nibor100
1,031 posts
msg #145708
- Ignore nibor100 |
12/24/2018 10:04:50 PM
@KSK8,
Here's a tip for you, if you want to own a thread on SF make it read only, otherwise expect to get what you ask for....
Being a mathematician and retired cost analyst I've seen many, many claims of profit making strategies that don't hold up when actual data is provided for all to see and analyze. Which is the primary reason I asked all of my questions which I still believe to be serious in nature.
So I hope you are as patient with me as I use my own patience and understanding to sort thru the assorted text claims and rants you've provided until some actual data appears. ( note the use of the English language in regards to the words patient/patience)
1. Historical backtest data when further back than 3 years is often full of survivorship bias due to stocks being delisted over time and secondly the universe of stocks that the backtest may have been conducted upon, may differ in many ways from the current universe of stocks that SF deals with.
2. Max Drawdown info is key info for many traders to consider when choosing which strategies to follow.
Even my simple 10 days quicklook of your filter results showed 1 of your short picks surging higher after the open by more than 70% before the days close, which usually is not a good situation for shorts to be in.
Other times Max Drawdown data leads to discovery of chains of consecutive losing trades that some traders might not want to deal with. But without any actual data who knows?
3. I'll be among the first to praise your strategy and filter if the data required to answer those questions is ever provided and proves out your claims.
I've come up with a few more less serious questions related to the filter you provided:
a. what is the point of your "offset is 0" line it looks like its superfluous to me.
b. What is the purpose of the 30 filter lines after the mysterious "close is 10% below high" line? Those lines seem to have absolutely nothing to do with finding the stock selections that we are supposed to short, unless there is something else you haven't told us about this strategy?
Perhaps those lines present some additional decision criteria for evaluating whether or not to pull the trigger on the shorts? Bur, surely you wouldn't need them since the strategy is already so perfect.
You could have just as easily claimed 20 years of results and 98% accuracy or 1 month of results and 100% accuracy and it would still read to many of us as all text no fact!
By the way I have no problem being in the same boat as Mactheriverrat.
Actually it is a preferred boat to that of the hype dinghy that you seem to be captaining.
Thanks again for your patience,
Ed S.
|
KSK8
561 posts
msg #145711
- Ignore KSK8
modified |
12/25/2018 12:38:06 AM
nibor,
In statistics when you analyze a stochastic series (this is nothing to do with the stochastic oscillator)..the representative sample for an infinite population is 967 (or above) cases tested. Only then you can be probabilistically sure that a rule has been born. Over the past 14 years for this specific system there has been well over that sample number.
To answer your drawdown concern in a simple approach;
Show me a stock produced by this filter that closed over 70%. Good luck finding one because scenarios like so don't exist. And 14 years of data tells me it won't! RULE OF 967!
Also you can't make the argument that it won't work live trading because it does work !! LOL
Now to address your minor questions;
The point of the "offset is 0" is for those who don't have a subscription would still have the ability to look back into data to see the historical results. Now the purpose of the columns based upon price movement are all for my own leisure because despite being 100% mechanical I feel it is always educational to know what the stock has been doing prior to the criteria transpiring and that is how I gauge price action prior to the current session.
"it is a preferred boat to that of the hype dinghy that you seem to be captaining"
You can insult me all you want but at the end of the day you haven't posted a better system filter. I'd like to see you do so, but I am confident you CAN'T.
Cheers & Merry Christmas!
KSK8
|
shillllihs
6,045 posts
msg #145714
- Ignore shillllihs
modified |
12/26/2018 3:37:14 AM
Not that I was trying to create a better filter than anyone for shorting but after partying for 2 days I felt I needed to create something. You guys don’t need to like it but what do you all think?
I call it,
Fukin Good Short
Maybe I didn’t create the whole thing but altered some stuff, I’m really drunk so not sure but looks real good.
Does this beat your filter?
|
KSK8
561 posts
msg #145715
- Ignore KSK8 |
12/26/2018 11:33:00 AM
If you were truly intoxicated and were able to create such a filter I applaud the endeavor but I don't think this filter adds any generic value to mine unless I am not seeing something in it. It does look appealing though.
|
shillllihs
6,045 posts
msg #145716
- Ignore shillllihs |
12/26/2018 3:09:24 PM
Yeah was drunk.
Not an improvement on your filter but possibly a better overall filter.
Price above 5 and more easily shortable stocks. But whatever, don’t use it.
|
KSK8
561 posts
msg #145717
- Ignore KSK8 |
12/26/2018 3:46:15 PM
Its NOT difficult to short stocks under a $1 using Centerpoint. I don't know what old fashion broker you use shillllihs but go google Centerpoint Securities.
But high-5 for the effort!
Maybe one day you'll craft an improved version.
|
Mactheriverrat
3,156 posts
msg #145721
- Ignore Mactheriverrat
modified |
12/26/2018 9:35:17 PM
deleted - you wanted a shorting filter
|
KSK8
561 posts
msg #145722
- Ignore KSK8 |
12/26/2018 10:30:30 PM
Mac,
I encourage you to improve upon it rather than compete, perhaps tinkering with the criteria may yield more plays or more of a return, the possibilities are endless!
|
Mactheriverrat
3,156 posts
msg #145723
- Ignore Mactheriverrat
modified |
12/26/2018 11:00:22 PM
Roger that! The possibilities are endless!
;-)
|
johnpaulca
12,036 posts
msg #145724
- Ignore johnpaulca |
12/27/2018 12:13:01 AM
Why would anyone try to improve on a filter that yields a 95% win rate? If this correct then it would be the best filter ever posted on SF and kudos KSK8.
|