Backtesting: wierd performance results

[gammascalper]

I am backtesting a futures strategy in which I am never long (i.e.i sell then cover, sell then cover etc).

When I look at the "Strategy Analysis" report all the "Long Trades" performance metrics are 0 or N/A as you'd expect but there are significant differences between "Short Trades" and "All Trades".

Since all trades are in fact short trades I would expect them to show the same values.

examples shown as (All Trades / Short Trades)

Adjusted Gross Profit (40,451 / 51,906)
Adjusted Gross Loss (13,484 / 13,484)
Select Gross Profit (30,617 / 30,617)
Select Gross Loss ( (7,019) / (41,247) )

Sharpe Ratio ( 1.12 / 5.11)

The "All Trades" values seem more realistic from experience, but given the discrepancies I am not sure I even trust those calculations now.

Anyone have any ideas what is going on ??

[zadombei]

I use quite a few short strategies in Stock Watch Pro http://www.crontech.com/swp backtester and all of the results look valid and make sense. It is weird that you get discrepancies since we are using the same feed...