Trading with Market Statistics V. Other Entry Points

[jperl 11]

NEWBIE has come a long way since his early days of using technical analysis. He no longer trades by the seat of his pants. He has a good quantitative feel for market statistics and he simply follows the statistics wherever it wants to take him. He knows that the volume distribution function contains all the information that he will ever need to institute a trade. He knows about the peak volume price, PVP, and can pinpoint that with good precision on his charts. He knows about the distributions average value, the VWAP, and he can follow it as it slowly evolves during the day. He knows about market volatility and he can quantitatively measure it using the standard deviation, SD, of the VWAP. He knows how to determine the market's skew from the difference between the VWAP and the PVP (skew is proportional to VWAP - PVP). He has a simple entry technique, entering at the VWAP in the direction of the skew, a good profit point measured by the SD and a good stoploss point at the PVP.

 

(As an aside, a discussion of distribution skew, also called kurtosis, can be found at this Wikipedia site. http://en.wikipedia.org/wiki/Skewness

We use the Karl Pearson definition of skew which is (VWAP-PVP)/SD )

 

But he wants more. He's discovered that trade entries at the VWAP don't occur all that often throughout the day. He knows the market can give more if he just knew where else he could enter a trade beside the VWAP.

 

NEWBIE is about to have an epiphany.

 

Suppose he enters a short trade at the VWAP, exits the trade at the 1st SD. Then what does the market do? If it rarely returns to the VWAP, then the only other thing it can do is drop below the 1st SD. Now here is the epiphany.

 

Another entry point is at the 1st SD itself.

 

NEWBIE knows this has to be a good entry because the volume distribution function being skewed to the downside, (VWAP-PVP<<0) will remain skewed to the downside only if the price action stays below the VWAP. Only two conditions will change this, a)The market stalls near the 1st SD such that the PVP abruptly changes to near the 1st SD or b)the price action takes the market back up to the VWAP and higher. We will discuss these two conditions in later threads, but first things first. Watch the 31 minute video and see how NEWBIE takes a trade at the 1st SD. Where is his profit target? The volatility is still in force. His profit target can only be one place, the 2nd SD.

 

YMshortJuly26

 

NEWBIE is about to have a second epiphany. He's about to learn how he might change losing trades into winners by changing his ideas on stoploss placement. Check out [thread=2189]part VI[/thread] to find out how.

YMshortJuly26.swf

[darthtrader 10]

congrats newbie;)

 

not much to add, I love how your building on each idea in the videos.

 

one thing, If you were starting over as a trader would you just watch 3 or 4 markets with less correlation (ym, gold/silver, ags) and then only take trades at vwap and 1st std dev? that should give more than enough opportunity I would think?

[darthtrader 10]

out of curiosity and with something you posted in these threads about stops...if price in that video had actually gone back to the vwap would you have added to your short position if all else stayed the same since probability would have even been more in your favor compared to your entry?

[jperl 11]

congrats newbie;)

 

not much to add, I love how your building on each idea in the videos.

 

one thing, If you were starting over as a trader would you just watch 3 or 4 markets with less correlation (ym, gold/silver, ags) and then only take trades at vwap and 1st std dev? that should give more than enough opportunity I would think?

 

Yes, for sure. However stay away from anything that does not have high liquidity. Less than 100K contracts/day, and you are hurting for good statistical data.

[jperl 11]

out of curiosity and with something you posted in these threads about stops...if price in that video had actually gone back to the vwap would you have added to your short position if all else stayed the same since probability would have even been more in your favor compared to your entry?

 

You are catching on fast darth. In the next thread I will be discussing scaling-in

[Guest cooter]

Yes, for sure. However stay away from anything that does not have high liquidity. Less than 100K contracts/day, and you are hurting for good statistical data.

 

There are quite a few contracts that trade less than 100K/day, and are still quite volatile and liquid.

[jperl 11]

There are quite a few contracts that trade less than 100K/day, and are still quite volatile and liquid.

Volatile for sure, but you don't want to be trading anything with large spreads .

So cooter how about giving us a list that you consider liquid and I will take a look at the volume distribution function.

[Guest cooter]

Sure, but define "large spreads" :p

 

For example, try the CBOT grains:

 

Dec Corn

Sept Wheat

Nov Beans

 

There's more than enough action here to make a decent living, IMHO.

[houghton 10]

Hello Everyone,

 

My first post after spending nearly two months absorbing all the information available on this site. Let me first congratulate the many contributors who do so in a professional non-rancorous manner. It is truly refreshing.

 

I have two questions regarding this thread. First, is there a TradeStation eld for plotting the PVP? Second, has the method for calculating/plotting the SD for VWAP been settled?

 

Thanks,

Karl

[thrunner 10]

Thank you Jerry for the summer lessons I came back from vacation found that you have really done some interesting work. My question for this video is you said "there is no reason why it {target} shouldn't hit the 2nd standard deviation, the volatility of the market says it should". How exactly are you calculating this? It seems to newbie and me that 95% of the prices have occurred within SD2, so going beyond 95% is not a high probability trade. Are you basing this on skewness calculation? (eg, higher the skewness, the greater probability of exceeding SD2). Is there a skewness risk involve in this trade? http://en.wikipedia.org/wiki/Skewness_risk

[jperl 11]

Thank you Jerry for the summer lessons I came back from vacation found that you have really done some interesting work. My question for this video is you said "there is no reason why it {target} shouldn't hit the 2nd standard deviation, the volatility of the market says it should". How exactly are you calculating this? It seems to newbie and me that 95% of the prices have occurred within SD2, so going beyond 95% is not a high probability trade. Are you basing this on skewness calculation? (eg, higher the skewness, the greater probability of exceeding SD2). Is there a skewness risk involve in this trade? http://en.wikipedia.org/wiki/Skewness_risk

 

Very good questions thrunner and good observation as well.

First let me clear up one poorly understood idea about the SD. Most traders think that 68.3% of the data falls within 1 SD and 95% falls within 2 SD. This is only true for the normal or gaussian distribution. For skewed data, that is data that deviates from normal behavior, the best estimate can be obtained using Chebysev's inequality, which states that no less than 50% of the data falls within 1.4 SD, and no less than 75% falls within 2 SD. No less than 89% falls within 3 SD. These numbers are quite a bit different than that for the normal distribution.

 

These numbers are of course lower limits. The exact values could be computed from the distibution function. But I don't think there is much to be gained knowing that say 55% of the data rather than 50% of the data fall within 1.4 SD.

 

Another important point which I stated as a theorem in the SD thread, is that for any arbitrary distribution, computing the SD with respect to the VWAP yields the smallest SD possible. What that means in practice is that if you compute the SD with respect to any other price (eg the 1st SD price), you will by the theorem get a larger value for the standard deviation. This implies yet a larger volatility at the 1st SD than it does at the VWAP.

 

These two pieces of information taken together suggest to me that getting to the second SD (computed with respect to the VWAP) is not all that unreasonable although of course with greater risk than trading at the VWAP. Getting to the 3rd SD however is problematic.

 

I will discuss in the next thread, about what to do when you take a trade at the 1st SD and the price action does move against you. There is still room for pulling a profit out of the trade.

[darthtrader 10]

Volatile for sure, but you don't want to be trading anything with large spreads .

So cooter how about giving us a list that you consider liquid and I will take a look at the volume distribution function.

 

Hey Jerry, can you look at corn? Corn traded 100k contracts today, beans/wheat were much less.

I don't think gold trades enough to get a good sample.

 

Do you get that many different vwap setups on the russel vs ym? I'm not familiar with the russel really to know how correlated it is vs ym.

maybe just watching corn/russel/ym would be a good start.

[jperl 11]

Hey Jerry, can you look at corn? Corn traded 100k contracts today, beans/wheat were much less.

I don't think gold trades enough to get a good sample.

 

Do you get that many different vwap setups on the russel vs ym? I'm not familiar with the russel really to know how correlated it is vs ym.

maybe just watching corn/russel/ym would be a good start.

 

Well if you just trade the VWAP itself, you should follow 4 or 5 different instruments. YM, ES, NQ, ER2, ZC would be good choices. I only follow ER2 these days, since I trade all over the chart.

[darthtrader 10]

Well if you just trade the VWAP itself, you should follow 4 or 5 different instruments. YM, ES, NQ, ER2, ZC would be good choices. I only follow ER2 these days, since I trade all over the chart.

 

Very interesting on many levels.

 

I take it you focus on ER2 now because of volatility?

another thing is how would you order these contracts as far as what to look at as a newb for setups? the most interesting thing about your threads is every time its the total opposite of what i would do off the cuff. With that in mind I would think you would order them... ER2, ES, NQ, YM, ZC as far as what to look at first?

[jperl 11]

Very interesting on many levels.

 

I take it you focus on ER2 now because of volatility?

Actually ER2 recently has gotten more volatile than even I like. It's getting close to my risk tolerance.

 

another thing is how would you order these contracts as far as what to look at as a newb for setups? the most interesting thing about your threads is every time its the total opposite of what i would do off the cuff. With that in mind I would think you would order them... ER2, ES, NQ, YM, ZC as far as what to look at first?

 

For setups, it doesn't matter what contracts you look at. More important is what the volatility is compared to your risk tolerance. Once you establish a risk tolerance for yourself, and you know what the volatility is of the contract you are trading from the SD, you will know what to trade.

[Dogpile 10]

<<I have two questions regarding this thread. First, is there a TradeStation eld for plotting the PVP?>>

 

you can use Tradestations 'Matrix' window to see the PVP (longest volume bar)...

[houghton 10]

Thanks, Dogpile. That, of course, requires manual insertion of the PVP line onto the price chart. I'm wondering how Jerry plots his PVP which updates automatically as PVP changes during the day.

 

Karl

[nelo 10]

Jerry,

 

In (almost) all books is suggested to strive to achieve 3:1 RRR (reward to risk ratio). I do not doubt your method works (with enough practice) since I am watching it and there is nothing unreasonable. But the RRR seems more like 1:3 (taking trade on 1SD), so with one loser you have to have 3 winners to break even at least. This leads to huge and fast drawdowns - if it is not compensated with a great winning probability.

 

Every method has some psychological draw-backs, or requirements, for someone is more suitable scalping and for someone trend following. What psychology set-up or "traders mind" would you recommend and what expectation a trader should have to successfully follow your method?

 

Maybe you could also share some of your path, how did you discover your style. I believe most of us will appreciate it.

 

Thanks for great videos and sharing your experience.

[jperl 11]

Jerry,

 

In (almost) all books is suggested to strive to achieve 3:1 RRR (reward to risk ratio). I do not doubt your method works (with enough practice) since I am watching it and there is nothing unreasonable. But the RRR seems more like 1:3 (taking trade on 1SD), so with one loser you have to have 3 winners to break even at least. This leads to huge and fast drawdowns - if it is not compensated with a great winning probability.

 

Good observation nelo. I was wondering when someone would see this. The point I will make here, is the old paradigm about choosing trades with a 2:1 or 3:1 reward/risk ratio with fixed stop losses for most traders results in a slow bleeding of their account. We are going to cover this topic in more detail in the next thread, coming soon.

 

Every method has some psychological draw-backs, or requirements, for someone is more suitable scalping and for someone trend following. What psychology set-up or "traders mind" would you recommend and what expectation a trader should have to successfully follow your method?

Probably the most important mind set that I had to overcome, was giving up the idea of fixed stoploss on every trade and substituting the idea of risk tolerance instead. It wasn't until I did that, that I became a profitable trader.

 

Maybe you could also share some of your path, how did you discover your style. I believe most of us will appreciate it.

Thanks for great videos and sharing your experience.

 

My style is something that developed over many years. I was initially a strong proponent of classical technical analysis and traded futures and stocks for quite a number of years using classical methods. I oscillated back and forth between swing trading and daytrading, but was never satisfied with the results. Some years were profitable, other years were not. There was no consistency. I slowly came to the realization that classical technical analysis was not going to yield a consistent picture of market behavior. It was too heuristic. I wanted day to day consistency. I looked very carefully at market profile analysis. Realized that there was something there but it was woefully incomplete and in some cases just plain wrong. I wanted to be able to write my own software, but there were no good charting packages for doing that until ensign software came along. Being a student of molecular simulation theory, I knew enough about statistics to realize that the logic of the market could be found in a proper statistical analysis of the data. That coupled with understanding risk tolerance and trade management is where I am today. My trading is now quite consistent and I am happy to say has become quite enjoyable. I am both a teacher and student. I've been both my whole life and I am happy to share with you what I've learned about market behavior. There is still much about market behavior that I don't know and learning about the markets will be a lifetime experience.

[nickm001 10]

Jerry, Thanks for sharing your experience.

I am not sure if you answered this question before, but is the volume histogram, peak volume, VWAP and plotting SD around vwap something that is part of esign SW or did you have to program those yourself? I have checked esign site and found TPO based histogram, vwap is there, but did not see anything else from your tool set.

[jperl 11]

Jerry, Thanks for sharing your experience.

I am not sure if you answered this question before, but is the volume histogram, peak volume, VWAP and plotting SD around vwap something that is part of esign SW or did you have to program those yourself? I have checked esign site and found TPO based histogram, vwap is there, but did not see anything else from your tool set.

 

volume histogram: available

peak volume : available

VWAP : available

SD : not available but can be programmed by user as a DYO

Also if you ask Howard Arrington (owner of Ensign) he would probably make it available if there was enough demand.

 

My own software is presently not available for general use.

[nickm001 10]

That is a good news, that volume histogram and peak volume are available. I am not sure if the code for volume profile is open and can be expanded to calculate SD, which would not be too bad of a task.

 

I am tradestation user. All code for indicators is open, and can be used for further development. It is unfortunate that TS does not provide any of the above tools; user forum provided VWAP.. The rest is DYO.

[BlowFish 10]

I am sure I have seen a Tradestation study that does the PoC as a horizontal line that 'stair steps' when a new level becomes the new highest volume level. This is a great advantage when looking back at historical charts.

 

Unfortunately for the life of me I can't locate it.

 

Has anyone got 'pseudo code' (or ESPL would do) for the standard deviation lines. I would be happy to write and post EL code from that and would not have to think too hard about the maths :sleep:lazy:sleep: I know.

[Dogpile 10]

feel free to help out at the Tradestation forum site for this:

 

https://www.tradestation.com/Discussions/Topic.aspx?Topic_ID=66875&SearchTerm=VWAP&txtExactMatch=

[dbntina 10]

Jerry,

 

I am really enjoying how you use market statistics to trade...thank you for sharing the insights into your method. I am new to intraday trading and it is refreshing to see a new twist on plotting distributions (like market profile).

 

I had a more experienced trader tell me one time that you have to find a method that speaks to you. Being mathematically inclined I am really enjoying this way of looking at the market. It is speaking to me "so to speak":D

 

I am really looking forward to seeing some more threads. I am following the VWAP trade and I haven't got anything in the last few days on the ES and NQ (tradewise). Of course I have TS and I have a feeling that my VWAP and SD bands may not be correct.

 

Also to everyone else...I can code a little bit (programmer but not that familiar with TS) and am willing to lend my hand at helping out with the code in Tradestation.

 

Looking forward to more installments, even though can't really trade it due to indicator limitations :sad:

 

What a great forum, thanks for contributing Jperl! Willing to contribute but don't have many insights just yet.

 

One thing I am starting to see is that there are some good methods out there but if they don't "fit" you or "speak" to you they are very difficult to follow. I use to think that was a bunch of crap...now I am starting to see that it is true though.

 

dbntina