If you have a portfolio of 1000 stocks, there are 5 key decisions that one needs to take:

Buy to Open
Sell to Open
Buy to Close
Sell to Close
Do Nothing

In a quant based strategy like the one I have been working for the past few months , the signals help me get on to the position. However, there is a catch. As you get on to the position based on external asset price movements, you have to look at something relating to the portfolio, which is RISK. Risk based quant investing is something that is of key importance to stat arb based strategies.

Now as things stand, All I have is signals and based on those signals, here is how P&L looks like this :(PATHETIC to say the least)

pl

Look at how the portfolio rebounds when there is a huge vol. Anyway that’s not the point of the above P&L.The thing to be looked at is this:
This portfolio is not market neutral. So, how do you become market neutral. Well, it is part art and part math.

Market neutrality can also be achieved by Risk based decision making. That’s what I am working on. I hope to see this P&L becomes more well behaved when risk based decision making is put in place. Right now, the strategy is like a person on steroids!!..Ever reacting to the relative value of the stocks..That is not enough.
So, what should I be doing..The easiest thing to do is :

Get your trading signals for the day and then decide whether you want to trade the new positions or not based on ONE QUESTION. Do the trades increase my risk beyond a threshold. Now this is too too too easy and naive to begin with. Ideally there must be stock selection in the subset of potential stocks to be traded based on risk each of them brings in to the portfolio . If one thinks about this problem, it is not an easy question to answer as to what is the best set amongst the subset of stocks which are culled out by trading signals!!!! It would involve pairs / group based selection , ordering of the selections etc…

It is going to be tough to get this working..How to kill this low sharp ratio in the portfolio!!! ?

Hopefully trial and error on various risk based strategies should throw some light in the days to come!!..Hopefully !