FX Traders Are Being Replaced by Algos - Bloomberg
https://www.bloomberg.com/news/features/2024-07-0…;
How does this effect the S&T space as a whole in the coming years and specifically, how will it effect the hiring of junior traders 2026 and beyond? (Assuming this model is the weakest we will ever see and it only gets more robust from here)
Tangentially here's one from the field of math:
https://x.com/Thom_Wolf/status/1809895886899585164
https://x.com/JiaLi52524397/status/18088868801648…
The next question that arises is how 'safe' the quant jobs will be in the future.
Banks aren't letting third party AI models go near their shit until they have confirmation of security. The adoption of technology by financial firms is historically, very slow.
The issue of security does arise but there are two points that come to mind, would a firm be incentivized not to blow caution to the wind when having the prospect of not only making incredible profits but also driving down manpower costs? Secondly, a lot of the AI models are now having big tech companies step in as their 'patrons' for the lack of a better word (Microsoft for OpenAI, Amazon for Claude and well Google and Facebook doing their own thing) which would help to assuage the fear of data leaks to an extent.
The topic of security is also quite wide as it doesn't just come in the form of data leaks. The increased advancement and utility of these models will increase the likelihood of fraudsters or criminals using them for their own benefit.
There are also risks through corporate espionage or ransoms whereby a company's model, which is otherwise automated, is "infiltrated" through GANs. These can constantly feed the wrong information into the models that banks have created, leading to ransoms being wanted etc. It's rare but could happen, and you definitely wouldn't want to be the first - perhaps another -1 for these models replacing quants.
In this case, having patrons like the ones you mentioned would probably mean nothing to banks and their customers. An Amazon data leak in comparison to highly utilised financial algos being penetrated are worlds apart, especially in money terms (small fine for Amazon, huge opportunity cost for an algo using bank).
It's all speculation, really, but these things could happen.
It's an interesting topic for sure.
There's never going to be any conclusive answers, really. Quant jobs are as safe as any other job that can, theoretically, be performed by AI/ML. The issue that companies will face, for the time being at least, is whether or not they can really trust AI/ML to perform the same job as an actual person.
Algo's are quicker, sure, but you might soon face repeats of the 2010 flash crash and the repercussions that came with it.
Who knows what regulators will have to say about this, if anything at all. They could definitely place more stringent rules on these companies to protect investors and customers but, just like with the flash crash, regulations just meant companies got more creative, so that could be a +1 for an increased need for people.
We won't know for sure until someone creates a conclusive algo that can replicate human behaviour and, even then, humans will still be crucial to the role in one way, shape or form.
This is a very insightful answer! To the point of having another flash crash the bloomberg article says this:
'Chameleon, on the other hand, is better for executing larger transactions and allowing them to slowly dribble into the market without creating larger gyrations.'
According to the tinfoil, this algorithm should prevent the so-called 'fat-finger' error that caused the crash in 2010.
That's a good way of them dealing with it, I suppose. Takes away the uncontrollable speed element of it. I think this still would require constant monitoring by a qualified individual, though (if they're sensible)
There will only be so many qualified individuals required to oversee the models (the old adage of having 1 man do the work of 5) which led to question in the post of how the hiring for junior traders will be impacted in 2026 and beyond. From face value it does seem like we are looking at a drastic and permanent contraction of the trading side of the industry at least a la automobile manufacturing.
Minus tenetur quasi voluptatem repudiandae alias earum incidunt. Similique commodi laboriosam sequi est architecto accusantium eos.
Non odio qui non ipsam voluptas provident. Perferendis voluptatem cumque libero sunt odit. Distinctio id temporibus voluptas aut nisi dolorem eos qui. Numquam delectus quo maiores est rerum sit ducimus.
Quam optio praesentium dolor nisi velit. Dicta voluptates ad molestiae aut. Ea corrupti vero provident facilis fugiat mollitia.
Eaque sed voluptatem atque consequuntur incidunt. Pariatur et necessitatibus sit et distinctio itaque.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...