Are you able to go back historically and obtain statistical data given certain parameters for the S&P such as whether or not it was trading below its 200 day moving average when the 4 consecutive down days occured?
4 down days
- Thread starter fan27
- Start date
Too bad.
End of Story.
End of Story.
Here is an interesting twist to the system. Add "1 unit" after each loss (Go Back to normal size after a win). The profits nearly double and the MAX DD actually goes down.
Long Trades 91
Winning Trades 65
Losing Trades 26
Win Rate 0.714285714285714
Point gain\loss 43457.33
Dollar gain\loss 43277.33
Average Win 925.564461538462
Average Loss -649.475384615384
Largest Win 3714.87999999999
Largest Loss -2864.75
Max Consec Win 15
Max Consec Loss 3
Max DD -2864.75
Expectancy 475.553076923077
Long Trades 91
Winning Trades 65
Losing Trades 26
Win Rate 0.714285714285714
Point gain\loss 43457.33
Dollar gain\loss 43277.33
Average Win 925.564461538462
Average Loss -649.475384615384
Largest Win 3714.87999999999
Largest Loss -2864.75
Max Consec Win 15
Max Consec Loss 3
Max DD -2864.75
Expectancy 475.553076923077
And if you added 10 units after each loss it would be that much better
Think about it, you already know that your small population sample is win biased - so OF COURSE if you up your ante after a loss, eventually it'll work out favorably.
BUT, you only have a universe of 91 such trades garnered over 11 years (most of which was manically bullish) - doesn't seem like a reasonable population size from which to project future success.
A simple system based on the current position of the S&P, since 1.1.02. Similar results in previous years.
STOCKS: 30 stocks of the DOW
ENTRY RULE:
1. Stock is less than 5 of a 10 period daily RawK of stochastic.
2. SPY is less than 5 of a 10 period RawK of daily stochastic
3. SPY is less than 5 of a 10 period RawK of WEEKLY stochastic
4. Enter on open the next day.
EXIT RULE:
1. Stock is above 60 level of 10 period Raw K of stochastic.
2. Stock closes down.
3. Sell at close of that day.
RESULTS:
Average gain of 4.24% per trade
Percent Profitable: 81%
Average win/loss ratio: 1.43
This is not a trade recommendation. Just an observation.
STOCKS: 30 stocks of the DOW
ENTRY RULE:
1. Stock is less than 5 of a 10 period daily RawK of stochastic.
2. SPY is less than 5 of a 10 period RawK of daily stochastic
3. SPY is less than 5 of a 10 period RawK of WEEKLY stochastic
4. Enter on open the next day.
EXIT RULE:
1. Stock is above 60 level of 10 period Raw K of stochastic.
2. Stock closes down.
3. Sell at close of that day.
RESULTS:
Average gain of 4.24% per trade
Percent Profitable: 81%
Average win/loss ratio: 1.43
This is not a trade recommendation. Just an observation.
Quote from ArchAngel:
And if you added 10 units after each loss it would be that much better
Think about it, you already know that your small population sample is win biased - so OF COURSE if you up your ante after a loss, eventually it'll work out favorably.
BUT, you only have a universe of 91 such trades garnered over 11 years (most of which was manically bullish) - doesn't seem like a reasonable population size from which to project future success.
Good Point. I went ahead and tested this on the S&P 500 cash index going back to 1982 ($50,000 = 1 unit. Add 1 unit after each loss. Go back to 1 unit after a win). There were 3 down years:
1986 : -2392
1989: -854
1991: -111
Long Trades 223
Winning Trades 136
Losing Trades 87
Win Rate 0.609865470852018
Point gain\loss 55061.83
Dollar gain\loss 54617.83
Average Win 754.822205882353
Average Loss -552.183793103448
Largest Win 3913.60000000001
Largest Loss -2940.25
Max Consec Win 11
Max Consec Loss 4
Max DD -4316.26
Expectancy 244.91403587444
It will be interesting to see if it can outperform Buy and Hold over the next 20 years.
fan27