What is the logic in this? Most stuff does not work. Automating trash just gets you automated trash.
Or were you referring to a particular system?
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
How to research and verify trading ideas
- Thread starter talontrading
- Start date
- Status
- Not open for further replies.
What is the logic in this? Most stuff does not work. Automating trash just gets you automated trash.
Or were you referring to a particular system?
Quote from talontrading:
It might interest you to know that the additional filters I was going to apply to this were volatility based also. You can construct a reasonable proxy by taking one of the measures I listed above and calculating it over a 1-2 week timeframe, and then calculating it again over a 6-12 month timeframe. The ratio of those two measures will tell you if the market is trading at a premium to its long term vola or a discount. While not a standalone trading system can this be a useful addition to our system? let's see.
Going back to the spreadsheet I posted, the average return was 45.14 bp (basis points) with a stdev of 463.6. I collected (but did not post in the original sheet) a volatility measure for each trade at the time of inception. Does this filter have predictive value for the returns? A quick way to answer this is to do a simple linear regression:
Source | SS df MS Number of obs = 1075
-------------+------------------------------ F( 1, 1073) = 12.09
Model | 2571733.96 1 2571733.96 Prob > F = 0.0005
Residual | 228227966 1073 212700.807 R-squared = 0.0111
-------------+------------------------------ Adj R-squared = 0.0102
Total | 230799700 1074 214897.3 Root MSE = 461.19
------------------------------------------------------------------------------
return | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
volaratio | 125.7746 36.17133 3.48 0.001 54.80004 196.7492
_cons | -67.19465 35.2366 -1.91 0.057 -136.3351 1.94581
------------------------------------------------------------------------------
(Quick and dirty here, but an investigation like this is part of our process here so Im showing it.) The key elements here are the sign of the coefficient (+ which means higher vola = higher returns) and the t stat of 3.48, which is probably significant. (Don't get too caught up on R-squared, etc... this is not a complete model just a first foray into looking at the relationship.)
What happens if we only take trades when the volatility is greater than the long term for that market? Adding this filter drops the number of trades to 357 from 1075, increases the average trade to 102.27 bp(!) with a standard dev of 525.3. The median trade also raises to 143 from 106. Biggest and smallest trades are no affected.
Question: does this look like a worthwhile addition to the system? Thoughts??
And yes, I know BoWo will steal... but we're using some big words so I think we're safe (like spelling words in front of the kids so they don't understand lol.)
So far I'm a little suprised that no one has picked up on this analysis and ran with it. Understanding volatility and how it can make or break a trading idea is a conceptual must have in one's bag of tricks.
Its possibly that the analysis talon provided didn't have the intended impact. This could be because it conceptually went over the heads of some here, and its also possible people are still chewing on this fact. Its also possible that some just discarded it and moved on. But, in the hopes of getting some ideas out there, I have created a simple way to conceptualize what talon is doing.
Attached you'll find a spread-sheet that has the OHLC for the SP500 raw index (its called $INX in my symbol library).
What I've done is calculate something I've decided to call a volatility spread. Its been done before so it probably has a standard name associated with, but, for the purposes here lets just call it the following:
Vola_Spread = ATR ( 5 days ) / ATR ( 40 days);
The ATR lags were choosen to reflect 1 week and 2 months, respectively. These numbers are the result of an educated guess.
So we get a ratio that oscillates around unity vs. time.
The basic rules for analyzing the index are the following:
if Vola_Spread > 1.1 then we're at a discount (i.e. buy).
if Vola_Spread < 0.7 then we're at a premium (i.e. sell).
1.1 and 0.7 were the result of some "fitting" to the data, but, they reflect the nature of volatility - higher short term vola. usually indicates a discounted market and the reverse for lower short term vola.
The attached spreadsheet shows this effect clearly if one were to buy the index at the discount and short the index at premuim. The red line is closed index "points" and the blue line is the index itself.
Hopefully, this gets the discussion started as I believe its worthwhile to develop this concept.
Mike
Attachments
This did go over my head but I'm not sure if I speak for others. Trying to understand it as we speak
Quote from Mike805:
So far I'm a little suprised that no one has picked up on this analysis and ran with it. Understanding volatility and how it can make or break a trading idea is a conceptual must have in one's bag of tricks.
Its possibly that the analysis talon provided didn't have the intended impact. This could be because it conceptually went over the heads of some here, and its also possible people are still chewing on this fact. Its also possible that some just discarded it and moved on. But, in the hopes of getting some ideas out there, I have created a simple way to conceptualize what talon is doing.
Attached you'll find a spread-sheet that has the OHLC for the SP500 raw index (its called $INX in my symbol library).
What I've done is calculate something I've decided to call a volatility spread. Its been done before so it probably has a standard name associated with, but, for the purposes here lets just call it the following:
Vola_Spread = ATR ( 5 days ) / ATR ( 40 days);
The ATR lags were choosen to reflect 1 week and 2 months, respectively. These numbers are the result of an educated guess.
So we get a ratio that oscillates around unity vs. time.
The basic rules for analyzing the index are the following:
if Vola_Spread > 1.1 then we're at a discount (i.e. buy).
if Vola_Spread < 0.7 then we're at a premium (i.e. sell).
1.1 and 0.7 were the result of some "fitting" to the data, but, they reflect the nature of volatility - higher short term vola. usually indicates a discounted market and the reverse for lower short term vola.
The attached spreadsheet shows this effect clearly if one were to buy the index at the discount and short the index at premuim. The red line is closed index "points" and the blue line is the index itself.
Hopefully, this gets the discussion started as I believe its worthwhile to develop this concept.
Mike
I've begun looking at ways to filter out trades myself. Less trades and commisions and greater stability in returns sounds great. My concern with creating things like this volility spread indicator are how do I pick the proper time frames? I guess there's some fudge room on daily data but how do you come up with proper time frames on intraday data? Is this just a matter of trial and error, experience? Something like ATR(15)/ ATR(60) on a one minute chart seems feasible but I get nervious that I'm curve fitting when I try to get more specific than that on the period numbers. How do you know if the nature of volitility has changed and your filter is no longer accurate (before it's too late)?
A time filter has seemed beneficial to me on intraday trades. Only trading during certain hours of the day when volitility will be higher. How about only taking signals within a certain range of key prices (yesterday's close, overnight high/low, opening range high/low etc.)?
Even with a filter such as these, if you're using it with a mean reverting system how do you protect yourself from fading volatility that turns into a trend and fails to revert to the mean? Big spike in volatility, you get your signal, and then volatility dies but price is slowly drifting against you. Some sort of a time stop? Thanks again guys.
A time filter has seemed beneficial to me on intraday trades. Only trading during certain hours of the day when volitility will be higher. How about only taking signals within a certain range of key prices (yesterday's close, overnight high/low, opening range high/low etc.)?
Even with a filter such as these, if you're using it with a mean reverting system how do you protect yourself from fading volatility that turns into a trend and fails to revert to the mean? Big spike in volatility, you get your signal, and then volatility dies but price is slowly drifting against you. Some sort of a time stop? Thanks again guys.
Quote from Mike805:
So far I'm a little suprised that no one has picked up on this analysis and ran with it. Understanding volatility and how it can make or break a trading idea is a conceptual must have in one's bag of tricks.
Its possibly that the analysis talon provided didn't have the intended impact. This could be because it conceptually went over the heads of some here, and its also possible people are still chewing on this fact. Its also possible that some just discarded it and moved on. But, in the hopes of getting some ideas out there, I have created a simple way to conceptualize what talon is doing.
Attached you'll find a spread-sheet that has the OHLC for the SP500 raw index (its called $INX in my symbol library).
What I've done is calculate something I've decided to call a volatility spread. Its been done before so it probably has a standard name associated with, but, for the purposes here lets just call it the following:
Vola_Spread = ATR ( 5 days ) / ATR ( 40 days);
The ATR lags were choosen to reflect 1 week and 2 months, respectively. These numbers are the result of an educated guess.
So we get a ratio that oscillates around unity vs. time.
The basic rules for analyzing the index are the following:
if Vola_Spread > 1.1 then we're at a discount (i.e. buy).
if Vola_Spread < 0.7 then we're at a premium (i.e. sell).
1.1 and 0.7 were the result of some "fitting" to the data, but, they reflect the nature of volatility - higher short term vola. usually indicates a discounted market and the reverse for lower short term vola.
The attached spreadsheet shows this effect clearly if one were to buy the index at the discount and short the index at premuim. The red line is closed index "points" and the blue line is the index itself.
Hopefully, this gets the discussion started as I believe its worthwhile to develop this concept.
Mike
Quote from wutang:
I've begun looking at ways to filter out trades myself. Less trades and commisions and greater stability in returns sounds great. My concern with creating things like this volility spread indicator are how do I pick the proper time frames? I guess there's some fudge room on daily data but how do you come up with proper time frames on intraday data? Is this just a matter of trial and error, experience? Something like ATR(15)/ ATR(60) on a one minute chart seems feasible but I get nervious that I'm curve fitting when I try to get more specific than that on the period numbers. How do you know if the nature of volitility has changed and your filter is no longer accurate (before it's too late)?
A time filter has seemed beneficial to me on intraday trades. Only trading during certain hours of the day when volitility will be higher. How about only taking signals within a certain range of key prices (yesterday's close, overnight high/low, opening range high/low etc.)?
Even with a filter such as these, if you're using it with a mean reverting system how do you protect yourself from fading volatility that turns into a trend and fails to revert to the mean? Big spike in volatility, you get your signal, and then volatility dies but price is slowly drifting against you. Some sort of a time stop? Thanks again guys.
You may want to go about this scientifically and let the market tell you. Generally it is not a good idea to set arbitray values.
One of the advantages of using real market information is that it is very reliable.
The starting point for doing science is not doing any inductive processes or reasoning. You may notice you are where you are (see your questions) through finding stonewalls as a consequence of induction.
The market operating matrix of values when market pace is crossed with market volatility, yields Gaussian distributions for both the rows and columns of the matrix. Nothing is arbitrary.
It does not mean you use this matrix to create anything. That has to be avoided by turning to science instead.
As the OP says he spends a great deal of time at the end of each year discovering the mistakes he has made. this was played out a few years back by a person whose handle was Acruary. What caused him to discontinue what you and others do here is a very intresting story.
I'd think it is NEVER a good idea to use arbitrary values which is why I'm asking how you come up with values that are not curve fit. I can come up with numbers that have worked that are general enough that they don't seem to be "curve fitting" but how do you know when they have become arbitrary? Back on ignore now.Quote from jack hershey:
You may want to go about this scientifically and let the market tell you. Generally it is not a good idea to set arbitray values.
Welcome Jack. Everyone please hold the discussion right here if you would please. There is something we need to do before going further. It will be a few hours before I can post more but please wait. Sorry and thanks.
Hi Jack,
Ok I'm not really an ET regular. Before I started this thread a few weeks ago I checked in once every few months and just browsed some threads. I had come across some of your posts and the general reaction to those posts, so I think I understand some of the issues we may find ourselves dealing with here.
As a sometimes professional writer, I believe that language is very important. Good language communicates our ideas with precision. Good language bridges the gap between writer and reader. Poor use of language obfuscates and confuses. Even the most complicated concepts, when laid out by a master, read clear, simple and true. In fact, that's often a good stand-alone standard -- if someone can make complicated concepts seem simple, they probably really "get it".
Please don't take this the wrong way, but the problem with your posts is that they often seem to not make sense because the language you use is idiosyncratic. I do not want to read someone talking about applying the scientific method to market data (a good idea btw), but writing about it in parable, hidden meanings, and generally using the mysterious language of oracles! There may be a time to burn some psychoactive plants and consult the Delphic priestess, but in market discussions we must strive for simplicity and clarity. Always.
Please take this chance to define your concepts precisely, using examples if needed, and say what you mean in the simplest terms possible. Some specific points from your post:
"market operating matrix of values". this, to me, reads as nonsense. I have a decent background in mathematics, econometrics and quite a long history as a trader, and I frankly cannot make heads or tails of this. I am assuming this is a meaningful concept to you so please explain in clear English what this matrix is and perhaps show us how to construct it.
"yields Gaussian distributions" - because the market operating matrix of values is not a concept I understand, we can't address this. Let me get out in front of this though by saying that, in market data, pretty much nothing is classically Gaussian. I am hoping your example of the market operating matrix of values will let us clearly see these Gaussian distributions.
Let's start there, please. Look at this as a request for clarity from someone willing to give your ideas a chance. Once we have addressed this core concept then we can dig deeper into your suggestion of applying the scientific method to market data.
I also abhor innuendo in a context like this. What precisely are you saying here? What is the story of Acruary?
Ok I'm not really an ET regular. Before I started this thread a few weeks ago I checked in once every few months and just browsed some threads. I had come across some of your posts and the general reaction to those posts, so I think I understand some of the issues we may find ourselves dealing with here.
As a sometimes professional writer, I believe that language is very important. Good language communicates our ideas with precision. Good language bridges the gap between writer and reader. Poor use of language obfuscates and confuses. Even the most complicated concepts, when laid out by a master, read clear, simple and true. In fact, that's often a good stand-alone standard -- if someone can make complicated concepts seem simple, they probably really "get it".
Please don't take this the wrong way, but the problem with your posts is that they often seem to not make sense because the language you use is idiosyncratic. I do not want to read someone talking about applying the scientific method to market data (a good idea btw), but writing about it in parable, hidden meanings, and generally using the mysterious language of oracles! There may be a time to burn some psychoactive plants and consult the Delphic priestess, but in market discussions we must strive for simplicity and clarity. Always.
Please take this chance to define your concepts precisely, using examples if needed, and say what you mean in the simplest terms possible. Some specific points from your post:
"market operating matrix of values". this, to me, reads as nonsense. I have a decent background in mathematics, econometrics and quite a long history as a trader, and I frankly cannot make heads or tails of this. I am assuming this is a meaningful concept to you so please explain in clear English what this matrix is and perhaps show us how to construct it.
"yields Gaussian distributions" - because the market operating matrix of values is not a concept I understand, we can't address this. Let me get out in front of this though by saying that, in market data, pretty much nothing is classically Gaussian. I am hoping your example of the market operating matrix of values will let us clearly see these Gaussian distributions.
Let's start there, please. Look at this as a request for clarity from someone willing to give your ideas a chance. Once we have addressed this core concept then we can dig deeper into your suggestion of applying the scientific method to market data.
Quote from jack hershey:
The market operating matrix of values when market pace is crossed with market volatility, yields Gaussian distributions for both the rows and columns of the matrix. Nothing is arbitrary.
I also abhor innuendo in a context like this. What precisely are you saying here? What is the story of Acruary?
Quote from jack hershey:
As the OP says he spends a great deal of time at the end of each year discovering the mistakes he has made. this was played out a few years back by a person whose handle was Acruary. What caused him to discontinue what you and others do here is a very intresting story. [/B]
Quote from talontrading:
Hi Jack,
Ok I'm not really an ET regular. Before I started this thread a few weeks ago I checked in once every few months and just browsed some threads. I had come across some of your posts and the general reaction to those posts, so I think I understand some of the issues we may find ourselves dealing with here.
As a sometimes professional writer, I believe that language is very important. Good language communicates our ideas with precision. Good language bridges the gap between writer and reader. Poor use of language obfuscates and confuses. Even the most complicated concepts, when laid out by a master, read clear, simple and true. In fact, that's often a good stand-alone standard -- if someone can make complicated concepts seem simple, they probably really "get it".
Please don't take this the wrong way, but the problem with your posts is that they often seem to not make sense because the language you use is idiosyncratic. I do not want to read someone talking about applying the scientific method to market data (a good idea btw), but writing about it in parable, hidden meanings, and generally using the mysterious language of oracles! There may be a time to burn some psychoactive plants and consult the Delphic priestess, but in market discussions we must strive for simplicity and clarity. Always.
Please take this chance to define your concepts precisely, using examples if needed, and say what you mean in the simplest terms possible. Some specific points from your post:
"market operating matrix of values". this, to me, reads as nonsense. I have a decent background in mathematics, econometrics and quite a long history as a trader, and I frankly cannot make heads or tails of this. I am assuming this is a meaningful concept to you so please explain in clear English what this matrix is and perhaps show us how to construct it.
"yields Gaussian distributions" - because the market operating matrix of values is not a concept I understand, we can't address this. Let me get out in front of this though by saying that, in market data, pretty much nothing is classically Gaussian. I am hoping your example of the market operating matrix of values will let us clearly see these Gaussian distributions.
Let's start there, please. Look at this as a request for clarity from someone willing to give your ideas a chance. Once we have addressed this core concept then we can dig deeper into your suggestion of applying the scientific method to market data.
I also abhor innuendo in a context like this. What precisely are you saying here? What is the story of Acruary?
A matrix has rows and columns. I crossed market pace with bar volatility. This generated a grid of cells filled with bar frequencies, I used 1600 bars. I keep it current by adding new bars and deleting old bars.
If a person examines the shapes of the contents of the rows or columns, he finds a simple shape is occurring. The name of this distribution plays a role in natrual occurances, By relating to the implicatiions of this type of natural occurance this makes the markets seem ammenable to using other natrural occurances as models for the market's nature. Thus the market is telling us rather than us assigning arbitrary contitions.
It is nice to find a natural distribution when woking up market data.
I use it along side the other components of the system: two hypotheses four parametric measures, and 9 cases of possible adjacent bar combinations. These allow the deduction of a pattern of market operation which is bilateral in its application. The pattern also serves to interlock the nesting of fractals all of whom also exhibit the pattern. A math consequence occurs: Binary vector math is used. It is worth noting the fact that probabilities do not apply and prediction is not needed.
All the above is a foriegn language that does not stand up very well to those who work from only a Conventional Wisdom language orientation.
In other words what I post is put to the Conventional Wisdom standard of exposition by Conventional wisdom followers. Conventional Wisdom is my second language it turns out.
My approach to extracting capital is to take the offer effectively and efficiently. Thus I am not free to use the Conventional Wisdom but must obey the dictatesof the market.
Crossing two variables like Market Pace and bar volatility to see the result (the cells are filled with bar counts) in terms of row and column distributions again is foreign to most traders. Why would knowing various market operating points be important? It does allow a person to see that the market migrates instead of jumping around.
Price has two measures: continuation or change. In trading, profit segments follow one another only separated by an end effect By measuring continuation routinely, a person gets to see when the end effect appears. Profits are taken and the next trade is begun.
The market dictated these measures. In deduction, Keynes and Carnap largly provided the foundations. Probability goes away and so does the need for prediction.
In the literature this was found to be a paradigm shift and no commonality resulted. Convention will prevail in the financial industry because of the business orientation. Amateurs do not have to be bound by jobs or working agreements. They just use services for a price and make money.
As I started to trade, I looked at the market's offer and how to take it. I concluded to be in the market and be on the right side of the market. It appeared that measurng the hold periods could be done. Their beginnings and endings could be measured as well. This consumed all the time the markets ran. The hypotheses for this were two in number and they were in the form of If volume, then price. All measures were first derivative oriented. On charts this is sloping lines for the two variables.
Interlocking fractals meant that trading on the center one gave you a leading set of signals from the next faster and the slower one was just a firm correct unfailing certainty context. In terms of emotions this engenders support, comfort and confidence and not the CW feelings of anxiety, fear and anger.
What emerged from being competent and fully adressing sufficiency was a method much like driving a car. It is largely unconscious and you do not experience anxiety from not knowing or fear from what isn't possible to go awry. This ia very adaptive to how ATS's are generated, All fixed blatant unambiguous rule sets. As trading goes from crude to highly refined, a person just moves from one shell level of focus to another more detailed shell. The more skill, the more frequent are the trades until the limit of human capability is met. The multiple of the ATR retrieved ranges from 3 to 6 times. A beginner may not make 3x since he is sidelined when his mind fogs up.
The see one, do one and teach one approach is how to fast track acquisition of skills and knowledge.
The hypotheses are annotated in realtime. By doing one pane there is an outomatic transfer to related panes. A couple of fractals can be done on each pane. By using the YM to lead the ES, you always know from annotating YM what the annotations are going to be for ES.
Drills focused on skills additions is the CPM. Having only one patterm makes it all very straight forward.
Stock trading is done from a Universe and as time passes you set up trades for the week and just excute according to the calendar and the volume leading price in position trading. Ultimately you work into 100 cycles a year @ 10% a cycle. The batting order is circulated just after the Sunday meeting each week.
Commodities trading goes through skill levels and as a person gets better, he does faster fractal trading. Today trading rates were every two to three bars profit segments. Turns were easy to carve since it was a quiet market.
- Status
- Not open for further replies.