F1 teams and bond traders both use tech to eke out small gains for big wins
Formula One races are won or lost through small decisions and small actions. Each decision to pit, brake or overtake adds or subtracts milliseconds that could mark the difference between victory and defeat. So, too, in bond trading a series of rapid decisions factoring price, liquidity and the competition add or subtract the basis points that add up to profit or loss.
Both merge human intuition and nerve with cutting-edge technology to eke out the small increments that add up to big gains. And for both, lapses in risk management can lead to catastrophe. Both trading and F1 use data analytics, artificial intelligence, cloud computing and the interoperability of computer and data systems to gain the miniscule advantages in speed and precision that together add up to success.
Evolutions in technology have driven innovation in F1 and in bond trading. As computers went mainstream in the ’80s, spreadsheets and Bloomberg terminals began appearing on the desks of bond traders and F1 cars moved from being purely mechanical to having electronic systems on board. By the mid-eighties, sensors on components of an F1 car could send radio bursts of information to the garage before the car entered the pit.
The ’90s ushered in the widespread use of the internet and, at the end of the decade, the first electronic bond trading platforms emerged. At the same time, electronic control systems began appearing in F1 cars and the subsequent increase in data collection required the use of faster and more powerful computers in the garage and the factory.
Today, AI, data analytics and cloud are helping F1 teams develop higher-performing vehicles and better race strategy. In 2021, there will be 23 races, each on a different track in a different country. And each track will place different demands on the cars.
Massive amounts of historical and practice run data are analyzed to make highly precise modifications to the cars before each race to ensure optimal performance. During races, machine learning and analytics help teams integrate real-time and historic data to monitor the performance of the car and driver and make real-time decisions on strategy.
The optimal timing of pit stops is one such decision. F1 drivers are required to make at least one pit stop during a race to change tires, make mechanical adjustments and clean the driver’s visor. The strategy and execution of these stops is crucial and can be the deciding factor in their finishing placement in the race. F1 teams aim for a pit stop time of two to three seconds. The record, set by the Red Bull team in Brazil in 2019, is 1.82 seconds.
So, too, many trades for bond traders — such as smaller or routine trades in competition with other dealers — are necessary but can take a trader’s time away from executing larger, more complex or exclusive trades that are more profitable.
But now AI, cloud and interoperability are making automated execution of up to 30 per cent of these trades possible with limited or no intervention from the trader. As with F1, speed is important, and cloud has provided the speed necessary to execute automatic bond trades in three seconds.
In F1 the car dictates the winner. Even the most talented driver will not win in an inferior car. It may soon be the case that even the best trader will be less profitable without the best tools.
In both F1 and bond trading, the conditions under which decisions are being made are constantly changing. Predictive AI can help inform racing teams and traders. Some argue that traders and drivers will be replaced by AI — but this is unlikely to happen any time soon. Predictive analytics can tell them what is going to happen, but not what to do about it. For that, human skill and judgement are still needed.