UrmaBlue, question, do you think it's possible for retail to discern volume as a 'leading indicator' through tape reading? Or is it only possible with programming?
Kedwards, Thank you. It is most certain, as demonstrated below, possible to formulate leading indicators from total volume, buying volumes, selling volumes and the velocities of these volumes. From tape reading alone - unlikely. Peter Steidlmayer used to teach that the best indicators of future price have price as NO part of their calculation. Price based indicators such as MAs, Stochastic, RS, ROC & MACD can never lead price because they can't change until AFTER price changes. In the chart below the lines represent an adaptive moving average that comes with an adjustment for phase of the mid-points of each bar. This is the Jurik Moving average from Jurik Research and the best smoother we have been able to find anywhere and you can see at a glance that in every instance the lines are posted to the right of the price bars which means the average LAGS PRICE. The dots are an index of weighted biases based on different readings of both totals and velocities of total, buying and selling volumes. In almost every instance the dots are to the left of price which means they are LEADING PRICE. cheers
For those that have been able to reverse engineer these indicators none of this will be new information. For those that are still working on it, another hint - it's not just volume or transactions over time. The calculation involves both acceleraton and deceleration as well as measures of both buying and selling. While these spikes occur throughout the session they are present on almost every session extreme. The session extremes on both Thursday and Friday were all FORETOLD by this measure of commercial trade; On Thursday 08/27 the session low came at 0710+ and 7-8 minutes BEFORE that low there was a very intense spike in commercial buying as shown below: Some hours later, at 1144, and 1 minute before the session high there was a corresponding spike in selling: On Friday 08/28, just at the Open there was a sell spike that came some 7 minutes before and within .50 of the session high: Soon after the session high there was a buy spike that offered the opportunity for a "reaction trade" of some 4-5 points: Friday's session low occurred arount 0925 and was Preceded by a strong spike in intense commercial buying:
urma u mention buying and selling climaxes and the chart agrees with your analogy, do u know previous to the move if it was intense buying or selling, or just intense volume
Sometimes you will get whats called an volume spike before a reversal. For example, the market becomes over bought as all the buyers have spent their reserves to go long. Part of this volume extreme is of course larger traders trying to pick the top of the move and get in before the reversal. Since they are larger traders, they can take the heat of the market going against them a little longer before falling.
ammo, Thank you for your question. The calculation of this indicator involves much more than just volume over time. It involves acceleration, deceleration, buying volumes, selling volumes and other indications of the balance of trade. We describe the calculation of this indicator as - - "When taken in combination, the acceleration and deceleration of buying and selling volumes, total volume and the velocity/rate of change in the balance of trade reveal a certain dynamic that we find present at many, if not most, intra-session extremes." You will notice that in almost every chart posted above that the spike occurs well BEFORE the change in price. cheers
Cross currents from non-speculative trade can often confuse certain indicators. Premium arbitrageurs don't care about trading near an extreme they care that the premiums they are trying to capture are at the proper level. Because of liquidity issues almost all of this kind of trade occurs during the day session. Oftentimes during the nite session there is striking precision in the occurrence of these signals - last night was no exception:
Urma, does your indicator have to reach a defined level in order to be significant ? E.g. on one chart I see values of 30k+, on another it's in the range of hundreds. Thanks.
Marsupilami, Thanks for your reply/question. Yes indeed, there is what we call a "threshold of significance." That threshold is dependent on three factors - 1) the instrument/market 2) current settings for that particular session as read from our Market Heads Up Display (HUD) as demonstrated here as well as 3) certain parameters within the indicator itself. In the example from the nite session, naturally the threshold would be lower as the nite session only trades 10 - 15% of the volume traded during the day session. cheers
Hey Urma, thanks for your last reply in regards to tape reading vs. programming. Another question, could you show us an example of when the system got it wrong? In the bigger picture, I think this system is a fade strategy in anticipation (or in lockstep really, as demonstrated in some of your examples) of the new trend. But when does it get it wrong? For example, is it ever wrong in a continuation pattern? i.e., market gaps up on news, and flatlines. Have you ever seen a spike in commercial paper selling, but then price continues higher 5 points on ES? Or, perhaps my question really is, are the commercial traders you follow ever wrong?