Monday, 15 August 2011

Weekend trading work

The week's trading has led to further key questions to address on Saturday and Sunday

The first question concerns exits from trades - the question of profit targets versus reversal signals. On the one hand, we seem to have missed some huge wins recently by having a fixed profit target, but at other times wins are achieved that would be missed if we just waited for a signal reversal. What is best overall? My guess is that it is better to take the profit target exits - although there seem to be loads of huge profits around, actually there are also lots of profit-target gains as well - we just don't remember them as well as the "missed" huge wins.

Secondly, there is the question of the impact of waiting till end of time period before adjusting levels, versus making intra bar adjustments. The problem here is that the latter is not testable as there are possible intra-bar whipsaws that I can't detect. So I need to ensure that the current testing is based on what we could definitely get i.e. by only changing levels at the end of the bar. Then if this result is ok, we will know that we have a result that could actually be traded.

Next, it is clear thast the protocols in place for what happens after a profit target exit has occured, need tightening. And we could do with sorting out a better opening trade methodology.

But I am also thinking about some of the parameters of the current systems and trying to bet a feel for what is the impact of changing them? e.g. if one set is changed, avoid some poor trades, but get worse fills on all trades - would it be better to get more trades but better fills? On Sunday I spent the entire day doing a test of our CL parameters for an increase in levels - more trades, but better fills. I processed 15 days, and result is clear. It is better to do more trades but at better fills. So that change will be made for certain

Late on Sunday Jerome called and we were able to run through the latest system spec and review Aug 11th in detail for new rules. He is then going to work on these during Monday and Tuesday.

Some articles in the papers this weekend about algorithmic trading and HFT, seeking to blame these for the current market vol. No doubt regular investors are not to blame in the slightest. I have actually been keeping loads of articles at the moment on the current market difficulties - will no doubt make fascinating future reading!

So this week will be about the new system spec and more research and I am hoping to process at least a few weeks from the last 3 markets. Then I am back thinking about possibly increasing my investment in the trading - maybe double current position sizes and split it 66-33

And then there is issue of other external investors. My next major review is planned for around half term in October. Maybe after that I could think about approaching one or two possible investors.

So much to think about!!

No comments: