Idea for a new indicator

[Antraman]

Hi there,

Have been playing around with NT5 charting tools, what an addition to NT5! Awesome.

Anyway, I have an idea for an indicator that you guys might want to build into a future version. Its something I have been using when looking for entry points in share trading, and that is calculating the number of bids to a certain depth, and the number of asks to a certain number of levels of depth, and weighing one against the other in an effort to forecast future short term price movements. As we all probably know, more bids (buyers) pushes the price up, and more asks (sellers), pushes the price down.

By totalling the number of bids against asks, it is possible to foretell which direction the next sale will turn, therefore making it a bit easier to a) pick a finely tuned entry point, and b) have an idea of whether the market will go long or short.

It has been pretty successful for me with shares, as it is a slower market - the numbers of bids and asks in a stock's market depth chart change pretty slowly and I can calculate these pretty quickly in my head, but with the futures the numbers of bids and asks changes very rapidly, making it next to impossible to make such calculations mentally.

I see on the SuperDOM that bids and asks are shown to a depth of 5 levels, how difficult would it be to have these figures filtered into a program that weighs each up against the other and outputs a +/- figure, indicating a possible future long or short trend?

Anyone got any throughts on this?

cheers,

[NinjaTrader_Ray]

imported post

I think it was NTV2 that had a depth meter on the SuperDOM that did exactly as you described below. It was not that successful in that not a lot of people incorporated it into their trading.

I have been thinking about adding the CVD (Cumulative Volume Delta)indicator which is in my mind a better volume based indicator than OBV. What it does it plot the differnce between volume trading at the ask vs trading at the bid. It does this using one level of course.

Ray

[Antraman]

imported post

Interesting. What a shame I missed it.

The CVD does sound very similar to what I am thinking of. I might check it out.

[Antraman]

imported post

Been checking out the CVD (Cumulative Volume Delta) and while it at first seems similar to what I was asking about in post below, it is a bit different.

CVD appears to measure the volume of TRADED bids and asks, which means it's a lagging indicator. I was really looking for something that looked a bit more to the future, something that weighs the current bid and ask orders already placed, but not yet traded. CVD uses only 1 level of depth, as it is recording actual trades, however, I believe that the 2 and subsequent levels also have an influence on actual price movement, and also orders can be cancelled anytime after they are placed and before they are executed, so the trading environment is changing all the time. It is this environment that I am hoping to record with an indicator.

My suggestion (please excuse me, I haven't put a *lot* of thought into this) on how it could work is like this:

You have a horizontal zero line, with a label "O". On the vertical axis above the zero line would be no.# of asks, and below the zero line no.# of bids.

In the indicator options, you could select the depth (from the last traded price level) that you wish to measure, being anything between 1-5 (which is the number of levels that depth is shown in figures on the DOM).

Then a negative value could be applied to asks, and a positive value applied to bids, and the bids and asks are then added, to give a +/- figure. This figure would relate to bid/ask power, which could be represented by a signal line which could trace over the zero line, moving above or below to the extent of the difference between bids/asks (power). When the total of bids is greater than the total of asks, it will give a positive value, and vice versa.

The vertical axis would then have an incremental negative number scale under the zero and a positive incremental scale above the zero, which gives a readout of the quantity of the difference between total bids and total asks, so it would appear as in the attached graphic below...

To add detail, you could have a separate line for each level of depth, and have all 5 levels shown if you wanted, and each line can be edited to a different colour representing closeness to the traded price. Each level of depth could calculated separately, so you get a separate readout for each level.

You could take your data from the script that creates the text for the figures of cumulative depth in the DOM.

Does anyone think this is a worthwhile indicator to create? IS it possible to do in Create New Indicator Wizard?

cheers,

[Antraman]

imported post

This is the sort of thing I am thinking about...

http://www.nielsenindicator.com/

[cody]

imported post

Besides the complexity of guessing if bids/offers are real or simply meant to "spook" the market, how is this different than Money Flow? Or combining Money Flow with OBV?

[Antraman]

imported post

I don't know much about Money Flow or OBV indicators as I have not studied them yet, so can't answer that part of the question.

I see no "guessing involved in knowing whether bids/asks are real or not" - trades are put on and cancelled all the time, and the cumulative depth shows the number of bids/asks at any one point in time. The numbers change over time, and as they get closer to the price action, the more definite the figures become, to the point that they are actually traded and one knows exactly how many trades were executed at that price. The disparity between the numbers of bids and asks at level one depth determine where the next movement will go in order to satisfy the filling of orders.

This I think is covered by the initial posts on another thread http://ninjatrader.mywowbb.com/forum5/681.html

As I said, this method has worked pretty well for me in share trading, as the change in number of orders is quite slow, and I can compute the weight of bids against asks mentally, and can observe and generally predict small movements in the price action as a result of these varying bids and asks (as the Nielsen Indicator shows), so I believe there is a logic behind it, but in futures the changes in orders is too rapid for me to compute mentally hence my desire for a CPU to help me.

I also mentioned that it is really only to try and predict very small changes in price, for when one wishes to fine tune their entry point, not as an indicator of large, general trends. I think that has some bearing on the issue.

If you think its a load of dingoes kidneys, then please explain why...I don't see how your final comments in the above mentioned thread relate to my initial question or to Underground's question in that thread.

cheers,

[cody]

imported post

I dont have any problem with any indicator. I'm also happy to deploy capital to capture bid/offer spreads - if your idea helps me do this, thanks!

My question stems from orders I can place that reside on my PC, or broker server, and show say 100 shares on an ECN order book, while my true amount may be significantly in excess of 100. So, while your indicator sees my 100 shares if you bid the market as per your indicator, my selling increases - although you only see 100 shares. This is common in various arb strategies.

Which exchange are you using your idea on?

[Antraman]

imported post

I've been trading for a few years on the ASX through the Commsec and Westpac Broking online interfaces. I am presently studying and running sim in the Euro/USD futures with a plan to switch over soon.

I realise there are orders on the ASX and other exchanges whose quantities are undisclosed, I am still trying work out what goes on with that...and I realise this fact will confound the bid/ask market depth to a certain extent. Does this mean that the CVD will not display trades of orders with undisclosed quantities?

On the Commsec interface, any orders of undisclosed quantity are indicated by a "/u" in the market depth ladder, and they are very rare...

What are arb strategies?


cheers,

[Antraman]

imported post

double post - sorry

[cody]

imported post

Undisclosed orders, are also sometimes conditional orders, as in the order will activate only if the bid is a certain size, so it may wreck havoc with your plan. This is pretty common in NY as even retail brokers offer this.

Money Flow will show you if money is flowing in/out and may be an indicator you're interested in.

arb=arbitrage

[buderim]

I think it was NTV2 that had a depth meter on the SuperDOM that did exactly as you described below. It was not that successful in that not a lot of people incorporated it into their trading.

I have been thinking about adding the CVD (Cumulative Volume Delta)indicator which is in my mind a better volume based indicator than OBV. What it does it plot the differnce between volume trading at the ask vs trading at the bid. It does this using one level of course.

Ray

Hi ,

Did the delta indicator ever get made ?

Ian

[NinjaTrader_Ray]

Please take a look at some of the BuySellVolume indicators, this may be of value. CVD has not been implemented.

[rally]

Hi Ray,

I tried the BuySellVolume and BuySellPressure indicators, but nothing is showing, just blank panels, do these indicators work only in the live version of NT?

[rally]

Hi Antraman,

I like your idea, take a look at the flash demo on the left hand side of this:

http://www.tradingtechnologies.com/xtrader-pro.aspx

I'm very surprised CVD has not been implemented in NT. Also does NT have any plans to implement the other 2 volume indicators below the CVD in the above demo video?