Human Traders are an Endangered Species!
That prescient title is a quotation from a presentation given by Professor Dave Cliff of the University of Bristol at the Innovation and Algorithmic Trading conference hosted by LMAX at University College London at the end of February. I was sat at the front throughout Dave's talk, and I found myself laughing frequently. Perhaps for that very reason any shots of the back of my head have been left on the cutting room floor in the videos that have just been made available on the LMAX Trader community website. Here's Dave's extremely interesting and even more amusing presentation from that event on the subject of "Robot traders and evolving markets":
Here are Dave's take aways (at around 27:00) to try and tempt you to watch the whole thing from start to finish. You won't regret it, I promise!
- Algorithmic trading is here to stay
- Human traders are an endangered species
- Machine optimised algorithms are the future
- The global financial markets are a large scale complex IT system
- A wind tunnel is a useful thing… but it tells you nothing about whether Gower Street will be congested at 5 PM
- We need system level simulations too
Here's the UK Government project that Dave refers to at the end of his talk – Foresight Project on The Future of Computer Trading in Financial Markets and here's why he thinks it's important:
As far as I know no one has a tool for measuring tail risk at the systemic level (32:45)
That is no joke!
Filed under Trading Systems by 
Leave a Comment
Comments on Human Traders are an Endangered Species!
Thanks for posting this video. It is working now.
Really amusing and absolutely brilliant. So robots are better than we are. Now what I have to do is find a genius, brilliant almost fearless and errorless robot!
Thanks for posting this. One significant error: at 30:27 I say "…a few tens of millions…" but I should have said "…a few tens of billions…" the paper I am talking about there, by Michael Kearns, Alex Kulesza, and Yuriy Nevmyvaka (all at the University of Pennsylvania) is published in the Journal of Finance. You can download a preprint of it from here: http://www.cis.upenn.edu/~mkearns/papers/hft_arxiv.pdf
Hi Dave,
Thanks very much for the additional information, and for giving your thought provoking presentation in the first place!
It's now over a year since the original "flash crash" that you discuss in the video. As I recall one of attendees expressed an interest in trying to trade the next one. He's had a few chances, because since then there have been quite a few "mini flash crashes", with high frequency algorithms allegedly the culprits. There was an unusually interesting one only last week on the natural gas futures, which lost around 8% in a few seconds. There's a detailed analysis of what went on over at Nanex.
Hi Jim
Sorry, I hadn't seen your comment until today. Yes, that natgas event in June was very strange: the growing oscillations in the price in the moments leading up to the 8% crash are seriously screwy.
BTW, in September the UK Government Office for Science's "Foresight" project that I mentioned here, and that you linked to in the caption on that video, published its initial findings in a Working Paper along with the 16 associated "driver review" (DR) documents that the project had commissioned as supporting evidence. All of these documents are available for free download as PDF from here:
http://www.bis.gov.uk/foresight/our-work/projects/current-projects/computer-trading/working-paper
The review documents DR3 and DR13 cover in a lot more detail the stuff in the first 20minutes or so of my talk; document DR4 talks in much more detail about the large-scale system level failure issues that I move onto at the end. DR4 does mention the natgas event in a footnote.