A small matter of programming

Gratulations ;)

Can you give us some more info about your system? What time horizon do you look at? How long you expect trades to look at?
 
60-90 minutes typically, although there's exit-fast code if it's makng a lot of profit (twice expected). Lookback period is 120 minutes. It puts a 15 minute break between trades so it won't re-enter immediately, gives the whole thing time to cool off.

Going to fix today's bug findings then leave it running overnight. You can follow the ensuing chaos at http://twitter.com/FoxQuant
 
Nice. Well done. Lets hope you get profitable ;)

I hoppe to get my strategy running mid december ;) Still so much to do ;)
 
Not dead, just didn't want to post every "Today, I made a small tweak. It did nothing" update...

Today was the third live test of this version. First test made about 30 GBP, second lost about 80 USD, but they were both small-scale proof of concept tests...

7 roundturns, 6 profitable, trading 50-100k each time, across EUR/USD, GBP/USD, EUR/GBP and USD/CAD. Final results:

  • 34 GBP profit
  • 190 USD profit
  • 45 CAD loss

That gives me a total profit for today of £120, approximate, After profit/losses to date, that's about £100 profit still.

Good start I think. The trader needs a lot of work still, but good to have a solid success under my belt...
 
Sorry, the day job had me tied up for a while, and I didn't run the trader for a while. My initial profits ended up at £50 up after everything, then I took a break until end-Feb. I'm now £220 up, down from a peak of £370 up (this morning).

Hammered a bucket load of bugs out today, which is good. Always seems like my profitable trades are during the paper-trading tests, though (I should be about £200 up today, if I'd been trading live all day instead of half on paper).
 
Holy thread necromancy, Batman!

I just wanted to wrap this thread up; I have eventually given up. While my attempts at automated trading are seemingly improving, this mostly means I'm losing money less quickly. However, given that I have no invested nearly 3 years work into this, with no sign of a pay-off at any point in the foreseeable future, I have to give up. My free time is also under substantially increased pressure (I have gone from being half of a team developing research software used by around a quarter of the institution I work for, to leading that team and reporting directly to the CIO, with plans to expand to other institutions).

Interactive Brokers and IQFeed have been fantastic, and I would highly recommend both to anyone following this path. I would also not recommend anyone follows this path. That should help confuse a few people.

Long term, I suspect I may end up trying to do a similar job for a financial institution, and I hope I can at least salvage some of the investment through experience gained.

Good luck to you all!

Edit: Oh, just realised I hadn't posted at all in the interim. Essentially, I've now lost all my profits to date, and have made a minor loss before costs (and a more serious loss after).
 
Top