Morgan Stanley’s Pete Eggleston seeks the help of scientists to tackle problems like persistent illiquidity in bond and currency markets
Brain signals and white water rapids don’t usually turn up in the same sentence as banking. Unless you are talking to Pete Eggleston, head of Morgan Stanley Fixed Income Trading’s Quantitative Solutions and Innovations Group in London, who taps experts in fields such as fluid dynamics and neuroscience for practical insights on how to better navigate markets.
“People in the banking industry are smart, but we need input from other disciplines to generate the kind of ideas that are sophisticated enough to be useful for clients in today’s increasingly complex markets,” he says.
Eggleston and his team of “quants,” as quantitative analysts are called, build trading tools—algorithms and analytical suites called dashboards—that help bond and currency traders think about when to jump in or out of the markets.
Roaming the halls of Imperial College London and other universities comes naturally to him and his team, none of whom ever thought their science backgrounds would lead to banking. Eggleston, a former chemistry PhD student, has always kept in touch with old academic friends and talked shop with them.
Now he sees input from people in areas like biology, quantum physics and neuroscience as a necessity rather than a luxury addition to the process of building trading tools. “Existing financial theories and tools are important of course, but sometimes their use doesn’t stretch far enough to deal with the problems our foreign exchange (FX) and fixed income clients are now facing,” says Eggleston. “That’s why we need scientists – to help us think more laterally around a problem.”
Reduced risk-taking among banks and other changes since the 2008 financial collapse have irrevocably altered the trading conditions in global fixed income and currency markets, says Eggleston. “There are things traders call air pockets, where you just can’t get anything traded and the price gaps. That situation just feeds on itself and, before you know it, volatility can spike significantly and liquidity dries up, creating turbulence in the markets.”
Eggleston’s team is now actively seeking experts on the ebb and flow of rivers, because they think it might open a door to fully understanding how to develop analytics that can track moves in liquidity in real-time across a range of asset classes. "We might be able to find the answer by applying fluid dynamics theory,” he says. "If you think about it, trading volume in the markets is analogous to the flow of rivers, so it does make sense.”
The team first collaborated with scientists on a major project five years ago, when the first Greek crisis caused fixed income markets to seize up by triggering wild swings of risk-on and risk-off moves in the currency markets.
It all began after Eggleston met up with an old neuroscientist friend. “I was trying to track correlations between many markets, and the trouble I was having was that financial models are based on historical data, so by the time the correlations were discovered, the ability to trade ahead of a risk-on or risk-off occurrence was gone,” he says.
“I was out with my old colleague, who was telling me about how he was developing models that measured the correlation of the many different physical reactions that occur from brain impulses arising from a heart attack for instance. It dawned on me that he was struggling with a very similar problem to mine - we were both looking for causality across wide arrays of time in series of data with different frequencies.”
Eggleston was using classic financial methods to analyze correlations, but they were not very useful beyond comparing two assets at a time. The model the neuroscientists were developing was far superior, because it was able to track many different variants. Eggleston’s team started pouring reams of financial data into the neuroscientists’ number-cruncher. Ultimately, they built a warning signal, in the form of an algorithm, that gave a probability of risk-on and risk-off bouts in the currency markets.
Eggleston's team has developed an award-winning execution dashboard that uses traditional financial measures of liquidity—trading volume and the cost of trading—to help currency traders identify different liquidity regimes during the day. The dashboard breaks up the day's trading into four different types of liquidity “regimes” based on volumes and costs. It emits traffic lights of green, amber and red to help traders evaluate when and how to execute trades in real time.
To further develop the liquidity metrics and expand its use to include other asset classes, is one area where fluid dynamics might come into play, says Eggleston.
Put simply, fluid dynamics could provide the computational foundation for measuring liquidity changes in bond markets. A part of a river that’s deep and running smoothly is analogous to a market with plenty of trading volume and little volatility. Then there are rapids, where gushes of dangerous white water can suddenly lead to nothing but treacherous rock. That's similar to highly illiquid markets, where bad news can send would-be buyers of securities scuttling to the sidelines.
“The team only came up with the idea of looking at fluid dynamics a few months ago, so we could be wrong and there could be many other more applicable theories,” says Eggleston.
“But we've taken the first most important step, which is coming up with an idea."