Filter Exchange · super green alert>spy>166 at close today>weekly 165-166 calls
|
jimmyjazz msg #113675 |
5/24/2013 11:49:56 AM
I really struggle to understand his posts. I'm familiar with max pain theory, and in my observations, it doesn't seem all that reliable. Right now SPY is at $164.80, and $166.50 seems a long way away.
But I hope he's right! Just so long as GLD doesn't drop below $132.50. :)
|
Filter Exchange · super green alert>spy>166 at close today>weekly 165-166 calls
|
jimmyjazz msg #113673 |
5/24/2013 11:28:47 AM
You expect SPY to pop 2 points in the next 4.5 hours?
|
Filter Exchange · 5% A WEEK FILTER (BASED ON TRO'S CROCK POT)
|
jimmyjazz msg #113604 modified |
5/22/2013 6:39:26 PM
Now, see how a 9% stop loss and an 8% limit order outperformed the other settings for EVC over the last 6 months? If you look at the data, you'll see it causes more holds until close on Friday (where I assumed a sell), but most of those are profitable, anyway, and it turned out to be better to let the winners run all the way to an 8% limit as opposed to cutting them short at 5%.
Is this valid theory for a modification of the filter? I don't even know how one would do it, since it requires optimization of every instrument that is flagged.
By the way, EVC tripped my 5% limit and closed for a nice profit today. It would not have tripped an 8% limit, and worse, it fell off the cliff after I sold to close at over 12% below my purchase price, so I dodged a bullet there. A 7% limit would have worked, though, so I "left a little on the table".
Interesting strategy.
|
Filter Exchange · 5% A WEEK FILTER (BASED ON TRO'S CROCK POT)
|
jimmyjazz msg #113603 |
5/22/2013 6:31:16 PM
Yeesh, this is barbaric. Let me try this:

|
Filter Exchange · 5% A WEEK FILTER (BASED ON TRO'S CROCK POT)
|
jimmyjazz msg #113571 modified |
5/21/2013 8:15:56 AM
Sorry to resurrect a dormant thread, but I decided to take a look at one specific stock that was caught by this filter and see if I could optimize the stop loss and limit values to maximize compounded weekly returns over the 6-month scan period. The stock is EVC, which was flagged before the open yesterday as having a high reward rating along with low risk, leading to a very high "performance" rating (38).
I took the open/close/high/low data and looked at the effects of modifying the 10% stop loss and 5% limit settings. It turns out FOR THIS STOCK that over the past 6 months, one could have done better by using a 9% stop loss along with an 8% sell limit. I assumed that a weekly low below the stop loss triggered the loss, which (I believe) is a more conservative approach than Kevin used in his backtesting. Here is a chart of the various weekly equity curves that came out of this analysis. Thoughts? Is this playing with fire?
OK, I need help posting a JPEG file. I tried the [IMG] code but that didn't work.
|
Filter Exchange · NEW SYSTEM WITH INTEGRATED 5% STOP LOSS
|
jimmyjazz msg #113379 |
5/14/2013 11:30:21 AM
Being a complete newbie to SF, I'm not sure I captured the intent of this filter in constructing a backtest:
not ETF
close 1 day ago above 150
average volume(21) 1 day ago above 250000
average volume(21) 1 day ago below 2500000
rsi(2) 1 day ago below 20
adx(10) 1 day ago above 30
low more than 2% below close 2 days ago
close less than 5.99% below close 1 day ago
Note that I have attempted to optimize by reducing the range of volume and price. This was to (a) improve performance and (b) cut down the # of trades.
Other than those price & volume constraints, should this generate the correct set of stocks?
|
Filter Exchange · NEW SYSTEM WITH INTEGRATED 5% STOP LOSS
|
jimmyjazz msg #113359 |
5/14/2013 1:19:39 AM
Next day only? (i.e., place a limit order good for the day at 6% below the prior close?)
I'm not getting the same kind of simulation results on SF so I just want to make sure I'm matching your model. No doubt I have something messed up.
|
Filter Exchange · NEW SYSTEM WITH INTEGRATED 5% STOP LOSS
|
jimmyjazz msg #113356 |
5/13/2013 10:18:38 PM
Kevin, can you clarify -- does the filter on page 1 merely identify a set of stocks from which you are looking for ANOTHER 6% pullback? I'm speaking specifically about the system backtested to create the equity curve.
If that's true, what was the entry signal used -- 6% from the close?
|
Backtesting Support · Question about results
|
jimmyjazz msg #113337 |
5/13/2013 8:17:02 AM
Thanks -- fixing the headers will go a long way towards clearing things up.
|
General Discussion · Trading ^SPX closing prices
|
jimmyjazz msg #113330 |
5/12/2013 2:30:26 PM
Yes, I'm aware of the Yahoo database. The problem is that systems based on SPY in SF do not line up pricewise with the same instrument in stockcharts. It becomes impossible to do a quick check on the SF results w/o having to manually alter every price.
What I want to do is work on a system using index data, as IF I could actually trade the index.
Alternatively, if someone could tell me an ETF proxy to ^SPX that does NOT pay dividends or split, that would work.
|