Please explain HFT market making/arbitrage
Background - I'm a student one year away from graduating a STEM subject trying to explore as many avenues in securities based finance.
Recently met a European firm at a careers fair which refereed to themselves as being a HFT in multiple asset classes involved in market making and statistical arbitrage, they were offering a trading position which only required basic programming knowledge.
I understand that HFT essentially means you program an algorithm to trade for you at much faster speeds than a human is capable of. I understand that this can be used to analyse a market incredibly quickly, allowing you to continuously produce accurate bid and ask prices or discover arbitrage opportunities which can be acted on, all in a fraction of a second. But what would you, as the human trader, be doing in that job?
Do you just program and monitor the algorithms?
How do you develop the original trading idea behind the algorithm, or do you design the algorithm to manifest it's own ideas?
Do you really care much about incoming fundamental data from the market?
Is it anything like a normal trading position working as a market maker?
And connected to these questions, do you think it's a good place for a graduate to start out? Do you think it's a business which will grow and be profitable and eventually be quite lucrative for experienced individuals involved in the industry? And do you develop skills that could mean you could move into other areas of finance, such as HF's?
Would really appreciate some advice on this
Thanks
(And yes I have searched through previous forum topics but information of this subject is quite scarce)
Humans will always be needed to judge and manage the risk their algos take.
As for exit opps, kids on this fucking forum need to stop asking about exit opps from prop trading/market making. The exit opp is retirement or another firm. If you want to go to a quant HF go do a PhD in stats or CS.
Thanks for the response
Ok, get that. But still don't understand how judging and managing risk is going to take up a whole day of work or is massively skilful? Surely the algo would be pretty damn good at managing their own risk in normal conditions, so the trader would only need to step in when the market is acting very oddly because something very unusual is moving the market but this kind of thing happens quite rarely so I don't imagine this can fill a traders day. So in my head a trader in this position can just lie back in his comfy chair and watch the money roll in. But this obviously isn't true. So what else are you doing in this job or can you explain how risk management is so complicated?
I wasn't actually mentioning HF's as an exit opp, I just meant that if the industry were to contract or you didn't enjoy your job much you could move into another sphere of finance where your previous experience would still be useful.
Hedging is the most human part of the trading process imo. The machines will be able to execute your trades for theoretical gains, but usually it's up to the human to determine whether or not it's prudent to fully/partially hedge or even hedge at all.
The difference in how you hedge and manage positions can mean the difference between realizing the full potential of your near-mechanically-perfect trades and throwing away all your edge if the market moves against your deltas.
Alright, that makes sense.
In most firms is it common for the trader to be involved in the design/purpose of the algo or does someone specialising in programming design the algo and then passes it on to a trader to monitor and hedge it's risk? Just interested in knowing how much of a creative input, in terms of coming up with trading ideas and turning them into programs, is involved in the 'trader' role.
It varies significantly by firm, but from my experience its usually the quantitative strategists/researchers that are analyzing the data and designing strategies. Depending on how strong their programming background is they will either do the implementation themselves or will work with a developer to code everything up in c++. This is where the process becomes less exciting in my opinion as it takes a lot of testing and tweaking to get the fills and performance shown in backtesting. In terms of who is actually monitoring these strategies, again it varies, but usually it is the guys who have researched and developed them. Who is going to have a better idea of how well the strategy is running, someone who's done the analysis and knows the inner workings, or some trader who's never seen it?
Its a very difficult question to answer because it varies so much not only by firm, but also by the skills that you bring to the table. The more you know, the more opportunities you will have regardless of where you work.
Qui expedita ipsam optio consectetur aperiam architecto consequuntur. Mollitia quo quasi et incidunt et ad. Qui voluptatem a nisi a expedita sint.
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...