nadtim
|
|
| Total Posts: 8 |
| Joined: Jun 2012 |
| |
|
I've developed an Algorithmic Trading System and I’ve started talking to firms which are expressing a desire to take a closer look.
What information would they expect to see before committing funds?
How much information should I give them about how the system works?
Thanks! |
|
|
|
 |
|
Like you know, the daily PnL, transaction costs, asset-class and contracts, scability, required infrastructure .... I would think in most cases you can get pretty specific without really giving the strategy away.
In my experience it's impossible to have a high conviction rate in somebodys else's system. Often it's more a wager on the person running it. Also, I believe, it's much better to run several systems. For example I often see funds that follow one monolithic approach run that one system and the fund into the ground. So selling just a system doesn't sound like a good idea to me, but then I have no idea what we are talking about here. |
|
|
 |
|
| I'd add that if you don't have a reasonably long live trading track record with the system you should be very vocal in addressing every assumption you made to get to your results and parameters. If the accuracy of the backtests/results are in question, people will not likely pursue the idea, regardless of its performance. |
|
|
|
|
FDAXHunter
|
| Founding Member |
| Total Posts: 8114 |
| Joined: Mar 2004 |
| |
|
Never gets old: Whose order flow are you going to be processing?  |
Salman Pushdie |
|
|
nadtim
|
|
| Total Posts: 8 |
| Joined: Jun 2012 |
| |
|
> macrotrader and djfostner thanks for your advice.
> FDAXHunter currently the system would be handling trades for hedge funds.
At this point, in addition to a 12 month track record, folks would like to know
more about the internals of the strategy before scaling up the capital.
Since twice in my life, folks have taken my ideas to make money without me,
I'm somewhat hesitant to repeat the experience again. I don't want to be paranoid
so where should I strike a balance?
I've heard it stated that if two quants discuss a trading strategy and then go off and
implement it separately, after a year one quant's risk adjusted alpha will be higher
than the second quant's risk adjusted alpha. Does this ring true?
|
|
|
|
|
YukaRedux
|
| Now with added evil |
| Total Posts: 574 |
| Joined: Dec 2004 |
| |
|
"Does this ring true?"
You mean, what's the likelihood of two people being able to discuss in *complete* detail, including every corner case, a strategy and then implement independently that *exact* same strategy in *every* regard, right down to their fills...? I'll give you zero, plus/minus zero. |
The things you own end up owning you. |
|
|
