Here at Littlefish FX, our primary business is in quant models and building trading systems.
Now, over the past few years quants have had a lot of attention and primarily, in my opinion, bad press. I have no intention of setting the record straight (it’s honestly the only time in my life I could consider myself anything close to a bad boy so I’m taking it…). However, what I thought I would do is give you a bit of insight into what we do in the bowl to unlock a little bit of the mystery around quants and these “evil” algorithms…
I will focus on what we do here at Littlefish FX, and in some cases this will be a little bit different to the norm, but if you are really interested in learning more about quant trading I suggest you read ‘Inside the Black Box‘ (check out our Book Club here), it’s a very easy read that breaks down all the components of a general quant team.
What is a quant?
So let’s start with what the difference is between a quant (quantitative analyst) and a discretionary trader.
My view on this is pretty simple. A quant uses consistent mathematical models in order to set rules for placing trades, while a discretionary trader uses discretion to determine the best trades to place.
In my view a quant will have a hard set of rules that could be coded by a computer to set all trading decisions. Now these could be rather complex – but there are clearly defined rules.
The second part is that, for a quant, the rules will have to be proven and this takes one of two forms typically. Either a theory is formulated and then proved and the theory becomes the rules. Or the rules are defined by data mining historic data. As a matter of fact, I actually think the reality is that these two now overlap a lot in the quant world as models advance.
The role of quants at LFX are straight forward in description and challenging in execution. The whole point of a quant is to generate what we call ‘alpha’, that is a return in excess of a risk-free rate (think of this as the return you get on your savings account in a bank) and the FTSE100/S&P500. On top of that, we must also ensure we are uncorrelated to the equities markets (FTSE100/S&P500) and manage our risk exposure.
Essentially, the quant at LFX needs to design and build trading strategies by coming up with a theory then testing it over 10 to 20 years worth of data for consistency then set rules for inclusion within our portfolios.
They then need to work out a broad range of parameters that could be used with the system; we use a lot of portfolio systems to guard against the dreaded curve fitting, which means each trading strategy has to work regardless of the parameters supplied.
In general though, quants will take a few different approaches. You either need to look at really long timeframes of test data to ensure consistency through the widest range of possible scenarios, or you need to look at really short timeframes and accept curve fitting – ensuring your model can adapt to changing conditions.
The problem most people see when trying this is they find something that works amazingly well for two or three years and then it blows up as the market conditions change. For the most part most of those ‘scam’ robots you see in FX may well have been well intentioned: they were just naive to how they were curve-fitting to specific conditions.
So once our quant has his theory, has proved it and we have a set of parameters, we then run it on cost models and risk models to ensure it is good enough to actually run. If it is, which by this point is a very small subset of what we started with, it then goes into testing.
So there’s your starter for ten. Stay tuned for next time when I will take a look at the different types of models we use, the strategy, the cost and the risk models.
Got a question for Sam? Tweet him directly at @LFXSam or contact us here…
Get your hands on all the trading plans, indicators and strategies required to gain a comprehensive understanding of the Order Flow trading techniques successfully used by the traders here at Littlefish FX. Find out more here.