How to research and verify trading ideas
- Thread starter talontrading
- Start date
- Status
Let me preface this post with a disclaimer... I'm a real ass... Confident to the point of arrogance... confrontation and aggressive to the extreme. So accept my apologies in advance.
And let me tell you spinn... you're sounding like an idiot. Do a little reading and research here and i can pretty much guarantee you'll learn something.
this is what i do for a living... every day for many years and i really do know what i'm doing. don't take my word for it... have an open mind and look at what's going on in this thread. i'm not asking you to blindly accept my authority only to think about this before you dismiss it as a trip to the casino.
And let me tell you spinn... you're sounding like an idiot. Do a little reading and research here and i can pretty much guarantee you'll learn something.
this is what i do for a living... every day for many years and i really do know what i'm doing. don't take my word for it... have an open mind and look at what's going on in this thread. i'm not asking you to blindly accept my authority only to think about this before you dismiss it as a trip to the casino.
not sure where to start here... i guess i should have expected the kind of response this thread was greeted with but it came as a little surprise. let's talk in general about how i look at trading ideas before we did into the specifics of this one. that will give bwolinsky time to absorb and see if he will "accept" the papers i posted as being academic content. (what a freakin' joke lol.)
anyway... the 30,000 foot overview here... very biggest picture. when presented with a new trading idea i usually run through a more or less 3 step process. it's not really linear... like do step 1, check the box and move on to step 2. it's usually more of a fluid thing with some back and forth and interaction and it helps to have other traders around to bounce ideas off of. good things come out of those kinds of interactions (in contrast to the prevailing attitude i've found in these groups which is "i have my very special super secret method and i'm not going to share it but i will tell you all the great trades i could have made with it and will post them as if they're real trades and i did actually trade a 1 lot sometime last year so i'm a real trader i promise.") anyway... here's the outline:
1) Is there a reason this idea should work?
This would include market structure factors, big picture economic reasons, etc. The basic question we're looking at here is does this make some kind of sense? A failure here doesn't mean we stop, but I'm very reluctant to trade something that I can't understand a reason for. (In fact, I never have done so.) I have learned things too when i find something that works, but i see no reason for it... then digging deeper i find it really works... then i come to understand why and learning has taken place. you have to be very critical but also open and flexible. it's a difficult mindset.
2) Statistical testing
As much as I enjoyed the math lesson from riskfreetrading last night, I confess I was annoyed because this is an area we (my trading team) have put a lot of thought and energy into... I educated myself and even got a degree to help in this area and we have collaborated with some truly brilliant people. I think we do this kind of testing as well as anyone in the industry, but I'm not terribly modest lol. The basic idea is to disassemble the pieces of the system and test them separately. Coding a system in metastock or tradestation, running an equity curve, playing with new indicators and settings and optimizing is the farthest thing imaginable from what we do. Really... it's impossible to imagine something less similar. I have a set of technical tools that I know work... some as setups... some as exits... some as added confidence... because they have been through our testing process. We can talk about this in some more detail later.
There are of course a plethora of issues like data mining and walk forward testing (which isn't at all what it's cracked up to be btw) to deal with here.
3) "as if" equity curves with reasonable assumptions
This is another piece of the puzzle and not a very important one. Also lots of issues here (survivorship bias, robustness and trading costs to mention a few) that need to be dealt with intelligently. Once we have generated a test run we also look at it with our position sizing algorithms and run some monte carlo to find a worst case result... which is usually pretty bad btw.
=======================
So... I think it would be most productive to look at the system posted here through these steps, dont you agree? That was always my original intent... so I was a little pissed to be attacked so quickly. let's open the discussion to why this idea might work. what is going on with these additions and deletions that might impact how the stocks trade? ideas???
anyway... the 30,000 foot overview here... very biggest picture. when presented with a new trading idea i usually run through a more or less 3 step process. it's not really linear... like do step 1, check the box and move on to step 2. it's usually more of a fluid thing with some back and forth and interaction and it helps to have other traders around to bounce ideas off of. good things come out of those kinds of interactions (in contrast to the prevailing attitude i've found in these groups which is "i have my very special super secret method and i'm not going to share it but i will tell you all the great trades i could have made with it and will post them as if they're real trades and i did actually trade a 1 lot sometime last year so i'm a real trader i promise.") anyway... here's the outline:
1) Is there a reason this idea should work?
This would include market structure factors, big picture economic reasons, etc. The basic question we're looking at here is does this make some kind of sense? A failure here doesn't mean we stop, but I'm very reluctant to trade something that I can't understand a reason for. (In fact, I never have done so.) I have learned things too when i find something that works, but i see no reason for it... then digging deeper i find it really works... then i come to understand why and learning has taken place. you have to be very critical but also open and flexible. it's a difficult mindset.
2) Statistical testing
As much as I enjoyed the math lesson from riskfreetrading last night, I confess I was annoyed because this is an area we (my trading team) have put a lot of thought and energy into... I educated myself and even got a degree to help in this area and we have collaborated with some truly brilliant people. I think we do this kind of testing as well as anyone in the industry, but I'm not terribly modest lol. The basic idea is to disassemble the pieces of the system and test them separately. Coding a system in metastock or tradestation, running an equity curve, playing with new indicators and settings and optimizing is the farthest thing imaginable from what we do. Really... it's impossible to imagine something less similar. I have a set of technical tools that I know work... some as setups... some as exits... some as added confidence... because they have been through our testing process. We can talk about this in some more detail later.
There are of course a plethora of issues like data mining and walk forward testing (which isn't at all what it's cracked up to be btw) to deal with here.
3) "as if" equity curves with reasonable assumptions
This is another piece of the puzzle and not a very important one. Also lots of issues here (survivorship bias, robustness and trading costs to mention a few) that need to be dealt with intelligently. Once we have generated a test run we also look at it with our position sizing algorithms and run some monte carlo to find a worst case result... which is usually pretty bad btw.
=======================
So... I think it would be most productive to look at the system posted here through these steps, dont you agree? That was always my original intent... so I was a little pissed to be attacked so quickly. let's open the discussion to why this idea might work. what is going on with these additions and deletions that might impact how the stocks trade? ideas???
I am neutral on the OP (don't have a feeling one way or the other), but the proof rests on the proclaimer, not on the forum...
Quite a few things posted on ET prove to have some value, but readers have to sort through a lot of chaff.
But there are going to be a lot of naysayers and a lot of supporters.
There have been many starting "I will teach you" threads, almost none of them were what they originally said. So you would be on rare ground, if true.
you know... no one is possibly more cynical than i am so i do understand what you're saying. if i was selling something or asking for students or something then you have a point. however in this case i would think the appropriate course of action would be for the forum to read with an open mind and evaluate the ideas... again that's why i chose this first system as a lead in because it's easy to evaluate.
The first thing I look at when I have a new idea (my own or somebody else's) is: How well would this fit with my needs as a trader and as a person? I agree with lynx that are are many perfectly valid reasons why a serious trader would not trade the S&P500 add/drop system for a year.
BTW, I filed it in my ideas bucket, but probably won't trade it any time soon. I agree with the OP that this is likely a profitable system, because I have seen the effect up close.
BTW, I filed it in my ideas bucket, but probably won't trade it any time soon. I agree with the OP that this is likely a profitable system, because I have seen the effect up close.
- Status