In 2017 UBS, a Swiss investment bank, sold off their trading room in Stamford Connecticut. Their trading room was among the largest in the world and at its peak housed over more that 5000 traders [1]. Some of you might correctly point out that between 2006 and 2016 the subprime mortgage crisis happend, and you’d be right. However, that is not the only reason why dealing rooms are becoming less crowded. What eventually happend to UBS’s jewel illustrates another movement at dealing-rooms around the world, the rise of algorithmic trading.
When I reach out under my desk I find a remnant of the past, an old black phone trough which orders where manually placed on global markets. I look over my desk I imagine what this room would have looked like 15 years ago, however, in reality I find myself surrounded by algorithmic trading platforms and automated order systems.
When looking at other big financials we see that algorithmic trading becomes more and more standardised in an attempt to increase margins on trades as regulations strongly increased in their complexity. Last year JPMorgan, the worlds largest investment bank by market capitalisation [2], successfully launched their digital robot to execute trades; LOXM.
LOXM is one of the first fully implemented machine learning based bot to execute trades. JPM’s bot uses historical trading data in its decision making process, and continues to learn with every trade [3]. Making it unique trading tool in the industry. LOXM was designed to beat the industry’s benchmark by leaving behind rudimental algorithms by shifting towards deep reinforced learning that is capable of executing more complex trades. [4].
Still I am a bit sceptical for these bots. You probably have seen that video where two Alexa (Amazon’s Siri) communicate with each other [5], see video below.
I wonder what kind of situation the global capital markets create if banks and other financial institutions decide to completely shift towards implementing AI bots to track markets and execute trades. Will we see Alexa like situations?
1: https://nypost.com/2017/04/19/ubs-has-officially-ditched-its-massive-trading-floor/
2: http://banksdaily.com/topbanks/World/market-cap-2018.html
3: https://www.ft.com/content/16b8ffb6-7161-11e7-aca6-c6bd07df1a3c
4: https://knect365.com/quantminds/article/10d3b420-fe65-4269-b1da-ab555a509958/the-latest-in-loxm-and-why-we-shouldnt-be-using-single-stock-algos
5: https://www.youtube.com/watch?v=8zR0Hxojce4&t=65s&frags=pl%2Cwn
