optionplayer333
801 posts
msg #113740
- Ignore optionplayer333 |
5/28/2013 12:19:31 PM
the dip is worthless if the money is not behind it>graphs only tell you to look and see if the money will make your play move>on 5-23 money behind tsla at 84$ went to 105 today> money behind spy at 164.35>4 billion $ s worth
. before csco earnings>+1.2 billion $ s on a 21$ stock>
|
SAFeTRADE
646 posts
msg #113761
- Ignore SAFeTRADE |
5/28/2013 8:46:14 PM
Personal preference. Change to whatever time period you prefer.
|
dknoonan
27 posts
msg #113843
- Ignore dknoonan |
6/1/2013 4:04:21 PM
Mahkoh,
How did you apply extra pullback percentages, in SF or after the backtest in Excel? And what did you use as your entry price? I too think it is worth trying to improve these very promising results.
Also, I am wondering if there is a way to add an additional condition to wait until price turns upward before purchasing.
|
mahkoh
1,065 posts
msg #113845
- Ignore mahkoh
modified |
6/2/2013 4:39:35 AM
I exported the results into excel and evaluated a new column stating ' if drawdown is greater than 1%, display loss/gain column + 1, otherwise leave blank'. The entry in that case is open after day of signal minus 1%. Average result is slightly better at the cost of bigger avg drawdown and fewer trades. I believe the best option is to double down at some point. However that will (hopefully) leave a substantial amount of capital unused and it may go against one's trading rules.
You could add something like 'roc(1,1) > roc(1,1) 1 day ago' to filter for stocks where the rate of decline is decelerating.
|
mahkoh
1,065 posts
msg #113846
- Ignore mahkoh
modified |
6/2/2013 3:08:53 PM
This is a better alternative: Signal has triggered up to 5 days ago and today was the first up day
Edit: not quite happy with the fourth line "count(ROC(1,1) 1 day ago > 0,4) < 1 " which restricts a number of possible candidates.
If the signal happened yesterday it will exclude stocks that had updays prior to the signal. Not sure if this can be coded properly.
|
dknoonan
27 posts
msg #113849
- Ignore dknoonan |
6/2/2013 10:44:23 PM
Mahkoh, thanks for the response. I've also tried various changes, but they haven't been improvements -- including a conditional entry of high 1 day ago.
|
Kevin_in_GA
4,599 posts
msg #113854
- Ignore Kevin_in_GA |
6/3/2013 8:34:11 AM
Just a comment - it is human nature (or at least the nature of folks here on SF) to tweak things, thinking they are improving them. I still do this, but am moving away from this idea and now try to write filters that are as simple as possible, and then leave them that way.
This filter is an example of showing how a basic concept like buying a dip can be quantified against a large set of data (the S&P 500 over 5 years). I am almost certainly cherry picking the best combination of settings by backtesting 7200 variations - probably already overfitting the data. More tweaking is not likely to improve the results.
|
aferguson3
9 posts
msg #113874
- Ignore aferguson3 |
6/3/2013 4:03:14 PM
Sorry I know this is a stupid question, but I am noticing while doing backtesting for my strategies your ROC (7,1) above 2 works really well.
Does ROC(7,1) above 2 mean ROC(7) and ROC(1) are both above 2?
I use Fidelity active trader pro and it will only allow me to set indicator with ROC (7) and ROC (1) and not ROC (7,1).
Thanks
|
mahkoh
1,065 posts
msg #113876
- Ignore mahkoh |
6/3/2013 4:44:34 PM
The second number draws a moving average for the signal line, setting it to 1 removes that. You will want to use ROC(7).
|
aferguson3
9 posts
msg #113877
- Ignore aferguson3 |
6/3/2013 5:05:43 PM
So an exit signal would be when ROC(7) is above 2?
|