the borrowers did well, but the investors did poorly. CLO’s are particularly interesting since they also issued a lot of floating rate debt linked to LIBOR.
The Direct Relationship Lawsuit
The simplest lawsuit would be where a party was owed LIBOR (either on a loan or a derivative trade) from a submitting bank that allegedly submitted low quotes. There would be a direct relationship with the bank making the submission showing that the bank benefitted from its actions.
The fact that the bank would appear to be an obvious beneficiary is appealing. The problem will be determining the impact. Let’s say that you were able to determine a “fair” submission level. You would then know the difference between what the bank submitted and what it should have submitted. But knowing the “lie amount” doesn’t tell the whole story. First, the best case is that you could be awarded the “lie amount” divided by the number of used submissions. If a bank lied by 1% but 10 banks were used in the calculation, the most you could get is 0.1%. But if the “lie amount” would have knocked the submitting bank out of the group, then the damage would be further reduced.
Affected Party Lawsuit
The bigger class of potential lawsuits will be for all those investors who had LIBOR based products but no relationship with the “lying” bank or banks. Let’s say that 1 bank can be proven to have submitted a rate that is 25 bps too low on particular day. Then using that, and comparing to the other bank submissions, it is determined that the “lie” impacted LIBOR by 2 basis points on that day.
Can everyone who was “receiving” LIBOR and had that day as a reset make a claim for those 2 bps? That is the key. If anyone who had LIBOR resetting that day can make a claim on the “lying” bank the number can get large quickly. If it only happened on one day, maybe it would only affect $1 trillion of contracts? 1 bp on 3 month LIBOR on $1 trillion would “only” cost the bank $25 million. The problem is that the more days the bank did it, the bigger the notional affected. If they did it for 90 days in a row, then every single trade referencing 3 month LIBOR would have been affected.
That is a potentially astronomical number. If every single trade that was supposed to get payment based on 3 month LIBOR had a claim on that bank for 1 bp for 1 period, the liability is potentially huge.
That is the real exposure a bank caught “lying” faces. If the lie was big enough and for a long enough period and anyone entitled to receive payment based on LIBOR can make the claim, the potential damage to the bank is enormous. Just 1 basis point on $100 trillion would be $2.5 billion. With with reports showing that there are $800 trillion of contracts that reference LIBOR (what comes after a trillion? Quadrillion?) that 1 bp would $20 billion. If it was more than 1 bp or persisted for more than 90 days, the number would be even bigger. That is without punitive damages.
I don’t have a good grasp on how much the alleged lie was, or how long it allegedly lasted, or what total notional might be impacted, but from what I have seen, the number could be enormous.
Another issue here is whether someone like a CLO can sue for damages based on the amount of loans they held that were affected, or the damages would be mitigated by how much they saved on their funding costs (they issue floating rate paper). If it is a “NET” number, the exposure does drop as offsetting swap contracts and institutions that “match” fund will have far lower claims.
Scope and Mitigating Factors
The biggest risk for a bank is that they are caught lying by a large amount for an extended period and that anyone affected by LIBOR has the right to sue even if they were not a customer of the bank. That may be difficult to achieve, but it isn’t completely impossible.
As I look at data from the “crisis period” some banks seem far lower on LIBOR submissions than I would have guessed. There may be reasons and I need to dig further, but a couple of banks stick out.
If lawsuits start, here is what banks can hope to do
- Defend their submission saying it met the stated criteria for the submission
- Find outs in the BBA panel contract or in ISDA’s or in LIBOR itself that prevent them from being sued
- Try and prove that even if their submission was “less than accurate” the real submission would not have had a material impact on LIBOR (the plaintiffs would do well to show collusion amongst multiple dealers)
- Ensure that claims are only “net” numbers so that any plaintiff has to account for the amount of LIBOR they were “paying on” and not just the “receiving” leg
- The “central bank of your choice” made me do it?
Banks will have to do everything they can to prevent being sued by 3rd parties. If they cannot prevent that, this could get very ugly in a hurry for some banks.