|
PureBytes Links
Trading Reference Links
|
Oddball followers,
Those of interested in my previous post should have a look at this.
Before getting into this new variation, I want to point out that you can
actually increase the profits and slightly reduce the trade frequecy of
Mark's original system by simply specifying that the system should not trade
before 11:00 ET, instead of 10:00. Try changing those time parameters in
the original code (change 0930 to 1000) and applying it to a chart and see
what I mean. I guess advancing issues numbers are more stable by that time
of day. Anyhow, on to my next subject.
In my previous post, I showed a variation of oddball that used moving
averages as a directional bias to only take trades in the direction of the
trend. It reduced drawdowns significantly and cut the number of trades in
half, but the payoff was lower profits (although it showed a better average
profit per trade).
In a different variation, I was able to actually increase profits (compared
to Mark's unedited code) and still cut the number of trades in half. The
result is that the total net profit goes up to 367K, (from 339K) but the
number of trades gets reduced from 700+ to 373. This puts the average
profit per trade at close to $1000, more than twice the original code.
However, that gigantic pesky drawdown is still there, and it still has a
relatively large flat period of about 7 months.
Instead of using moving averages as the qualifier of directional bias, I
added an hourly advance decline line (NYSE advancers - NYSE decliners) and
only went long if this was above 0, vice versa for shorts.
I am posting the performance summaries for all three, in this order:
1. Mark's original code ("Oddball Original")
2. My first variation using moving averages ("Oddball Directional Bias1")
3. The new Directional Bias variation that I have described here ("Oddball
directional bias2")
Improvements welcome and encouraged. I would like to hear opinions as to
which of these variations is better. Like I said before, I have trouble
weighing tradeoffs when tweaking systems. If anyone can figure out how to
ditch that drawdown without hurting profits, I'd like to hear from you.
Stops don't work, incidentally. By the way, here's the new code:
{ Data1 = S&P 500.
Data2 = NYSE Advancing Issues
Data3 = NYSE Advance/Decline}
If time > 1000 and time <= 1600 then begin
Inputs: RL(7), BZ(3), SZ(1);
If C data3 > 0 and RateOfChange(Close Data2, RL) > BZ Then Buy;
If C data3 < 0 and RateOfChange(Close Data2, RL) < SZ Then Sell;
End;
TEST PERIOD IS BETWEEN NOVEMBER 30TH 1998 AND JULY 26, 2001
[1] ODDBALL ORIGINAL
TradeStation Strategy Performance Report - Oddball - Original SPC5.CSV-60
min.
Performance Summary: All Trades
Total Net Profit $339,375.00 Open position P/L $625.00
Gross Profit $1,232,125.00 Gross Loss ($892,750.00)
Total # of trades 707 Percent profitable 47.10%
Number winning trades 333 Number losing trades 374
Largest winning trade $25,125.00 Largest losing trade ($9,000.00)
Average winning trade $3,700.08 Average losing trade ($2,387.03)
Ratio avg win/avg loss 1.55 Avg trade (win & loss) $480.02
Max consec. Winners 8 Max consec. losers 10
Avg # bars in winners 10 Avg # bars in losers 6
Max intraday drawdown ($61,325.00)
Profit Factor 1.38 Max # contracts held 1
Account size required $61,325.00 Return on account 553.40%
[2] ODDBALL DIRECTIONAL BIAS1
TradeStation Strategy Performance Report - Oddball - Dir. Bias SPC5.CSV-60
min. (11/30/98-7/26/01)
Performance Summary: All Trades
Total Net Profit $270,675.00 Open position P/L $0.00
Gross Profit $690,300.00 Gross Loss ($419,625.00)
Total # of trades 380 Percent profitable 55.53%
Number winning trades 211 Number losing trades 169
Largest winning trade $25,125.00 Largest losing trade ($11,875.00)
Average winning trade $3,271.56 Average losing trade ($2,482.99)
Ratio avg win/avg loss 1.32 Avg trade (win & loss) $712.30
Max consec. Winners 13 Max consec. losers 6
Avg # bars in winners 8 Avg # bars in losers 6
Max intraday drawdown ($22,475.00)
Profit Factor 1.65 Max # contracts held 1
Account size required $22,475.00 Return on account 1204.34%
[3] ODDBALL DIRECTIONAL BIAS2
TradeStation Strategy Performance Report - Oddball Directional Bias2 -
SPC5.CSV-60 min.
Performance Summary: All Trades
Total Net Profit $367,875.00 Open position P/L $950.00
Gross Profit $933,575.00 Gross Loss ($565,700.00)
Total # of trades 373 Percent profitable 50.67%
Number winning trades 189 Number losing trades 184
Largest winning trade $23,575.00 Largest losing trade ($16,000.00)
Average winning trade $4,939.55 Average losing trade ($3,074.46)
Ratio avg win/avg loss 1.61 Avg trade (win & loss) $986.26
Max consec. Winners 8 Max consec. losers 7
Avg # bars in winners 19 Avg # bars in losers 10
Max intraday drawdown ($66,375.00)
Profit Factor 1.65 Max # contracts held 1
Account size required $66,375.00 Return on account 554.24%
|