Could Financial Trading Be The Model Of How To Use AI Effectively?

There is a lot of fear around AI, but financial trading could be a bellwether for using it properly


The financial trading sector has been demonized over the past 30 years. We have even seen it in works of fiction where some of the most evil and depraved characters have worked in the sector. For instance, Patrick Bateman, the psychopath from Brett Easton Ellis' 'American Psycho' naturally worked for a fictional investment firm. However, the sector has done nothing to stop the stereotype of money-obsessed clones with no morals, especially as the banking sector nearly took down the entire world's economy in 2007 and is hence seen as single-handedly responsible for the surge in dangerous right-wing politics that has seen a terrifying increase in hate crimes and discrimination in the world.

They are seen to have managed to keep hold of their jobs because they are the 'experts' who know how complicated financial transactions need to be done, how to pick the best stocks, how to make the best investments, etc. The behavior of many in the sector does little to gain them respect amongst the general public, though, and there are several cases which have further dented financial trader's image. For instance, in 2017, an HBOS banker was jailed for a £1 billion fraud in the UK, Bernie Madoff was jailed for 150 years for a multi-billion dollar Ponzi scheme, and a London City trader was sentenced to 14 years in prison for manipulating the Libor rate. It is safe to say that despite the perception that they are needed, there is little love for those working within the banking and financial sectors amongst the majority of the world's population.

However, there is a common belief that this may soon be a thing of the past as AI is increasing in power and the kind of analysis-focussed jobs that traders do seem like prime examples of roles that are ripe for replacement by the technology.

For instance, AI goes against the idea that people like Carl Icahn, Ray Dalio, and John Paulson are the best options for magically turning pennies into billions of dollars. There is certainly something to say about gut feeling within investments as each are worth billions, but when you consider that each also manages funds that have performed poorly over the past three years, it seems that gut feeling is no longer good enough in today's data-driven and second-by-second investment environment. For instance, Carl Icahn's fund dropped by 18% in 2015 and 20.3% in 2016, whilst John Paulson's company Paulson & Co went from managing $36 billion in 2011 to managing only $10 billion in 2017. This compares to an AI algorithm invented by John Alberg, co-founder of hedge fund firm Euclidean Technologies, and Zachary Lipton, a researcher at Amazon, which generated a 17% annualized return, which is even better than the 14.4% produced by a standard statistical model.

AI also has huge potential within regulations, reporting, and fraud detection - things that banks have been criticized for in the past. Criticism for this is not necessarily a slight on those within banks who's job is the regulate what's happening internally, but instead just the fact that over the last 30 years the digitization of the world and huge increase in digital transactions has meant that there are now billions of transactions every day. Trying to accurately identify the fraudulent ones is not something that a human or even basic AI system could even attempt. However, there is also a huge amount of internal monitoring that needs to take place as the financial industry is one of the most heavily regulated in the world. Given the huge increase in digital communications within the workplace, platforms like Skype, Slack, and traditional emails all create recordable and analyzable data sets for AI to work through and notice questionable practices.

We often think about automation and AI impacting those with relatively manual or low paying jobs, for instance factory workers, translators, or call centre workers, but the reality is that the pure numbers-driven nature of the finance industry means that AI is perfectly suited to understanding it better than almost anybody currently working within it today. Decisions can technically be made better and faster whilst taking in infinitely more information by AI than humans, with feeds from markets, past data, similar market conditions, and even social media being analysable in seconds.

However, despite the huge potential for AI, it is only acting as a tool used by humans to help them in their roles. There are hundreds of financial companies who are investing in AI and deep learning technologies, but there hasn't been a huge layoff of traders. This isn't to do with an increased loyalty to their workforce compared to other sectors, as the financial sector has been historically willing to lay off staff, with Citigroup laying off 50,000 staff in November 2008 (the second largest layoff in history). Financial institutions are the one place where profits are guaranteed to be put above all else, so it perhaps shows that they see more value in using AI as a tool rather than a replacement for their workers.

There have been funds created with AI at the core of their investment strategy who still don't use it without major human influence. For instance, Babak Hodjat, co-founder of Sentient Technologies ,who invest using AI, has said that their machine learning techniques tend to find patterns within their data that don't actually hold up to real-world scrutiny. The closest to a full AI-driven model is Cerebellum Capital, which uses AI to create and assess investment strategies, but even then the actual trades are still done by humans.

AI being used as a tool is something that we are likely to see across multiple companies because it makes considerably more sense than simply replacing people with AI. Customer service call centres are a prime example; customers despise not having the ability to talk to a human on the other end of the phone when they have an issue with their order. So, the company who pulls the trigger on a full AI-driven customer care system may save some money on salaries, but that is likely to be offset by an increasing number of customers refusing to use their service. Instead AI can be used for screening or answering questions rather than dealing with complaints. In manufacturing, AI can be used to remove the menial jobs for workers and allow them to do the more complex or dexterous work.

Ultimately, AI has the potential to do huge harm to people's livelihoods, but only if we adopt in specific ways. Ironically, those in the financial sector, which is perceived as being immoral and profit obsessed, are the ones who may be showing other industries how to use it properly. How long this continues to be the case remains to be seen as the technology is developing at a rapid pace, but for the time being, the industry seen as one of the most immoral is acting as a bellwether for AI use. 


comments powered byDisqus
Data flow small

Read next:

Let The Data Flow