Barometer of Fear

by Alexis; Stenfors

  ***

  My attention shifted more and more from the Nordic countries to Japan. There were two reasons for this. First, the Japanese yen and the Canadian dollar happened to be part of the ‘Scandi’ desk at Citibank in London. I still don’t know how the two currencies, which were so vastly different, had ended up with us, but the desk setup required me to follow what was going on in both Japan and Canada. Second, and more importantly, Japan was heading towards a financial crisis.

  Up until 1995, whenever a Japanese bank was in trouble, the government had intervened by arranging the merger of an insolvent bank with a sound bank. With this framework, the Japanese banking sector was perceived to be safe by financial market participants. In August 1995, however, the government let Hyogo Bank default. The bank had $37 billion worth of assets and it was the first bank failure in Japanese history. Suddenly, the perception of Japanese banks changed.1 Others became reluctant to lend to them. The Japanese banks were massive, and a number of them had previously embarked upon ambitious expansion programmes abroad. As they typically did not have retail networks outside Japan providing them with deposits, much of their foreign currency borrowing was done in the interbank market. In the Japanese yen market, this was not much of a problem, as Bank of Japan, the central bank, ultimately was able to provide liquidity. However, they also needed foreign currency, and US dollars in particular. Bank of Japan could not print US dollars, only the Federal Reserve could.

  So the Japanese banks came to us in the interbank market. Lending directly to them was out of the question. The credit lines had been either filled up or withdrawn. And it seemed like every single foreign bank was in the same situation. As the Japanese banks struggled to borrow dollars directly, they turned to the FX swap market instead. By entering into agreements to buy US dollars against yen now, and simultaneously to do the opposite at a predetermined date in the future, they effectively ‘borrowed’ US dollars that they needed and ‘lent’ Japanese yen (which they had or could obtain).

  With Japanese traders now rushing to the FX swap market, for the first time I realised that this market could deviate quite substantially from what theory said. Basically, the banking crisis in Japan seemed to have messed up the mathematical equation. Theory said that the US dollar LIBOR should reflect the interest rate level at which banks were lending US dollars to each other. This might have been the case for most banks, but it certainly did not apply to the Japanese banks. They had to pay a significant premium to access the US dollar money market, and this premium was reflected in the FX swap prices we quoted day in, day out.

  I was instructed to watch out for a range of four-letter codes calling on the Reuters Dealing 2000-2 machines: SUMG (the dealing code for the FX swap desk at Sumitomo Bank in Tokyo), TMFT (Tokyo-Mitsubishi), IBJK (Industrial Bank of Japan), SNWT (Sanwa Bank), and so on. Whenever one of them called on the machine, they had only one purpose in mind: to trade on the price I quoted. And as a market maker, I had to quote them a two-way price almost immediately. If I let the phone ring (or in this case the machine beep) for half a minute or so, I would automatically break one of the many unwritten rules in FX trading.

  After having clicked on one of the incoming calls, the first three lines of a conversation could look like this:

  # 3M USD/JPY 100

  # -140/-138

  # -140

  The trader at the other bank first asks me to quote a three-month FX swap price for $100 million against Japanese yen. I then quote -140/-138 (a bid and an offer). The trader deals at -140. Within seconds, the remaining parts of the transaction are agreed and confirmed, such as the FX spot rate used to work out exactly how many US dollars and Japanese yen will be shipped back and forth (or, put differently, the actual interest rates at which we have borrowed from and lent to each other), as well as the payment dates and bank details. The short conversation finally ends like this (depending on ‘fat fingers’ and the keyboard shortcuts used):

  #THANKS AND BYE

  #TXBBIIB

  #END LOCAL#

  Given the state of the Japanese financial system at the time, the Japanese banks needed to get hold of US dollars and so were constantly looking for bids in the FX swap market. The more they traded (or were expected to trade), the lower the bids became. They became easy to ‘read’. Rather quickly, a two-tier market evolved. Non-Japanese banks faced no funding issues, so their market continued to function smoothly. Japanese banks, however, had to pay an additional premium in the FX and money markets. This extra cost later became known as the ‘Japan premium’ in academic journal articles and economics textbooks.

  Japan also happened to have a unique derivatives market, with not one but two money market benchmarks: LIBOR and TIBOR. LIBOR was set in London, mainly by non-Japanese banks. TIBOR, however, was set in Tokyo and its panel was mainly composed of Japanese banks. Some derivatives were indexed to LIBOR, others to TIBOR, with some clients preferring LIBOR, others TIBOR. Before the crisis, it did not really matter which one you used, as they tended to be almost exactly the same every day. Now, this symmetry was distorted as the Japanese banks had to pay a premium to access funding. The ‘TIBOR–LIBOR spread’ (the difference between the two benchmarks) turned into a kind of barometer of fear in relation to the Japanese banking system. It was possible to put a number on this fear, and this number went up and down. More than this, bets could be put on this number.

  After a series of bank capital injections by the Japanese government, the Japan premium more or less disappeared around March 1999. The market slowly began to return to normal. However, the atmosphere on the STIRT desk was now different. The launch of the euro meant that a massive business could be built around a brand new currency. Still, this did not compensate for the fact that a number of currencies had disappeared from the face of the earth. If you had been a specialist in one of the major currencies, such as the deutschmark, the French franc or the Italian lira, you would probably fit in. If your career had been built around the Austrian schilling or the Portuguese escudo, on the other hand, your future was much more uncertain. However, the birth of the euro had been preceded by financial crises not only in Japan but also in South Korea, Russia, Malaysia, Indonesia, Thailand and the Philippines. The newly formed ‘emerging markets’ desks needed experienced traders and currency experts in a range of markets. A brilliant Dutch guilder trader sitting opposite me became a Malaysian ringgit expert almost overnight. I remember talking to him about ‘the old KLIBOR’, after he had informed me that Kuala Lumpur actually had its own LIBOR. The Finnish markka faced the same euro destiny, but it had not been my main book for several years now. My key focus was on Japanese yen, US dollars and the other Nordic currencies.

  Some of the derivatives traders I used to hang out with after work decided to return to Citibank Sydney. I also felt that it was time for a change and asked whether I could be transferred to New York for a couple of years. My manager was supportive of the idea and plans were drawn up. Over lunch at Christopher’s in Covent Garden in October 1999, however, he told me that the situation had changed. I could either wait a few years, hoping for a posting to New York, or I could become chief dealer in Tokyo now. Although I was hesitant about getting management responsibility, I was truly excited by the prospect. For as long as I could remember, I had been interested in foreign languages and cultures. I had visited Japan once and loved it. Two weeks later Maria and I were looking at apartments in Tokyo.

  ***

  During the following eight years, many things happened. I became a father, I returned to London, I joined Crédit Agricole Indosuez and I lost my father. The financial markets became much more sophisticated, bigger, more competitive, perhaps more ruthless. And as the derivatives markets expanded, the importance of LIBOR grew exponentially.

  Every day arrived with a precise timetable for economic data releases. It could be the UK inflation number, the US unemployment rate, the Japanese retail sales figures or the Australian GDP (gross domestic product). These numbers would
have an impact on how central banks shaped their assessments about the economic outlook. This, in turn, would influence their monetary policy strategy. At the end of the day, it was about predicting if, when and by how much the central banks would change the official interest rate in the future, because this would affect LIBOR by roughly the same magnitude. We watched every step the central banks took, scrutinised every word they said – anything that could provide a clue to the future direction of LIBOR.

  The interest in LIBOR, however, was mutual. Just as much as it mattered to traders, it mattered to central bankers. This is why. Assume that the UK inflation rate falls, and the Bank of England therefore decides to lower the base rate (the official interest rate used for lending to other banks) to meet its inflation target of 2 per cent. The interest rate cut is supposed to be transmitted immediately to the interbank money market rate (where banks lend to each other). Unless people think that the Bank of England will change its mind and reverse the rate cut the following day (or week or month), banks will probably decide to lower the one-week rate, the one-month rate, the three-month rate, and perhaps even interest rates with longer maturities. Then, as banks compete with each other for customer business, they lower the interest rates they charge for variable rate loans, overdrafts, and so on. Rates offered to savers are also lowered and mortgages might get a touch cheaper. As the lower interest rates gradually filter through to the real economy, inflation begins to creep up towards the Bank of England inflation target of 2 per cent.2 This process, the ‘monetary transmission mechanism’, can be seen as the channel through which a specific interbank money market rate – the three-month rate, say – is generated. The central bank does not determine the three-month interbank money market rate. The banks do. However, the central bank has considerable power to influence it.

  How do banks know what the Bank of England base rate is? The answer is simple: it is on the front page of the Bank of England website. Whenever it is changed, a press release is sent out and press conferences held. The Bank of England even announces a calendar with exact dates when they might consider changing the base rate. Over the years, this process has become increasingly open and transparent. How do central banks know what the three-month interbank money market rate is? This is more complicated, because trades between banks are secret. Instead, LIBOR is used as a gauge for this interbank rate.

  During the years prior to August 2007, changes – or expected changes – in the official central bank rates filtered through smoothly to LIBOR, lending support to the assumption that a central bank rate change leads to a proportional change in interbank money market rates. Central bankers, having grown accustomed to a seemingly transparent and well-functioning money market more or less without credit and liquidity issues, could rely on the first stage of the monetary transmission mechanism. As LIBOR could be relied upon, focus could be put on the channels affecting the real economy, and on new methods to minimise monetary policy surprises and to increase transparency.

  In the old days, central bankers often flexed their muscles by taking the markets by surprise. As a trader, it was difficult to figure out when they were meeting, what they were saying, what they were thinking. A radical shift was now taking place, spurred by influential academics arguing that transparency was a good thing. Step by step, central banks began to lay their cards on the table. Their meeting schedules were announced up to a year in advance, meaning that surprises were unlikely to happen in between. The minutes from the meetings were published, helping traders understand what they were saying. The identities of the people attending the meetings were revealed, making it easier to grasp the composition of their monetary policy committees. Who was a ‘hawk’ (advocating higher interest rates) and who was a ‘dove’ (preferring lower interest rates)? How hawkish were the hawks and how dovish were the doves? Some central banks even went so far as to publish their own interest rate projections and the probabilities of this actually materialising.

  And, as central banks gradually became more and more open, the attention paid to these meeting dates became intense. For a STIRT trader like myself, the meetings became exciting ‘events’. Everyone had an opinion about what would happen, and bets were placed in the financial markets accordingly. Generally, there was a consensus about the direction of the next interest rate move by the central bank. However, if they were going to hike rates, it could happen now or they could wait until the next meeting. If everyone shouted ‘Unchanged!’ after the newsflash appeared on our screens, the rest of the trading day became an anti-climax. Those who had put bets on nothing happening could book some profits and those aiming for some action had to lick their wounds. An unexpected rate move resulted in chaos, with traders trying to make sense of it all. Larger profits or losses were booked on such days. On some rare occasions, traders also disagreed on the magnitude of the potential rate change. ‘Fifty!’, referring to an unusually large 50 basis point change in the official central bank rate, could result in tears of joy and despair across the dealing rooms.

  During the final minute before the anticipated press release, my heart rate would increase. Hearing nothing but my own heartbeat, fuelled by numerous cups of triple espresso and cans of Red Bull, it could be unbearable. My hands would be too soaked in sweat to control the mouse, so I would go to the bathroom and wash them with extra soap. I would then obsessively tidy up my desk to make sure everything was in the right place: the computer screens, the two telephone sets, the broker microphones, the calculator and the blotter showing the trades I had done so far that day. I made sure I had a few extra blank sheets printed out in case the afternoon turned out to be messy. I would mentally prepare a list of banks to call once the announcement had been released, and what kind of prices I would ask them for. Because I knew that once the bright red newsflash appeared, there would be no time to read the full press release by the central bank, let alone the whole 50-page monetary policy report. Within a second, my switchboard and Reuters Dealing machine would light up like Christmas trees with incoming calls. It was the ultimate adrenaline rush.

  ***

  The fact that there was an important theoretical difference between the interest rate at which banks traded with the central bank and the interest rate at which banks traded with each other was gradually ignored.

  This difference, the ‘money market risk premium’, has two components – in effect, taking account of the fact that central banks can print money endlessly (which private banks cannot) and that central banks are annexed to the state and therefore are unlikely to go bust (whereas private banks can default). The first component is liquidity risk.3 Liquidity is normally referred to as the ease and speed with which an asset can be turned into cash. For instance, a painting by Picasso is relatively liquid, as Sotheby’s or Christie’s would probably be able to sell it easily and quickly should you want to put one up for auction. The artwork by your cousin who just finished art school, however, is likely to be illiquid. That is not to say it is worthless – simply that, in order to get a decent price for the painting, it would take longer and be more difficult to find a buyer for a relatively unknown talent. The same principle applies to the money markets, and can take two forms.

  Market liquidity risk is the fear of not being able to borrow back the money once you have lent it (or sell back the painting once you have bought it). It is also the fear of causing the market to panic after having borrowed only a fraction of what you need (imagine desperately having to sell 100 Picasso paintings). It could also be the fear that a temporary rise in the interest rate turns out to be long lasting – that, for some reason, the market becomes too slow to return to some kind of normality. A common way to quantify the market liquidity risk is to look at the bid–offer spread (i.e. the difference between the price quoted for the sale [bid] and purchase [offer]) of the ‘thing’ that is being bought and sold relative to the price of that thing. Anyone who has tried to resell a used car will have discovered that the bid–offer spread of used cars is relatively wide. Unles
s you are an expert in the business, buying a used car from one dealership will most likely result in a sizeable loss if you change your mind and go to another dealership looking to sell it the next day. The same goes for foreign currencies you buy at airports or in popular tourist destinations. Doing exactly the opposite transaction at exactly the same time is generally an extremely expensive exercise – perhaps not in total money lost, but definitely in relation to the amount you are exchanging. In this respect, the various financial instruments in the money, FX and derivatives markets between banks and between banks and large corporations tend to be very liquid and the bid–offer spreads are tight.

  Funding liquidity risk, on the other hand, represents something different. It is the ease with which a bank can obtain funding from others. It represents the fear that the lenders will suddenly require some kind of collateral (that you might not have). The fear that it will become more costly or impossible to roll over short-term borrowing. Or, in the absolute worst-case scenario, the fear that depositors will begin to queue up in front of the ATM to withdraw funds because they think you are about to go bankrupt.

  Even though it might be useful to distinguish funding liquidity risk from market liquidity risk, it can be very difficult to separate them as they can be highly interconnected. If there are problems with funding liquidity, often there are also issues with market liquidity. However, the reverse does not need to be the case. If the market dries up – ahead of the release of important economic data or a central bank policy announcement, for instance – the market can be regarded as illiquid temporarily, but without having any real impact on funding liquidity. Even the most liquid markets tend to be ‘dead’ during the World Cup final – this is not to suggest that concerns should be raised regarding the health of the financial system during those 90 minutes. Funding liquidity issues, however, do tend to automatically affect market liquidity. For decades, recurring year-end cash squeezes have been common even in liquid currencies such as the US dollar and the Swiss franc. This type of liquidity risk can happen regularly, even quarterly or monthly, but they need not be serious for the financial system as a whole. If the events are well anticipated (which they sometimes are), central banks can simply inject ample amounts of liquidity to maintain financial stability.