jperl

There is no proof Naveen, because my statement about equality is only partially true. It's true for the median value by definition. You can see why it is true for a normal distribution and other well known distributions when the mean equals the median. For all other non analytic distributions what you can say and prove is that the mean-median< 1 standard deviation The proof is here proof Short answer is yes. I qualify it with the statement, that there needs to be sufficient data to be statistically meaningful. So very low volume stocks would not work.

If you have followed and understood all the "Trading with Market Statistics" threads, from the very basic VWAP trades in [thread=2008]Part III,[/thread] and the SD trades in[thread=2130] PartV[/thread] to the more advanced trade types involving breakouts in [thread=2232]Part VII[/thread] and counter trend trades in[thread=2285] Part VIII,[/thread] then you are ready to apply your new found knowledge to the fast and furious world of scalping Scalp trading has many definitions depending on whose doing the scalping. The usual definition is trading for ticks rather than points. But this is a purely heuristic definition. My definition is more quantitative and is based on when I start my volume distribution computation as follows: a)If I start my volume distribution computations at the opening bell and continue until the closing bell, then that's a normal non-scalping trading day b)If I start my volume distribution at any time after the opening bell, and watch it for short periods of time, then I'm scalping. You know what a) means from the previous eight threads. What does b) mean? Until now, I have talked about the volume distribution when it starts from the opening bell, and runs until the closing bell, that is regular trading hours. However, there is no reason why you could not start the volume distribution computation at any other time during the day, let it run for say 15 minutes or so, trade off of it, and then restart it again. That's what b) means, and that's what I call scalping. By doing this you are essentially looking at the market statistic over a short time frame and asking the same questions with the same responses as given in the last eight threads. The net result will be, you will be taking many more trades and trading smaller standard deviations. This is what is meant by "trading for ticks rather than points". Scalping requires entries, exits, scale ins, scale outs, reversals and closes with one mouse click, otherwise you will miss the opportunity. To do this, you have to use a DOM (Depth of Market) or something equivalent as part of your trading platform. Watch the video and see how I do scalp trades using the DOM. ER2ScalpTrades Addendum: I was going to present a thread on the use of Hold Up Prices (HUP) which I've mentioned many times in these threads. I've decided not to do it now because it's quite complicated and I haven't yet found a simple way to present it. So I am going to delay that presentation until another time. ER2ScalpAug24.swf

That's actually saying a lot, considering that it applies to any type of skewed disitribution of arbitrary shape.

Unfortunately not. It's the very nature of skewed distributions that the "tail" on one side is not the same as the tail on the other side. So a tail test just doesn't make much sense.

Glad you find the statistics useful Dogpile. As far as the Shapiro Effect is concerned, it's not a pattern in the usual sense of patterns as defined by Bulkowski, but rather a simple statement of implied market direction. The breaking of a high/low is just my interpretation. You can use any other wait period of your own choosing.

The answer to your question emmster is a definite maybe. If the price action stays below the VWAP and breaks out below the 1st SD, then the bias would be down. But until that happens, all you will get are oscillations between the VWAP and the 1st SD below it. See the breakout thread [thread=2232]Part VII[/thread] for further clarification.

I looked at some of those algorithms. With an unweighted variance, they are simple. The computation with a weighted variance is a messy business and not worth the effort. I don't think they will improve computation speed.

If you want to weight the variance computation by the volume, that's exactly what you would do as described in [thread=2101]Part IV[/thread]. Think of it this way. A 10 contract trade can be thought of as ten 1-contract trades. To compute the variance you would have to include all 10 contract trades in the variance computation. That's identical to multiplying the square by 10. correct no

You are quite correct for static distribution functions. Note the emphasis on the word static. If you knew the distribution function in advance, you would know exactly how to trade it. As you correctly point out, if the distribution had positive skew, you would go long every time the price action dropped below the PVP and vice versa for negative skew. Similarly for no skew distribution, always trade toward the VWAP. This is the classical reversion to the mean theory. The problem as you realize, is that in a real market, the distribution function is dynamic. Note the emphasis on the word dynamic. Reversion to the mean, does not necessarily occur. Price action itself renormalizes the distribution as more prices are added to the time series. As a consequence, the relation of the price action to the VWAP and PVP becomes critical in choosing trade direction. Trading AWAY from the VWAP then becomes the more likely scenario. Trading TOWARD the VWAP is then relegated to symmetric distributions, which is the classical case. Trading away from the VWAP then defines the trend (if in fact you want or need a definition)as UP when the price action is above the VWAP and DOWN when the price action is below the VWAP. This definition breaks down of course when the distribution passes through the symmetric state. At that point there is no trend.

This is correct as defined in [thread=1990]Part II[/thread] Yes, this is also correct This is approximately correct. You can get reversion to the mean in any environment including a skewed environment. In a skewed environment you would look for a trade at the VWAP. In a non skewed symmetric environment, do nothing at the VWAP. Almost but not quite. You have to know where the price action is relative to the VWAP and PVP. For example for a positive skew your bias is long provided price action is above the VWAP. If price action is at or near the PVP, even if the skew is positive you don't have a bias until the breakout occurs. This is incorrect. Look for longs only when the price action is above the VWAP. Below the VWAP, wait for the price to move above and retrace before taking a long. Otherwise wait for the break out at the SD below the PVP and go short. Also incorrect as mentioned above. You trade short if price is below the VWAP, otherwise wait for a breakout. You might want to reread [thread=2232]Part VII[/thread] concerning break out trades. Don't know what you mean by current bias. If the skew is close to zero, there is no bias. Well this depends on your trading style. In a symmetric distribution you have a dilemma, in that you don't know how long the symmetry will last. This means you could take a trade at any of the SD's either long or short. In one case you would be trading countertrend, the other with the trend. That's why this is not for newbies. Using something like the Shapiro Effect will help. I think you meant the NQ video, the first video was a YM long. In any case the first NQ trade was a legitimate breakout trade into the low volume zone. It doesn't matter how close the VWAP is to the PVP, only that price action is in the PVP zone. Dangerous? yes, but nevertheless workable. Again, this will depend on your trading style. An advanced trader would take the trade if he thinks the VWAP is going to continue on down. Why miss the opportunity? If the Shaprio effect indicated the price action is going to continue down, pull the trigger. Yes, I would agree witht this interpretation. You would either exit the trade at a hard stop if the trade moved against you, or possibly reverse the trade for a further move down depending on how far below the 2nd SD you did the reversal.

Nick, the new SD is computed using the new VWAP. See post 14694 for an example.

I'm surprised at that Steve. Howard is usually receptive to DYO studies.

Good question Nick. Short answer is there are other situations where reversion to the mean might occur. Example being a skew sign flip does do it sometimes. But other than that I don't know what else would.

In the discussion about VWAP in [thread=1990]Part II,[/thread] we introduced the concept of skew, a measure of how the volume distibution deviates from a symmetric or normal distribution. The sign of the skew allowed a new trader to decide in which direction he/she should look for a trade setup. Positive skew meant look for long trades only. Negative skew meant look for short trades only. We have yet to consider trading aspects in markets with symmetric distributions. Like breakout trades discussed in [thread=2232]part VII,[/thread] trading symmetric distibutions is an advanced concept. Not for newbies to be dabbling in. A symmetric distribution is one in which the skew is very small or zero Skew = (VWAP-PVP)/SD ~= 0. There are a number of implications of this definition as follows: 1)a small skew means the VWAP is close to or equal to the PVP 2)Given a small or zero skew, it means that price action has moved across the VWAP at least once, otherwise the volume distibution could not be symmetric. Now comes the kicker: 3)If the distribution is to remain symmetric, it must continue to oscillate across the PVP and hence the VWAP. This implies trades of the following type: If price moves to the 1st or 2nd SD above the VWAP pull the trigger SHORT. If price moves to the 1st or 2nd SD below the VWAP pull the trigger LONG WOW- that's completely opposite to everything you've been told in the last seven threads. Up until now, every trade was taken moving AWAY FROM THE VWAP. Now you have to learn to take trades moving TOWARD THE VWAP. To trade a symmetric distribution, everything you have learned in the preceding threads is turned upside down. To complicate the situation, the condition for a symmetric distribution is fuzzy. It's defined with skew approximately but not necessarily 0. There is also no guarantee that it will remain small. For example, suppose the skew is slightly positive and price action is around the 1st SD below the VWAP. You would look for long trades back toward the VWAP. But it is also possible for the price action to continue on down with the VWAP crossing the PVP and continuing on down. Like the breakout trade, trading a symmetric distribution has to be done with great care. By its very nature, a trade taken toward the VWAP in a symmetric distribution is a counter trend trade. For example, when price is below the VWAP, the trend is down as defined in Part II. If you trade toward the VWAP then, you are taking a long entry in a down trending market. Similarly for shorts. Look at the first video and see if our trader can decide if the distribution is symmetric. Symmetric YM Trade Advice: If you want to counter trend trade in a symmetric distribution, use the Shapiro Effect discussed in post 16541 to decide on the entry. If the countertrend trade is taken at the 1st SD below the VWAP and the trade fails (price action drops below the 1st SD) you have two choices. 1)reverse the trade and take your profit at the 2nd SD or 2) hang on and scale in at the 2nd SD for the counter trend move back to the 1st SD. Our trader in the first video was so sure that he would not want to take a short trade. But now watch the second video and see what our trader thinks now. In the second video, our trader takes 3 trades, the first a standard breakout from the PVP area as discussed in [thread=2232]Part VII,[/thread] the second a counter trend trade, and the third in the trend direction after a retrace. The last two trades demonstrate the use of the Shapiro Effect when the distribution is symmetric. NQsymmetric trades Clearly trading symmetric distributions is as difficult as trading breakouts. The choices can be quite contradictory. YMsymmetricAug14.swf NQsymmetricAug16.swf

I realize that breakouts can be touchy trades, but try using the Shapiro effect to trade the breakouts (see post 16541 ). Sometimes there will be a retrace and you will be able to catch this trade for a big move.

At the open, you don't have any data for today. If I am not using any HUP from previous days (yet to be discussed), then I wait until the range of the bars is smaller than 1 SD, before considering any trades. This of course means I miss all that great action at the beginning when the market takes off in one direction or oscillates rapidly. So be it. We will talk about trading at the HUP on the open in an up coming thread.

Well condsider the profit in each case. Trade 3 contracts at VWAP, exit all at 1st SD. Profit 3SD Trade 3 contracts, remove 2 at 1st SD, remove 1 at 2nd SD. Profit 2SD for the first 2 contracts and 2SD for the last contract. Total profit 4SD. This one is not a good idea. If you trade 2 at the 1st SD and price moves against you back to the VWAP, you are going to want to put 2 more contracts on, so that break even is a 50% retrace. If you put only 1 on, break even is a 67% retace, leaving you hardly any profit potential back to the 1st SD. Which brings me to another point. If you are using risk tolerance as a trading philosophy, and you are comfortable trading only 4 contracts. Your initial entry better be 2 contracts or less. Do you understand why? Again if I am trading using risk tolerance, and 3 contracts is my limit, your only choice is 1 contract on, keep two in reserve. In this mode, if you enter 1 contract at 1st SD and market moves back to VWAP, you can put 2 more on and have break even at 33% retrace. That's usually a pretty good bet.

Not sure Nick what you mean by VWAP leading the PVP. If you mean for example that when the VWAP is less than the PVP and stays that way, then yes you would only look for short trades, until there is either a change in the sign of the skew or price action moves back to the PVP. At that point you are going to have to rethink any new trade entry. Ok, I think I understand what you may be asking. Let me phrase it as a question: Is there a distinction to be made between a)PVP moving to the price action and b) price action moving to the PVP. My initial thinking about this was similar to yours. That if the PVP suddenly jumped to the price action, it was still trying to establish itself and hence the price action would continue in the same direction. But after seeing a lot of this type of behavior (that is, PVP jumping to the price action) and noting that sometimes it continued in the same direction and other times it would reverse, I've concluded that the best you can do if you want to trade in this region is wait for the break through an SD or VWAP before taking a trade. In some cases this means taking a trade in the same direction as the trend, in other cases it means taking a trade against the trend. Is it dangerous? Yes. That's why I've waited seven threads to talk about it. As far as the stair stepping of the PVP goes, I think each stair step down say has to be considered on it's own merits. When the PVP drops from one point to a new point, you have to re-ask yourself the same question, is a reversal iminent. Will the price action take the market up and break out the nearest SD above the present price. You can also do the following thought experiment. Suppose you walked into a room for the first time and you saw a monitor displaying a real time chart with the price action at the PVP. Is there anyway for you to tell without going back in time, whether the PVP jumped to that price action or the price action moved to the PVP? The answer is no. Until the price action breaks through the nearest SD, you really can't say much about which direction you would look for a trade. Now compare that to walking into the same room and seeing the VWAP below the PVP and the price action below the VWAP. In this case you would know immediately which way to look for a trade.

Glad you asked Walter. That's coming up in the next thread on trading symmetric distributions.

The simple answer to your question walter is yes. What complicates the anwser however, is that the price action is in a region (low volume region) where a break out against the skew is very likely. Also, price action is on the wrong side of the VWAP, so the trend is up. In addition, price action is above the PVP which acts as a HUP (Hold UP Price) for the short. So be cautious here. Use the Shapiro effect to enter the short, but bail out immediately if the SD is violated to the upside. Similarly for longs. VWAP itself is available as a Design Your Own (DYO) study in ensign. For the SD's I'm sure if you send a note to Howard, he would be happy to write a DYO to compute the SD.

OK, sorry db, I had it backwards. In any case here is my data for the PVP values of NQ for August 15 on a 2 minute chart (times are east coast). You can compare these with your computations: TIME, PVP 9::30 to 10:18 ,1907.25 10:18 to 11:00 ,1915.50 11:00 to 11:16 ,1914.50 11:16 to 11:32 ,1912.00 11:32 to 11:34 ,1915.50 11:34 to 12:30 ,1912.00 12:30 to 14:18 ,1907.25 14:18 to 14:40 ,1912.00 14:40 to EOD , 1907.25 A lot of jumping around

There is something I don't understand dbntina. On your 2 minute chart, it looks like the tick by tick PVP is constant for the entire day at around the 1907 value, whereas on the 1 minute chart it jumps up for a short time and then back down again. Why the difference?

You are correct bh_trade. PVP does jump around especially at the beginning of the day when the volume histogram is first developing. It's important to know the relation between PVP and VWAP in order to understand the price action. Hopefully these threads will help you understand how price action is connected to this relation.

Yes, that is correct. Dbntina has also verified this. Thanks

OK thanks dbntina. If you get a comparison, I would like to send something to Howard in this regard.