Is 5.0 a Profit Factor Good Enough?

[LauraRomans]

What is your opinion?

Trading the emini on a back tested scalping strategy (which includes broker fees):

Is 5.0 a good profit factor?

Is 85% a good success rate?

Is an average of trade of +$12.50 and -$13.00 OK?

Should the greatest cumulative loss be less than 3% of the annual return?

Is it reasonable to expect ALL loss trades to be filled but only 1/2 of profit trades?

What am I missing in this analysis?

Any and all opinions are welcomed!

 

[LauraRomans]

Even negative feedback would be appreciated.

 

[options-george]

Even negative feedback would be appreciated.

Hi Laura.

Just looked at your numbers, your winning trades make a profit of 1 tick, and losing trades make a loss of 1 tick - is that correct?

Win Rates should always be looked at in combination with average reward to risk ratios. So a win rate of 85% with a rr ratio of 1:1 is fantastic. It's not possibly to judge a win rate or rr ratio in isolation.

Without knowing more details, and assuming you are looking at a 1 tick scalping strategy on the emini, then I would say:
- your testing methodology is likely not detailed enough and does not give accurate reflection of live trading conditions
- your transaction costs are likely to be understated
- your expected profitability is likely to be significantly overstated

Hopefully that helps a little (even though it may be discouraging!).

That said, it is really good that you are investing time in testing to begin with!

 

[LauraRomans]

Thank you!

Yes +/-1 or 2 ticks on the emini.

I reflect all transaction fees for the broker and exchange. Orders use a limit so there is only slippage for the 15% losers.

The profit is clearly overstated at 5.0. I do assume 1/2 of the posite trades ocurr and all of the negative trades occur. With these assumptions the PF reduces from 5.0 to 3.0.

Am I missing anything?

 

[options-george]

Thank you!

Yes +/-1 or 2 ticks on the emini.

I reflect all transaction fees for the broker and exchange. Orders use a limit so there is only slippage for the 15% losers.

The profit is clearly overstated at 5.0. I do assume 1/2 of the posite trades ocurr and all of the negative trades occur. With these assumptions the PF reduces from 5.0 to 3.0.

Am I missing anything?

I have limited (but some) emini experience but I would think that if you are going for 1 tick winners then you are likely competing with highly sophisticated robots who look at significantly more data than you are looking at in your testing model, such as execution rates, order sizes, cancellation rates, placement rates, size of queues etc.

I would think it would be difficult for you to incorporate this in your testing model.

What do you think?

 

[LauraRomans]

Great feedback options-George.

Do you know of a broker and platform that offers data on the mentioned items (execution rates, order sizes, cancellation etc.)? Your point hits home. I see the need to do more detailed research.

 

[options-george]

Great feedback options-George.

Do you know of a broker and platform that offers data on the mentioned items (execution rates, order sizes, cancellation etc.)? Your point hits home. I see the need to do more detailed research.

I have an account with Trade Station, a US-based broker. They offer a lot of detailed futures information for as low as $20/month. However I would think that many other options for brokers for the same level of info.

Uranus (a T2w member) mentioned some others in this thread here.
http://www.trade2win.com/boards/futures-options/209876-free-real-time-data-ty-futures-contract.html

However it might be quite a challenge to take this data and use it for testing. I am lucky to be working with a very talented programmer who has done exactly that but I would think that it's a very difficult task for most people (other than for well-funded market participants such as hedge funds and banks).

IMHO, my advice would be to look at slightly longer trade durations and slightly bigger targets/stops to give yourself a chance of at least competing with other traders. And be careful not to underestimate your transaction costs and the potential slippages.

Also talk to SniperDayTrader on the forum here as he may have some helpful pointers for you - he also trades the emini but more on the time horizons that I was suggesting just earlier.

Good luck.

 

[LauraRomans]

Great info, thank you.

Currently, I use the US Tradestation for back testing. So far I'm testing price, volume, time and up/down ticks. Is there data for the other items such as lot size or cancels?

My programming is cude but sufficient.

Your suggestions are appreciated!