Strategy Results

[happypappy]

Hey guys, been working on an automated strategy for nearly a year now. I designed 3 of my own indicators and using those and 3 separate time frames I have got a strategy that makes money. *Happy Face*

I have run the strategy over 20 different entities, stocks, currency & futures and they all make money and the figures vary obviously. I've been researching MFE / MAE / ETD and Sharpe to determine if although it making money is wise to use.

I tested 01/01/2013 - 15/11/2013, I need to download the latest data.

The results are
Sharpe ~0.52 - 0.79
MAE ~0.42% - 0.60%
MFE ~0.80% - 2.10%
ETD ~0.84% - 1.56%
Trades over this year range from 40-120 entrys
Max downdraw is ~3.80% - 8.94%
Avg time in market ranges from 2.2 days till 9.4 days
Profit per month ~.90% to 2.6%
Cumulative profit ranged from 10.2% to 35.9%

What do I need to do to improve the strategy, is it the exit timing? Is the profit factor to small? Etc, any advise would be much appreciated.

[StockFundoo]

Hi HappyPappy,

Results look good, although Sharpe ratio needs some improvement. Low shape ratio means that strategy is probably taking excess volatility for a given return. Sharpe of anything over 1.5 is a good number. Anything less than 1 is difficult.
Otherwise the numbers look good.

Cheers!

[happypappy]

Thanks for the reply, would increasing my winning trades effect this? It I worked on it a bit last night and think I can improve this part.

[ABenfer]

I see posts like this all the time, and it's just not realistic to give you quantitative advice without digging into the guts of your strategy.

[happypappy]

Abenfer,

Thx for the response, well I guess I designed 3 indicators to try and determine the trend. The stock works best with 3 different time frames that were provide by optimization run against all the stocks. The fundamentals of how it determines entry and exits is not really relevant.

I'm using 3 time frames 8, 21 & 60 minutes to calculate the best entry points for both short and long. Currently it simply swaps position, so it never exit's the market once it starts it just keeps going and I think this is why the Sharpe is low. I may get 3/4 entry's wrong before hitting the correct trend, however for each wrong entry the loss is for example $70, for each correct entry it makes the gain is say $1000, so it makes money, doesn't matter if the stock / future or currency is bullish or bearish.

After the advise from the first comment I have started working on preventing the multiple wrong entries to determine if that will in fact increase the Sharpe. If it won't and you know how it will, ie the exit is to fast, I should hold for longer or more profit, or the exit is to late I am getting out after to far of a drop I can make adjustment to that as well.

I guess the real questions is, is the MFE / MAE / ETD and Sharpe good? If not how do I improve them?

My understanding is the MAE is great being low, the MFE is not so great, because it should probably be higher, the ETD is ok & the Sharpe is just below good

I've been told Sharpe is based on the accuracy of entries, so the more correct entries the better the results is this right?
I've been told the MFE is low because the profit range is to low, is this right?

[koganam]

Abenfer,

Thx for the response, well I guess I designed 3 indicators to try and determine the trend. The stock works best with 3 different time frames that were provide by optimization run against all the stocks. The fundamentals of how it determines entry and exits is not really relevant.

I'm using 3 time frames 8, 21 & 60 minutes to calculate the best entry points for both short and long. Currently it simply swaps position, so it never exit's the market once it starts it just keeps going and I think this is why the Sharpe is low. I may get 3/4 entry's wrong before hitting the correct trend, however for each wrong entry the loss is for example $70, for each correct entry it makes the gain is say $1000, so it makes money, doesn't matter if the stock / future or currency is bullish or bearish.

After the advise from the first comment I have started working on preventing the multiple wrong entries to determine if that will in fact increase the Sharpe. If it won't and you know how it will, ie the exit is to fast, I should hold for longer or more profit, or the exit is to late I am getting out after to far of a drop I can make adjustment to that as well.

I guess the real questions is, is the MFE / MAE / ETD and Sharpe good? If not how do I improve them?

My understanding is the MAE is great being low, the MFE is not so great, because it should probably be higher, the ETD is ok & the Sharpe is just below good

I've been told Sharpe is based on the accuracy of entries, so the more correct entries the better the results is this right?
I've been told the MFE is low because the profit range is to low, is this right?

Your answer gives the impression that what you really are seeking is general ideas/comments about automated trading system. If so, you might be better off exercising your Google Fu. (e.g., https://www.google.com/search?q=buil...ient=firefox-a).

Just as ABenfer says, it is unfortunately impossible to give specific comments without seeing the specifics of the strategy. OTOH, it is obvious why you would not want to give away your specific secret sauce.

However, ultimately, all trading is "managing money", "exits" and "entries", in that specific order. Managing money and position sizing relates to expectancy, which would depend on the rules (which you cannot disclose).

Ultimately, you need to determine whether your results, which seem to be made up of many small trades and a few large successful trades, are viable results. IOW, are your large successful trades outliers, or are they a normal part of the system. What are the confidence limits on your results?

You may want to read these threads, where I expounded on some of the mathematics of the issue:

http://www.ninjatrader.com/support/forum/showthread.php?t=43841&highlight=expectancy

http://www.ninjatrader.com/support/forum/showthread.php?t=58578&highlight=expectancy

[happypappy]

Koganam

Thx u for responding, I will take a look I did a lot of google before I came here. Maybe your links will explain what I am after. My secret sauces will be come available to all in due time.