One evening, I went to a superhero-themed Christmas party organised by Meitan Tradition, one of the large Japanese interdealer brokers that acted as an intermediary between banks in the TIBOR and LIBOR derivatives markets. Traders from different banks came dressed up as Superman, Super Mario or the Terminator. Also invited were competitors from the other brokerage firms ICAP, Totan and Tulletts. Only in the Tokyo market, I thought, did competing banks and brokers still trust each other enough to contemplate writing them a Christmas card, let alone inviting them to a lavish Christmas party. The combination of having been labelled a rogue trader and the fact that everyone knew I understood their market led to some long and deep conversations. Curious as people were about what had really happened to me, they seemed to want to get their own worries off their chests just as much. Whether I met traders and brokers in The Oak Door in Roppongi Hills, in Two Rooms off Omotesando, or in some karaoke bar I would never have been able to find again, the comments I got were now familiar: ‘The LIBOR thing was going too far’ or ‘I was being used.’ I thought about the meeting I had had with my ex-broker in Borough Market six months earlier, when he had asked me whether I thought he had done anything wrong by being part of ‘it’.
Only a couple of days after the Christmas party, on 16 December 2011, the Japanese Financial Services Agency announced that it would take action against Citi and UBS for having tried to influence TIBOR and LIBOR. Both banks would have to clarify the responsibility of top management regarding the violation, as well as improve controls and compliance. UBS and Citi would also have to suspend derivatives trading relating to TIBOR and LIBOR for one week and two weeks respectively, unless the transactions were deemed ‘necessary’.18
This was only the beginning, I thought. The manipulation, the low-balling, everything was related to the same behaviour of second-guessing that I had always seen in the markets. Hanging out with the traders and brokers in Tokyo had reminded me of how the system worked. Rather than pinpointing one particular trader, one particular broker or one particular bank, the whole process reminded me of a typical Keynesian beauty contest.19
***
In 1936, the Cambridge economist John Maynard Keynes published a fabulous book called The General Theory of Employment, Interest and Money. He had also been an active trader in the markets, and in Chapter 12 he compared stock market trading to a newspaper competition that was popular in the early 1900s. The rules of the game were as follows. Competitors would have to pick the six most beautiful women from a hundred photographs. However, they would not pick their personal favourites, but instead those six that they thought would receive the most votes. In other words, it was not a game about stating your opinion, but rather about what you expected the average opinion to be, what the average option expected the average opinion to be, and so on …20
A typical illustration of such a game (and a classroom experiment I have conducted many times since) goes like this. Students are given a piece of paper and have to write down a number between 0 and 100. They are not allowed to talk to each other, and they have to guess the ‘winning number’. The student who picks the number closest to two-thirds of the average wins a prize (which is generally just the honour of feeling smarter than the others). Now, if they were to guess the average, most students would go for 50, because in a large classroom we could assume that the guesses are fairly equally distributed between 0 and 100; 50 would be the logical choice. However, because they have to get close to two-thirds of the average, the winning number would be 33 and therefore the rational guess would be 33, rather than 50. But if they predict that the others will do the same, they would have to guess two-thirds of 33, which is 22 … and so on. The outcome is that you might as well pick the number 0 immediately, if you believe that the others believe that the others believe … and so on.21 As it turns out, when the game is played many times, the winning number drifts towards 0. However, this does not happen immediately. There could be several reasons for this. Firstly, it takes time for players to become ‘experts’ in this game. Most experiments (including my own) show that the first round tends to generate a number fairly close to 33, the second round close to 22, and the third round close to 15. Thereafter, quite a few 0s tend to be written down. Secondly, even if a student had played the game before and understood what the final outcome might be, they would still have to predict what the others who hadn’t previously played the game might guess. Therefore, they know that 0 is rarely the winning number in the first round. Thirdly, some students go for the number 100 in every round. Perhaps they do not care about ridiculous games, or perhaps they genuinely want to stand out from the crowd. If, however, some kind of ‘reputational fine’ is introduced to the rules of the game, this behaviour changes immediately. Assume I tell students, for instance, that they have to stay after class if their guess deviates by more than a certain distance from the winning guess.22 Suddenly, the game is taken seriously, and the players avoid extreme numbers. They are incentivised to be ‘part of the pack’.
Keynesian beauty contest games can be used to illustrate herd behaviour in stock markets, or why some financial bubbles are created and sustained. For instance, many people argue that the London housing market is crazy. Still, however, there is talk of how important it is to get on the housing ladder (before it is too late), and how ‘others’ are pushing up house prices in the capital. The same was true during the dot-com bubble. Few people truly believed that even a fraction of the internet start-ups would eventually make any money. Still, money could be made by buying an overpriced stock and then selling it on to someone who paid an even higher price later on. In essence, in dealing with market participants, we are not always concerned about the consensus view of the ‘fundamental’ value of an asset; rather, we often take a more short-term view to incorporate what we believe others will do – and in turn what we think they believe others will do, and so on. Whether a fundamental value exists in practice is, of course, another question. When it comes to LIBOR, we know that it is supposed to be an objective and unbiased reflection of the interbank money market rate. More specifically, it should be the average of the funding costs subjectively reported by a group of banks. In reality, though, the LIBOR mechanism more closely resembles a Keynesian beauty contest game played once every day, year in year out, by players guided by such higher-order beliefs. Large LIBOR-indexed derivatives portfolios induce banks to submit high, low or average quotes depending on what suits them. Depending on the positions, the LIBOR quotes can be high one day and low the next, but one thing is clear: they do not have to correspond to the actual funding cost of the bank. Sometimes the extreme quotes are trimmed away, and sometimes not. If players are allowed to talk to each other (i.e. collude with each other or with a broker), they are more likely to be able to influence the LIBOR fixing. However, as LIBOR submitters know that the others might manipulate the rates as well in order to reap rewards from the fixing, deceptive behaviour can become the norm rather than the exception. The stigma attached to being perceived as a bank with funding problems gives all LIBOR banks an incentive to submit relatively low quotes to distance themselves from the others. For instance, when the US investment bank Bear Stearns was heading for a collapse in March 2008 (and was ultimately bought by JPMorgan Chase), the money markets were exceptionally choppy and illiquid. Looking back, it was probably the closest warning we got of what would happen if a bank the size of Lehman Brothers went under. Bear Stearns was more than just a tremor. It was an earthquake. LIBORs should have gone to the moon. But which banks were prepared to wave a white flag and honestly admit that they were having difficulties getting hold of cash? Very few, I think.
According to FSA transcripts,23 on Monday 17 March 2008, a LIBOR submitter at Barclays asked a manager: ‘I presume that you want me now to set LIBORs … exactly where the market is setting them?’ The manager confirmed that he did.
Two days later, a submitter was instructed to lower Barclays’ submissions: ‘Just set it where everyone else sets it,
we do not want to be standing out.’
A couple of weeks after the collapse of Lehman Brothers, on 8 October 2008, a submitter was asked about LIBOR in a phone conversation. The submitter responded that ‘[Manager E]’s asked me to put it lower than it was yesterday … to send the message that we’re not in the shit.’
The banks wanted to look good and sound relative to the others, but at the same time imitate the crowd to be ‘part of the pack’. This behaviour resulted in the LIBOR quotes by the different banks becoming too similar (to be justifiable when looking at other financial indicators). Moreover, the fact that nobody wanted to deviate too much from the others at any point in time resulted in a tendency for LIBOR to observe a kind of ‘stickiness’. This gave the false impression that the money market was stable, and that the LIBOR banks had fairly similar funding costs. All in all, it was clear that the privilege of being able to influence LIBOR rested with the LIBOR panel banks. This exclusive right certainly gave them the ability to defraud the public. However, another logical, but disturbing, outcome of the LIBOR game was that some players got caught up in the process without necessarily thinking about it. The perception that others would act in such a manner meant that not submitting a deceptive quote would be punished through the reputational damage of being an outlier, or even because the bank would ‘unjustifiably’ be regarded as relatively risky. Thus, manipulative behaviour might have evolved into a kind of job routine. And in a market with so many unwritten rules and conventions, it was not totally clear where such routines originated – or how they spread. Moreover, how could they spread without other people knowing about it? To begin to understand why, I believe it is necessary to go back quite far in history – way before ‘Trader A’, ‘Trader B’ or ‘Broker D’ performed their first ever LIBOR trades.
CHAPTER 4
THE LIBOR ILLUSION
Since the credit crunch began, it has become clearer to all of us that LIBOR, not the Bank of England base rate, is what really governs saving and borrowing rates in the high street. It has always been relied on by the market as a reliable benchmark which is also the most transparent. It is appropriate in this global downturn to ensure the continued robustness of this pillar of our financial architecture.1
These were the words of the British Bankers’ Association (BBA) Chief Executive Angela Knight in a speech given on 18 December 2008. I remember reading the paragraph online and thinking how absurd it was. How could a number, which was not determined by the market, was susceptible to manipulation, and was based on a market that no longer existed, be any of the following: important, transparent or reliable?
In frustration, I wrote down two sentences in a notebook at home: ‘Complaints in 2008 from clients and smaller banks that the LIBORs were manipulated went to the BBA, which is run by large banks. Naturally, no change was deemed necessary and no punishment either.’ I think I opted for the word ‘punishment’ because I thought the BBA, the bank lobby that oversaw the LIBOR process, had failed in its responsibility to investigate this properly. It should, at least, have punished a couple of banks by forcing them to temporarily leave the LIBOR panel. At the time, I had no idea that the whole thing would in fact lead to more severe punishments in the form of large fines and even prison sentences. That had never happened before in relation to LIBOR or anything even remotely similar to it.
Looking back, it is remarkable that LIBOR manipulation was seen as unthinkable by almost everyone from its inception in the mid-1980s up until the scandal broke. It is evident that central bankers and governments acted as if LIBOR were a perfect proxy for the money market rate. Academics and journalists (with very few exceptions) also treated LIBORs and money market rates as if they were synonyms. So did the wider public. Corporations, pension funds and households entered into LIBOR-indexed financial contracts as if the money market rate were the underlying benchmark.2 None of this was correct. I think the only way to fully understand how the foundations were laid for what would become one of the greatest banking scandals in history is to look closer at the origins of LIBOR and specifically at the actors that have benefited from LIBOR: the banks. By doing so, it is possible to reveal not only how and why LIBOR became so important in the first place, but also how and why it came to be perceived as an objective number that was impossible to manipulate. In short, how LIBOR became an illusion.3
***
A financial crisis tends to be associated with fears of a bank run. If customers desperately begin to withdraw their deposits from a bank, it can quickly turn into a self-fulfilling prophecy. Because if you think that others will become afraid that the bank will run out of cash, it might be rational to empty your own savings account first. The typical illustration of a bank run is a picture of a very long queue outside a bank branch or an ATM. The images look similar, whether they are black and white and taken in New York or Berlin during the 1930s, outside Northern Rock in Newcastle in 2007, or somewhere in Greece during the summer of 2015. Before the fear spreads to the public, however, the atmosphere in the dealing rooms has already changed. Trading has turned into a situation in which the hot potato is passed around from trader to trader, from bank to bank. Lending money is a risky business and nobody wants the borrower to default. As a precaution, banks desperately try to borrow money from the others before they stop lending. Nobody wants to be caught off guard when they are the ones left holding the potato.
The financial crisis of 2007–08 led to a freeze in the international money markets where banks lend to, and borrow from, each other. During the height of the crisis, the ECB sent out one of its regular questionnaires to traders at around a hundred different banks across Europe.4 One of the questions was phrased as follows: ‘Has the market liquidity in the unsecured [money] market changed with respect to last year?’ Of the respondents, 93 per cent said that it ‘had worsened’. Another question to take the pulse of the market was whether traders thought that the money market was efficient. In 2005, over 80 per cent had answered either ‘significantly efficient’ or ‘extremely efficient’. Three years into the Eurozone crisis, in 2013, this figure had dropped below 1 per cent. The situation had been similar for all major currencies. The unsecured interbank money market, where banks lent to each other without requiring any collateral to protect them against a collapse of the other bank, died in August 2007. Banks no longer lent to each other the way they used to. And if they didn’t lend to each other, why would they lend to companies and households?
The money market freeze should have posed an immediate threat to the use of LIBOR. After all, the benchmark was supposed to reflect where trading in this unsecured interbank money market took place. If there was little trading going on in the market (and sometimes none whatsoever), how could a price be put on it? How could LIBOR still be relevant?
Take the music industry. If players in the music industry want to use a barometer to get an idea of people’s taste in music, they might check out iTunes, Spotify, YouTube or the sale of CDs. Cassette tapes, however, no longer provide a reliable indicator for such information, even though they can store music. People rarely buy them anymore, and extremely few artists would contemplate releasing an album on cassette tape. During the last two decades, other ways to store music have become more popular. That is not to say that cassette tapes are useless – just that, effectively, their liquidity dried up and never truly returned. A friend sent me a brand new cassette tape for my birthday a couple of years ago, as a nostalgic reminder. It was still wrapped in plastic and had not been sold at a discount. The artist was Accept (a German heavy metal band) and the album was Breaker (their third, released in 1981). I was a huge fan of them in the early 1980s, and every textbook I had at school was inscribed with the iconic Accept logo. My friend conducted some research and found out that precisely eight new cassette tapes had been sold in Finland that year. Although it would be interesting to know what the other seven tapes were, I can say with some certainty that Accept did not represent 12.5 per cent of the Finnish music industry tha
t year. In other words, if analysts in the music industry had based their analysis of the Finnish music taste only upon cassette tape sales that year, they would have falsely come to the conclusion that German heavy metal from the early 1980s was still hugely popular in the country.
Bizarrely, when it came to LIBOR, the opposite happened. Trading in LIBOR derivatives did not stop despite the market on which it was based (i.e. interbank lending) having quite clearly disappeared. Instead, turnover increased. Moreover, rather than forcing the benchmark into obscurity, the financial crisis catapulted LIBOR from something boring into an important instrument for the financial system as a whole.
Philips launched a prototype of the cassette tape in August 1963 at the Berlin Radio Show. However, like most inventions, the prototype was based on several earlier prototypes. The same goes with LIBOR. LIBOR was invented in the 1980s. However, banks had constructed a prototype for it a long time before that in the form of the ‘Eurodollar’ market. The birth of the Eurodollar market occurred in 1957, when banks created a market in Europe where US dollar deposits were re-lent to European institutions instead of being reinvested in the United States. Eurodollars thereby came to be defined as deposits denominated in US dollars in banks outside the US. These kinds of deposits were later denominated in other currencies, and these ‘Eurocurrencies’ in general (Eurodeutschmarks, Euroyen, Eurosterling and so on) came to represent borrowing and lending outside the jurisdiction of the central bank issuing the currency in question.
The Eurocurrency market grew very rapidly, from around $14 billion in 1964 to over $2,500 billion in 1988.5 This largely mirrored the increase in international trade and investment that took place at the time, although it grew at a much faster rate; world merchandise exports, for instance, grew from $176 billion to $2,869 billion during the same period.6 In particular, US multinational corporations in Europe sought cheaper, alternative ways to fund their foreign expansion. There was a demand for new and innovative ways to borrow money, and compared with the US domestic interest rate markets, Eurodollars had a number of advantages. Banks and large corporations with relatively high credit ratings were the main players, meaning that they were perceived as less risky. The deals were often large, resulting in lower administrative costs. Because they circumvented some strict US regulations at the time, Eurodollars could be offered at lower, more competitive rates. This, coupled with the sheer size of the market, resulted in tighter bid–offer spreads. Should you, for whatever reason, decide to offset a trade you had previously done in the Eurodollar market, the cost of doing so was smaller than in the old-fashioned domestic money market.
Barometer of Fear Page 10