Let's Compare Strategies

[koganam]

still using same basic concept. 2 days were in NT, and the other 2 were on my phone in my other account.

May 2013..actual results

Very nice. 11 trades, no losses, and a trade about every 2 days. Now, I wish I could get my grubby, greedy, little fingers on something like that!

[sledge]

Along those very lines, is the secret-sauce whose test results that you published, making you some good returns?

still using same basic concept. 2 days were in NT, and the other 2 were on my phone in my other account.

May 2013..actual results

[koganam]

I'm finding scaling up helps out a lot.

It appears I'm good at "guessing" 1 to 1.5 pts, from there out, it scales down.

So when I take a bunch of contracts off at 1 pt, I'm already in the + nicely, and the rest can play it out, and I'm stressing less over "I could have had that".

Along those very lines, is the secret-sauce whose test results that you published, making you some good returns?

[koganam]

Live yes
Mega-bucks no (could be if I scaled up)
Healthy Profit yes

I have continued to rewrite this program (relies on ATR) and another one I am using (relies on OBV). Each has only 6 variables (including trailstop and profittarget).

From my experience and from reading posts here and literature you have to minimize the number of variables to increase your chances of replicating backtests with live trading.

I have also realized that my discretionary trading can't compete with my automated trading, which is why I guess I have been putting in so much time programming.

I sent you a PM. Never got a response. Did the TCPIP gods lose it in the ether?

[sledge]

I'm finding scaling up helps out a lot.

It appears I'm good at "guessing" 1 to 1.5 pts, from there out, it scales down.

So when I take a bunch of contracts off at 1 pt, I'm already in the + nicely, and the rest can play it out, and I'm stressing less over "I could have had that".

Live yes
Mega-bucks no (could be if I scaled up)
Healthy Profit yes

I have continued to rewrite this program (relies on ATR) and another one I am using (relies on OBV). Each has only 6 variables (including trailstop and profittarget).

From my experience and from reading posts here and literature you have to minimize the number of variables to increase your chances of replicating backtests with live trading.

I have also realized that my discretionary trading can't compete with my automated trading, which is why I guess I have been putting in so much time programming.

[NFT-Trader]

Live yes
Mega-bucks no (could be if I scaled up)
Healthy Profit yes

I have continued to rewrite this program (relies on ATR) and another one I am using (relies on OBV). Each has only 6 variables (including trailstop and profittarget).

From my experience and from reading posts here and literature you have to minimize the number of variables to increase your chances of replicating backtests with live trading.

I have also realized that my discretionary trading can't compete with my automated trading, which is why I guess I have been putting in so much time programming.

[sledge]

So have you gone live?

Making mega-bucks?

Is it working?

I have spent the last 4 days changing my >50 parameter backtested strategy (see my early posted results) to the minimum number I feel still retains the nature of this strategy, which is now 19 parameters. Using this strategy and optimizing the best time frames for walking forward I think I have what I believe to be a good system.

The system uses 10 days to predict the next 10 days (including weekends). Here are my results after averaging two WFs for every 10 day period
Total Profit: $15,479 for 290 days
Average Profit Factor = 2.71
Average % Profitable = 52.69%
Cumulative Profit over 290 days = 65.43%

The graphs below represent cumulative profit (left) and investing the system using 1/4 Kelly Criterion (right)

Thank you to those in this discussion who encouraged me to look into walking forward analysis.

[NFT-Trader]

Walking Forward

Walk forward results are the only results that count, otherwise a simple optimization is just predicting historical data perfectly (especially when you >3 parameters, you mentioned you have 50!). Either way, looks like it's off to a good start, but 1 month of live results may not be enough, unless it's profiting every day.

I have spent the last 4 days changing my >50 parameter backtested strategy (see my early posted results) to the minimum number I feel still retains the nature of this strategy, which is now 19 parameters. Using this strategy and optimizing the best time frames for walking forward I think I have what I believe to be a good system.

The system uses 10 days to predict the next 10 days (including weekends). Here are my results after averaging two WFs for every 10 day period
Total Profit: $15,479 for 290 days
Average Profit Factor = 2.71
Average % Profitable = 52.69%
Cumulative Profit over 290 days = 65.43%

The graphs below represent cumulative profit (left) and investing the system using 1/4 Kelly Criterion (right)

Thank you to those in this discussion who encouraged me to look into walking forward analysis.

[marcow]

Along those lines, what is the lowest commission rate you've seen for futures, and who has it?

InteractiveBrokers, $4,- round trip for US futures.

[koganam]

Do you mind saying how much it was?
Seats are a little expensive.

Last they were quoting is $3.85RT for every retail trader. They were not showing any limit breaks, but I am sure they do exist if your volume is high enough. I do not know; I just think that there probably are.

[Adamdetx]

Do you mind saying how much it was?
Seats are a little expensive.

[koganam]

Along those lines, what is the lowest commission rate you've seen for futures, and who has it?

For retail traders who do not have a seat, the lowest that I have been quoted so far is from Trading Technologies.

[Adamdetx]

Along those lines, what is the lowest commission rate you've seen for futures, and who has it?

[koganam]

Hey guys, I hope it's not to late to post my latest. Normally I go for longer term bots, but here is my latest scalper.

Profit Factor 2.76
Net Profit $5210
Percent Profitable 91.33%

It only has a few days of data in the screen shot since it a really data intensive strategy.
I have been able to take some random sampling of days in Dec 2011 and Jan 2012 and the results are congruent with what I'm seeing in Sim now. By the way the screen shot is from running sim in real time. If I can just scrape some money together now I can run this thing.

That is an awful lot of vig you are giving to the broker!

[marcow]

No difference in results with slippage 0,1 or 2 ? That's very strange. It almost seems that slippage does have no effect on a sim account ? There has to be a difference between them. I mean if your stoploss gets triggered it sends a market order. with slippage 2 this order gets filled 2 ticks lower than with slippage 0

Please do try a run of the same day on MarketReplay with slippage 0,1 and 2.

I'll post more results as I have them.

yes, please !