Why would a group of competing traders voluntarily talk to and trust each other? Apart from the obvious reason that human beings are social animals, I believe an important reason in the FX and money markets in particular relates to ‘reciprocity’, sometimes referred to as one of the oldest forms of human co-ordination.
The structure of the FX and money markets differs in many ways from that of the stock market. For instance, stocks are traded on an exchange and therefore ‘centralised’. The FX and money markets are decentralised, which requires co-ordination to take place in a different way. The stock market has set opening hours and is driven by orders coming into the market. The opening hours of the FX and money markets are more fluid and dependent on market makers’ willingness to quote prices. In the stock market, it is easy to work out what a trader’s or a bank’s inventory or exposure is at all times. It is possible to have zero holdings in Apple shares. A bank cannot, however, have zero exposure to ‘money’. In the stock market, investors and traders try to value companies, which is why insiders who can access secret or non-public information might be able to get an advantage over others. In the FX and money markets, it is extremely difficult to define what kind of non-public information could potentially move the market. In fact, the kind of insider information that could fall within this grey zone is also something that makes these markets unique. Importantly, the FX and money markets, despite often symbolising free and unregulated markets, have an inherent link to the state. All money market instruments have a natural anchor in the official central bank interest rate. For example, from 30 January to 21 February 2006, more than half of all panel banks submitted a three-month US dollar LIBOR rate of precisely 4.570 per cent on every single day. It could be that these banks had formed a conspiracy and that the actual LIBOR rate should have been different on at least one of those days. However, unless market participants expected that the Federal Reserve would change the official interest rate in the near future, and unless any changes took place with regards to liquidity or credit risk, a more plausible scenario is that banks simply submitted the same rate as the day before. The competing banks did not need to communicate with each other to achieve such price harmonisation. Simply knowing the prevailing central bank rate (which was public information) and being aware of what LIBOR quotes the other banks had submitted the previous day (which also was public information) were enough for this to happen.
Another difference that makes the FX and money markets special relates to secrecy and transparency. As we already know, this was particularly true with LIBOR, where individual prices were neither tradeable nor had to be backed up by evidence of any trading activity taking place at the submitted quote. Prices in FX and money markets cannot be compared directly with LIBOR, which is an interest rate benchmark. There are, nonetheless, important similarities. As the economists Joseph Stiglitz and Bruce Greenwald point out, ‘interest rates are not like conventional prices and the capital market is not like an auction market’.25 Ultimately, prices in markets involving borrowing and lending depend on judgements around creditworthiness and access to liquidity. Such markets, therefore, become highly dependent on relationships. It could be the relationship between a borrower and a lender. It could be the relationship between a bank and a central bank. It could also be the relationship between a bank and another bank, or a trader and another trader.
It is impossible to assess whether the market will go up or down, or whether the other trader or client wants to buy or sell, ahead of every price you quote. You have only a few seconds to think. Often, you have to rely on your instinct, which includes a psychological assessment of the situation at hand. However, it is also impossible to assess what the liquidity will be like a millisecond after you have quoted the price. What if you suddenly end up with a hot potato and, when you try to get rid of it, every single counterparty refuses to quote you a price? What if the market liquidity, to which you and the other market makers are contributing, for some reason suddenly dries up like a desert? There is not enough time to forecast what the liquidity might be like the next second, hour or day. You have to rely on your instincts and, crucially, your assessment of the other market makers’ instincts. As a result, the market tends to function as long as market makers are willing to quote each other prices. Banks are market makers and liquidity providers in the FX and money markets. However, in order to be able to be market makers and liquidity providers, banks require other banks to also be market makers and liquidity providers. It is a game based on reciprocity and trust.
There are few written rules about reciprocity and trust; rather, they are more the widely accepted norms – something that is ‘the right thing to do’. The ‘run-throughs’ I mentioned earlier are a perfect example of such a norm. One of the first things I learned when becoming a market maker in FX swaps was to interpret and trade on, but also provide, run-throughs.
A sequence of numbers shouted down the speaker box by a broker first thing in the morning – ‘115-210-280-510-690-825’, say – would indicate a series of FX swap bid prices in a specific currency pair for different maturities that could be dealt on. It meant that another market maker had kicked off that day’s market by generously providing an interdealer broker with a set of prices that could be passed on to all their clients, which invariably included all the other market-making banks. The prices were valid for only a few seconds, which meant that, as a trader, you had to be quick if you wanted to trade on any of them. If the broker shouted ‘Off, one-month given at 115!’ it meant that someone else had been quick and had dealt at 115, after which the prices could no longer be dealt on. The original market maker might then provide a new, perhaps different, run-through. Or maybe another market maker would step in to have a go. There was no requirement to provide run-throughs. However, it was a way to support or make a market that essentially did not exist until prices were quoted in it. To provide prices – in other words, liquidity – in a market in which the banks were actively involved, without having any guarantee that others would do the same, was seen as the right thing to do. If you didn’t do it, why would anyone else?
Reciprocity and trust are endorsed by the trade organisation Forex. The latest version of the ACI Model Code, published in February 2015, states that ‘bilateral reciprocal dealing relationships are common in the OTC markets and often extend to unwritten understandings between Dealers to quote firm two-way dealing prices’. Further, the Model Code states that such informal reciprocal dealing relationships are to be ‘encouraged’ and are ‘a logical development in the OTC markets and play an important role in providing support and liquidity’.26 The trade organisation (whose motto is ‘Once a dealer, always a dealer’) sees such informal agreements as natural, to be encouraged in order to maintain trust, reciprocity and liquidity in the market place. However, it does not state specifically how these informal agreements ought to be achieved. The logic not only seems to encourage mutual understandings between competitors, it also turns a blind eye to the fact that trust and reciprocity are difficult to obtain unless market makers are allowed to talk to each other. How do you build trust in an environment where you cannot communicate with one another?
The bid–offer spread illustrates this well. In economic theory, a market maker should take into account various costs when quoting a two-way price. A volatile or uncertain market, for instance, should imply a wider spread, as it might be more costly to close out a position in such a market. The same goes for an illiquid market: the more such potential costs are added, the wider the spread should be. At the same time, competition between market makers should force this spread to be as tight as possible at each moment in time.
The question is whether this theory works in practice. During the late 1990s, three academics conducted a survey among more than 500 (mostly rather senior) FX dealers in Hong Kong, Singapore, Tokyo and New York, and asked what the main driver was when choosing the interbank spread.27 Interestingly, fewer than one-third of the respondents argued in favour of ‘poten
tial costs’, whereas two-thirds went along with the ‘market convention’. The main reason for following the market convention appears to have been to maintain an ‘equitable and reciprocal trading relationship’. Other important drivers were ‘market image’ and ‘firm policy’. Fewer than a tenth of the respondents claimed that ‘trading profits’ were the main driver for following the market convention. Unfortunately, following the FX scandals, it might be difficult to conduct a similar survey today. Nonetheless, the responses demonstrate how FX and money market makers have felt a ‘sense of duty’ towards the bank and its competitors with regards to liquidity provision. Traders are influenced by the prevailing market practice.
Indeed, the number sequence ‘115-210-280-510-690-825’ in the example above contained only bid prices. Often, but not always, the prevailing bid–offer spread was presumed to be known. The broker might even add ‘with standard bid–offer spreads’, confirming that the bank which had provided the prices assumed that all competitors knew which spreads were to be applied. A bid price of 115 might mean an offer price of 120, assuming the bid–offer spread was 5; ‘5’ had become a convention that was supposed to last until, for some reason, it broke down.
Violating the market convention, for instance by suddenly quoting a very wide bid–offer spread or not quoting a price at all, would have breached a (mostly unwritten) policy outlined by management to support the market in a certain way. It might simultaneously have tarnished the image of the trader, the desk and the bank. Violating the market convention would also have harmed the relationship with other banks and their traders, and might ultimately have resulted in sanctions or retaliation. Such sanctions could involve more limited access to liquidity in the market. This, in turn, would have led to less profit, less client business and possibly even a withdrawal from the market-making function. Violating the market convention was like shooting yourself in the foot.
Or so I thought. Towards the end of my trading career, a competitor suddenly decided to double, and then quadruple, the spread they quoted me. I did not want to retaliate. But within hours, every other market maker had followed this new ‘convention’. As a result of the extraordinarily wide bid–offer spreads, the market virtually died. This was bad news for me, but good news for some of the other market makers who had the opposite position to me. I was furious. But there was, of course, nothing I could do about it. Market conventions were not regulated. And market conventions did not necessarily reflect what would be best for the market as a whole. As with most things in these markets, the conventions were ultimately decided by relatively few, yet powerful, banks.
***
When, in 1993, I was taught to trade T-bills and bonds, there was no official rule book on how to call out for Swedish T-bills. The guidelines were more formal than children’s playground rules, but the enforcement mechanism if or when you broke them was anything but. A year later, when I quoted my first price in the FX market, I received similar guidelines and instructions from more senior traders. However, there was no involvement by any authority when it came to settling disputes. Instead, this was based on the overriding principle that the game was ‘tough’ but that the players would keep it ‘fair’. A player quoting slow or wide prices could expect slow and wide prices back. A player refusing to quote amounts agreed by everybody else could expect to find it difficult to close large positions in the future. Invariably, a player not adhering to the social norms could expect to receive less valuable information, be treated with less consideration and occasionally even be excluded from social events. The enforcement mechanism was based on retaliation – or, more precisely, the fear of retaliation.
‘Individuals may participate in social norms in part because of an expectation that others will also participate,’ Dean Harley wrote in the California Law Review.28 Richard McAdams, also with a law background, defines social norms as ‘informal social regularities that individuals feel obligated to follow because of an internalized sense of duty, because of a fear of external non-legal sanctions, or both’.29
There is nothing inherently ‘wrong’ with social norms – think of shaking hands, bowing, saying sorry, please and thank you. Just as there is nothing wrong in theory with a market maker calling a competitor for a price in an 11-month foreign exchange swap either. It is just that, for most currency pairs, you are supposed to stick only to certain maturities, the closest in this case being nine months or 12 months.
Similarly, there is nothing wrong with a London-based market maker calling a London-based competitor for a price after 4 p.m. It is just that you are supposed to respect local opening hours (the London close, the New York open, etc.) even if the market trades 24/7.
Nor is there anything wrong with a market maker calling a competitor for a price in Swiss francs against US dollars. It is just that you are supposed to ask the other way round: US dollars against Swiss francs.
There are many more such social norms and market conventions that, for example, govern how fast you are supposed to quote, the language you are supposed to use, how often you are supposed to restrict yourself from calling other market makers, how you are supposed to treat interdealer brokers or how large each trade ticket is supposed to be. With few formal rules and regulations, and the constant uncertainty in financial markets, such conventions become extremely important. Market conventions act to settle disputes quickly and prevent anarchy in an environment that otherwise would be very susceptible to collapse. Market conventions also allow expectations about the future to be harmonised and to become somewhat orderly – at least to a degree. However, it would be a misunderstanding to think that the sheer size of the FX market, or indeed the LIBOR-indexed derivatives market, automatically means that it is competitive. It would also be a misunderstanding to believe that the whole market has been involved in shaping its social norms, informal rules and practices over the years. Banks have always been in the driving seat. Market conventions might, of course, change at any time. However, if they become attached to the daily trading routine, or part of the glue that binds the industry together, participants in the market become increasingly confident that they are relevant and valid.30 They become part of the ‘culture’, to borrow a term from the UBS press release following the revelation of the bank’s involvement in the FX scandal.
A common theme with the markets and benchmarks that have been the subject of scandals in recent years is that they have largely been unregulated and lacked transparency. This, however, does not mean that they have lacked ‘rules’. It is just that many of the social practices and conventions have evolved without intrusion by governments or regulators.
When I was having the previously mentioned conversation with the managing director about the possible ramifications of uncovering the truth of the FX market, he continued by saying: ‘On the other hand, the biggest manipulators of all are the central banks. They bully the market all the time.’
The fact that central banks, which are ultimately responsible for printing a country’s money, could use their foreign exchange reserves to intervene in the market, argue in favour of capital controls to prevent FX trading from taking place, or influence exchange rates by changing the interest rates or issuing biased press releases, he saw as evidence that the market would never be subject to a thorough investigation. If the banks were to be found guilty of putting grains of sand into the machinery of the free markets, surely the central banks should take some blame? If this ended up being classified as a conspiracy, surely it would go all the way to the top? The central banks might not have had a vote in deciding how the market conventions evolved, but given their importance as stakeholders in the markets relating to the concept of money, they were often invited as ‘observers’.
Whenever I was asked to fill out a survey or questionnaire by a central bank, or if a central banker called me directly to ask my opinion about the market, I felt special. Being selected to ‘help the state’ because of my position as a market maker engendered a sense of pride and duty. Whatev
er I was doing as a job, it was of great importance. Despite the fact that the vast majority of trading was driven by speculation, it served as a reminder that my role as a market maker actually meant something. The state wanted, even needed, people who provided liquidity to the markets.
***
Although banks have received very large fines for their traders’ behavioural failings in the unregulated FX market, it seems like such behaviour was widespread. The conspiracy did not involve just one or two banks, but a group of banks controlling more than half of the global FX market. The conspiracy did not involve only junior and inexperienced traders, but senior traders, chief dealers and global heads of FX trading. More than any other market, the FX and money markets are quite literally about money. Banks, dominant in these markets, are also secretive by nature. Money, secrecy and traders are three ingredients that, when mixed together, can end up in a conspiracy.
Very few people (academics, regulators or others) bothered to look into the FX trading culture before the scandal broke. This might seem strange, given how large and important the market is. On the other hand, I think one reason it received so little attention was precisely because it was so large. From the outside, it seemed too global, too competitive, too efficient – simply too perfect to have space for something that could be referred to as a ‘culture’. As in the case of LIBOR, this perception was false.
‘People of the same trade seldom meet together, even for merriment and diversion, but the conversation ends in a conspiracy against the public, or in some contrivance to raise prices,’31 Adam Smith wrote more than 200 years ago. The contours of a conspiracy, however, do not need to take shape in a smoke-filled cigar club or an electronic chatroom. Conventions, whether they later are classified as conspiracies or not, can also evolve on the desk, during a dinner, a seminar, a cigarette break or perhaps even in a classroom.
Barometer of Fear Page 21