Trading with Market Statistics - Questions

[metalhead 10]

I'm hoping this thread will help me better understand the "Trading with Market Statistics" threads.

 

In Part III we learned:

 

Long Entry:

Price> VWAP > PVP

 

Short Entry:

Price < VWAP < PVP

 

 

In Part IV, we learned that Newbie takes trades at the VWAP in the direction of the skew mentioned in part III (so in accordance with the Price/VWAP/PVP rules we learned).

 

This is where I get confused. When I look over my charts, it seems like this doesn't work quite often. Consider the following chart. Also, I'm aware that I could be misinterpreting a lot of this, so please tell me if my comments on the chart are wrong.

 

Light blue line = VWAP

Orangle line = historical PVP

Blue dashed lines = VWAP SD

 

 

 

Like I said I'm still a noob at this, I just haven't been able to make it profitable so I figured I'd ask what I'm doing wrong. Thanks.

 

Everything prior to this point is referencing the first attached pic.

 

 

edit - weird... I made another reply but they got merged... twice. So... let me at least edit it so it makes sense...

 

Question 2: am I correct in assuming that it's best to do nothing when price doesn't line up in your favor, such as when price is below PVP but VWAP is above PVP? This question is referencing the second attached pic.

 

(also, note the two circled areas... wouldn't these have been long entries as defined in Part III? they all failed)

 

Thanks.

Edited December 2, 2008 by metalhead

[Head2k 10]

As I said in another thread, using VWAP, its associated SDs and skew of daily price/volume distribution is just a way to define trend and to find potential support and resistance areas.

If you think of it like this, if you understand it as a tool, you will find you need something more to make an entry. Let me compare it to traditional way of imposing structure on price data. If you use trend lines and horizontal S/R lines, does it mean that if you are in a downtrend and price turns upwards to test a potential resistance level that you sell at that level automatically? Or do you want at least some confirmation to validate your entry?

I am a newbie-wanabe-trader, but I guess that one should seek for some kind of compromise - what sort of confirmation can I afford to obtain still a good price, and OTOH what sort of confirmation do I need so I am not sucked into too many losers. And when to move stop to breakeven to protect yourself but also not be shaken out of profitable trades too often? I think your method should be more complex than "sell at that line of other line is below yet another line".

Jerry's threads are only the beginning, they serve for the familirisation with this kind of structure and with some implications that it brings. But the profitable method is something you have to make yourself (I am not yet there either).

 

Maybe Jerry has better points to say on this subject...

[jperl 11]

In Part IV, we learned that Newbie takes trades at the VWAP in the direction of the skew mentioned in part III (so in accordance with the Price/VWAP/PVP rules we learned).

 

This is where I get confused. When I look over my charts, it seems like this doesn't work quite often. Consider the following chart. Also, I'm aware that I could be misinterpreting a lot of this, so please tell me if my comments on the chart are wrong.

 

 

Metalhead,

Head2k has aptly pointed out that the Trading with Market Statistics threads are less about defining a trading method and more about establishing a way to analyze the price data. How you use this information will depend on your specific trading style. I know that newbies haven't established a trading style, so I've tried to present a way that newbies might get started in using market statistics to establish one.

It's up to you to modify what I've presented to suit your own needs.

 

If you enter a trade and the trade moves against you as you point out in your comments above, you have to have a plan as to what you will do. There are several ways you could turn the trade entries you indicated in your charts from losing trades into winning ones, including scale ins and trade reversals. You could have also passed up the trades by using something like the Shapiro effect which I discussed in post16541. Regardless of what you do, the market statistics should help you decide how to manage your trades.

 

 

Bump:

 

Question 2: am I correct in assuming that it's best to do nothing when price doesn't line up in your favor, such as when price is below PVP but VWAP is above PVP?

 

If you are a newbie, Yes, otherwise, No. When you read further in the Market Statiistics threads you will discover that you can develop ways to trade almost anywhere on the price chart. How you do this depends on understanding HUP and how it can influence price action.

Edited December 3, 2008 by jperl

[lj2500 10]

Hey MetalHead, what chartingprogram are you using? Tradestation? If so, where did u get the VWAP and PVP code (I have vwap code already, looking for pvp)

 

Thanks.

[metalhead 10]

SierraChart.

 

Oops i have to extend my message to at least 20 characters.

[Hlm 10]

Metalhead, a few comments from what I see.

 

(all in my experience and opinion)

 

One should not take a trade on a vwap crossover (transition) without signs of a)failure and b)restest. By failure I mean a solid break to the other side. The market must show interest in that direction. As for retest, this usually is a classic flip/pullback with no interest found in the previous direction. Most of the trades shown on the first chart are done right at crossover which I would suggest against.

 

Now, as for what I find to be one of the most important concepts. Jperl talks about the idea of "randomly" starting the charts. Though I personally would not suggest this, it will give you similar results to starting at the beginning of each day in today's market. I say this, because nowdays markets can be decently active during off market hours. We are also dealing with much larger timeframes than usual so a one day profile for the pvp/vwap may not be enough. This is why I suggest to traders idea of starting the calculations and "important" areas. For example...swings highs and lows. At which point wait for the "failure" through that area and then watch for the "flip point" on the other side of the vwap. A move up away from the vwap, and then a break down below it creates a nicer distriubtion (created by significant market action - swing high/low) to work with. In other words, base your statistics off of the basis of the market (hh/hl - ll/lh). Not only will this keep you with the trend, but it will allow you to see what time frame you are dealing with.

 

Maybe this will give you some ideas.

[metalhead 10]

Thanks for the replies.

 

Apparently I misinterpreted the original series of threads. I thought it was a specific defined trading system. Oops! :doh::crap:

[BlowFish 10]

Whilst Jerry does not present it as a system per se he provides ample information how you might trade it systematically. In particular there are a few techniques presented that cover potential entries over a variety of circumstances.

[stanlyd 10]

Metalhead,

 

One thing I know I used to do was to get so focused on what I was looking for (i,e, setups) that I would ignore the "bigger" picture. In your chart example I refer to the trend. Take a look at the chart and you will relize that the first trade worked out because why?, you took a trade in way of trending price action. Now look at the losing trades. Notice they are all counter to the the price action trend. So in those losing trades you would have called for the market to turn at those points. So maybe take a moment before entering and look at how the price is trending so you know if your are going with the current chart trend or against it.

[bh_trade 10]

A gentleman named dbntina has coded up the pvp for Tradestation. It is not as simple as plotting the indicator on your chart, you will have to insert a tick chart into your workspace, insert the indicator into it, then use the ade (all data everywere) utilities found on the TS forums to bring the actual pvp plot into the chart you intend to trade from. Check out the indicator section - I believe the indicator package and instructions are in there.

 

Hey MetalHead, what chartingprogram are you using? Tradestation? If so, where did u get the VWAP and PVP code (I have vwap code already, looking for pvp)

 

Thanks.

[BlowFish 10]

Metalhead,

 

One thing I know I used to do was to get so focused on what I was looking for (i,e, setups) that I would ignore the "bigger" picture. In your chart example I refer to the trend. Take a look at the chart and you will relize that the first trade worked out because why?, you took a trade in way of trending price action. Now look at the losing trades. Notice they are all counter to the the price action trend. So in those losing trades you would have called for the market to turn at those points. So maybe take a moment before entering and look at how the price is trending so you know if your are going with the current chart trend or against it.

 

Firstly there is an elegant and effective way presented early in the series on determining the trend. It is through the position of the VWAP in comparison to the PVP.

 

I think if you look at the chart without the benefit of hindsight the trend was not clear from the price action. In fact it looked like a trend less day despite running up and down (I too have the benefit of hind sight ). I would bet that the day had a symmetrical profile (for the data shown) and it could be argued that the the VWAP and PVP where 'close'. Under those circumstance it would not be wrong to consider counter trend trades as detailed in "Section VIII Counter Trend Trades in Symmetric Distributions".

 

There where a couple of areas that I think can be a bit 'tricky', one of them is deciding whether to follow sections III V & VII and when to follow section VIII.

[stanlyd 10]

I would have to respectfully disagree with you blowfish. Yes the vwap in relation to PVP is a good to determine trend ,however it depends on the timeframe you are looking at. Just because one timeframe does not have a "trend" does not say another smaller (or larger) one does not as well. Just placing the vwap and PVP from the start of the day along in my testing will not keep you in the other "smaller" trends that develop intraday.

And it's not hingsight. Let me give you an example based on metalheads chart.

Look at his chart around 12:03:30. That is the start of the move back to the vwap he has charted. Well for me at that point at 12:03:30 in which the a higher high was made. or in other words the slope of the last move down went from negative to positive, means "trend" change is in works and so armed with this information I would now be looking to get aboard that move(if you take trades towards vwap,which could be something I avoid )

So an entry would be at 12:24:14 AND at the vwap their would have been no reasosn to get out or take a short since that "trend" has not shifted to a negative slope. In fact in that trade I would have made the vwap a target to get all out of at least most out.

So I guess what I am saying is I do not see this as a trendless day. Maybe the "day" itself in macrostructure is trendless, but inside of that "trendless day" we do have a few trends to trade. I am open to some contructive feedback on this view as always.

Edited January 18, 2009 by stanlyd

[Hlm 10]

Very good points stanlyd, and that is why I find it more beneficial to focus on starting calculations at the trend changes (HH/HL - LL/LH) versus just at the open. In my opinion, in order to truly get the big picture, you must understand and be able to spot the fractal nature of the market. For me, there is no such thing as a consolidation or trending day when dealing with my strategy. If we spend the majority of the day in consolidation it's most likely because there is a larger time frame in transition. What does this mean for me? Well, look at the next higher time frame for a possible swing setup (if this transition is a pullback) or play the smaller time frame trends. By getting rid of these "stuck" time frames such as 15M or Daily, you start to truly understand how to read what the market is telling you. In other words, think of those as settings to simplify the information for quick reference, and not time frames. Also, by understanding the current cycle these fractal time frames are in, target profit and stops come very natural.

[BlowFish 10]

Stanlyd you are preaching to the converted here I am pretty much a pure price action trader. I was just pointing out that in the context of Jerry's method there is a simple and effective way of determining the trend of any sample of price data. The same points can be made about 'price action' the trend depends largely on how you sample the data. But of course that's just another way of expressing what you and Hlm are saying

 

Actually sampling data appropriately ('picking a time frame' if you like) for the moves you are trying to capture is a key component of trading. When I traded Jerrys methods I derived all his indicators from pure tick data and started calculations from a variety of start points depending on what I was trying to capture. Pretty much how you might use daily and hourly charts to determine intraday S/R levels to trade against on say a 5 minute chart.

 

I found the best way (for me) to trade the method was use market stats for the macro view look at faster stats (like an hourly sample) for a confluence/confirmation and then use traditional PA triggers. I should say I only traded the method for about 3 or 4 months but it 'worked' from the get go. It is really pretty robust.

[stanlyd 10]

agreed Blowfish. I think thats the point we and or new traders shoudld get out of this . The point that it's not a sytem so much that jerry presented, but a very powerful way of looking at the market data to meet your needs. That is where a lot of problems come in when starting out trading. I know i for one had no idea really what timeframe I was in and why i was in it. Like you said, it is a key component of trading for a trader to find that for themselves before applying the stats.