After Trading for several years now using TA and watching T&S all the time I have the feel for the price-action. I'm not a programmer so that's why I haven't tried the following idea. How about adding all the volume that is traded at Offer and adding all the volume that is traded at Bid for a certain amount of time (i.e. since the bottom of a trend - the starting point should be defined as soon the market is in a trend) This could give an additional information about the balance of power.. wouldn't it? Is someone already doing that or has experience with something similar?
not exactly... it would be adding the entire volume during a certain period. The key would be to find certain patterns that show you that a certain amount of contracts have been bought during the recent trend. Then I believe when I would follow that I could get some sort of key volume that was traded at offer during an upward trend that could indicate that the dynamics are changing.
Then you just want a rolling N period sum of volume. That shouldn't be hard to implement in almost any backtesting platform. --laz
hmmm.. it would seem that you would get random readings because the bid/ask ratio of trades would be different depending on why the market was making a bottom (if it is even making a bottom).. in other words, you would get a different result from an aggressive seller leaving the market as opposed to an aggressive buyer stepping in and bidding it up.. interesting idea though.. -qwik
I just believe that in a trend there is a point where all the money, that market participants in a certain market have, is in the market. So at that point there is no new money that can support the trend... except for the short-people changing positions.. but also at a certain point shorts go out of breath. In this case the only possible scenario for the market is to turn. I just ask myself if it is possible to see some sort of pattern in adding the volume that is traded at the offer and the volume that is traded at the bid. As I can see so far, none has tried that. I will put some more thinking in it and I'm sure that i will be able to find some way to translate this idea into a usable tool.
respectfully, i think both of these arguements might be worth reconsidering.. for one thing, even if it were true that all market participants with intentions toward a particular vehicle were 100% committed in a position, it isnt possible to know how many other participants might decide to take a position 2 minutes later.. also, supposing that you are watching xyz go up and are 100% certain that ALL long market participants and those intending to participate have entered the market.. there is no way to know if the market will turn.. suppose no one steps up to sell? what happens if the general market lifts the issue higher without any buys? its not possible to know what will happen.. -qwik
I believe the idea is good theoretically and would have worked 2 years ago. However I see 2 major problems. 1. Block Trades (sometimes they have no impact on the market, because they are settled outside the market. This program will account for them and give a false direction) 2. Funds playing. As of right now there are a lot of funds playing the stocks - throwing them left and right. Even though momentum is picking up - still there is a lot of uncertanty. I think human judgement is still needed to trade this strategy successfully.
I have to agree with the above, and add that the other variable is that the "uptick" (offer) volume would also include "short sales" which then adds to "short stock interest" which is yet again another consideration. Human judgement can certainly be assisted with automation and graphic displays, but common sense will usually prevail (no machine, yet, can beat the market consistently)......we've made some pretty good inroads, however. FWIW Don
You will go through several iterations of what you are chatting about. Right now, you have not considered the chart you want to look at as yet. When you do here are some of the next things that come up for any bar duration chart. You will find that above a certain rate of volume the market moves. This is the mid range volume and it supports trends. You will also see that the market goes laterally in another lower range of volumes. The lowest range is what you will will want to detect because it is the volume that just precedes the best break outs for trends. The last volume range is the one to use to make money the fastest. It is the highest volume range. For every market there are a set of these values. Naturally the great value in knowing them is that you can tell when the they change from one subset to another, you have a market change telegraphed to you. you are exploring that now and have not yet discovered there are several different kind of volume signals. Interestingly enough, when you monitor these, you see, usually, several minutes in advance, what is coming up. here in ET people are usually stuck on prediction. This is anticipation at it's best. Some people in ET do not know the difference and others do not thinthere is a difference. You have an advantage ovr all thse types who are stuck where they are stuck. It is a mellow place to be to have this advantage over so many others. I always couple the volume stuff with the BBid and Basked as well. I do their ratio so I know which way the market is going. The minority rules as has been discussed elswhere. Also, I can easily estimate when the Bbid and Bask are going to change their respective values. With these items working nicely, then I can spend my time monitoring the price to hone making money. This is the main event. There I use three simple sets of "monitoring search" alarms (Boolean constructs that are flawless in application). One set gives me a "perspective" of the market's climate. The important set gives me precise timing for changing which side of the market I am on. being on the right side of the market at all times assures a continuing stream of profits. And the last set simply, when necessary, trims the very short term perturbations that occur when some portion of the edge traders ae taking it in the shorts and are in reaction. Trimming means washing and reversing (one action) so that no drawdowns ever come up for me. I may do this trimmng up to 8 times a day. I do the ES mini. If you want the sets of values for volume and the three sets of "monitoring search" instructions for price, let me know. They go with a log. I posted my unscaled results per contract in ES for Friday elsewhere (12.3/contract). They are very different when the scaling aspect is added in to the mix.