Once we have good results from backtesting, what is the best way to implement trading using WL? I know of three so far:
Re-run the backtest each bar interval and manually place trades at a broker as soon as there is a signal,
Drag the strategy onto a chart, wait for signals, then manually place trades at a broker,
Use one of the connections to a broker to automate the trading, but I don't know how to do that.
Re-run the backtest each bar interval and manually place trades at a broker as soon as there is a signal,
Drag the strategy onto a chart, wait for signals, then manually place trades at a broker,
Use one of the connections to a broker to automate the trading, but I don't know how to do that.
Rename
Let me throw in a piece of advice:
If you have a promising backtest you should run your strategy on "future" data, i.e. some data you have never used to "optimize" your strategy before you risk real money. (This is not as much fun/excitement as real trading but less expensive)
There are several possibilities to do so:
1. The beginners way: Run your backtest on data until today, Then conduct a few weeks or months of paper-trading to make sure your strategy works on "unseeen data" (the quants name it "Out of Sample/OOS data)
2. Make sure you conduct all backtest on data which stops some months before "today". Then, after you finished all development, tweeking, optimizing adaption, improvement and after you are absolutely certain that your strategy is the best you can come up with... then run it on the data laid aside (the last months).
The second method saves you some weeks of paper trading. But the risk is high that you find out that your strategy does not work good enough on OOS data. So:
3. start again, change parameters, add filters and so forth. And here just a word of warning: If you repeat the third step too often, you are back at possibility one: Do some paper-trading...(Because your "unseen data" is no longer "unseen")
If you have a promising backtest you should run your strategy on "future" data, i.e. some data you have never used to "optimize" your strategy before you risk real money. (This is not as much fun/excitement as real trading but less expensive)
There are several possibilities to do so:
1. The beginners way: Run your backtest on data until today, Then conduct a few weeks or months of paper-trading to make sure your strategy works on "unseeen data" (the quants name it "Out of Sample/OOS data)
2. Make sure you conduct all backtest on data which stops some months before "today". Then, after you finished all development, tweeking, optimizing adaption, improvement and after you are absolutely certain that your strategy is the best you can come up with... then run it on the data laid aside (the last months).
The second method saves you some weeks of paper trading. But the risk is high that you find out that your strategy does not work good enough on OOS data. So:
3. start again, change parameters, add filters and so forth. And here just a word of warning: If you repeat the third step too often, you are back at possibility one: Do some paper-trading...(Because your "unseen data" is no longer "unseen")
There are two great ways to conduct quite realistic paper trading:
* WealthSignals
* Collective2
WL7 currently has a broker adapter for Collective2 which automates most of the work.
* WealthSignals
* Collective2
WL7 currently has a broker adapter for Collective2 which automates most of the work.
Thanks, those are all great comments.
I don't find Collective2 in the extensions list. Where did you find it ?
What about a TD Ameritrade (TOS) paper account ?
I don't find Collective2 in the extensions list. Where did you find it ?
What about a TD Ameritrade (TOS) paper account ?
QUOTE:
Collective2
DrKoch,
What has been your experience with Collective2? I have looked at the website off-and-on for years, and I have attended a couple of their webinars, but I see that you are an active(?) member there, so I would like to hear about your experiences.
Vince
QUOTE:
What has been your experience with Collective2?
Well, a few years back I had a successful strategy running on Collective2 which attracted about 100 subscribers for a few months.
Then the strategy stopped working (because the markets changed...)
This is not C2's fault of course.
I found the system/site quite stable and reliable.
QUOTE:
Then the strategy stopped working
Isn't that the story for all of us! ;)
BTW, have you ever thought of using the metrics of Quality Control to help quickly identify when a system stops working?
Vince
Another great way to paper trade - fully integrated with WL7 is using the Alpaca broker extension. You can start with a free paper trading account, and if you like it, move some money there.
Your Response
Post
Edit Post
Login is required