Tape Read (Time & Sales) Strategies

[carltonp 10]

My grammar was terrible in last post... Sorry about that.

[daVinciLite 10]

I'm hoping that someone like dVL will provide some strategies. All I need is a strategy and then I can go away and program it in Excel.

 

The best strategy is market understanding and the ability to put price in context of the current trading bracket, both long term and short term, along with reading the order flow.

 

The ability to think in probabilities and possible scenarios is something that is completely unnatural - especially for technical people like us. So learning that is a gradual process.

 

I am a discretionary trader first and foremost. But I can't stop myself from creating a mechanical system that builds on my discretionary tools because I am still a programmer at heart. I do have some free stuff for Sierra Chart users that I posted in the SC message board under (Warrior Trader).

 

Regards,

dVL

[carltonp 10]

The best strategy is market understanding and the ability to put price in context of the current trading bracket, both long term and short term, along with reading the order flow.

 

 

dVL

 

I'm working on that mate....

[gosu 10]

Gosu, that I would agree is a problem. One way to overcome that problem is write an excel formula to alert me to the frequency of bids/asks. I understand that when large institutions start to accumulate they do so in small lots. So, I will write a program that will watch for that kind of action. Other than that I have don't know how to deal with that issue.

 

I should add if you have any suggestions as to what I should watch out for I would love to hear them. At the moment I'm just studying the Tape with a simple setup. Once I see a pattern I will write a program for it. I'm just looking for patterns right now.

 

I'm hoping that someone like dVL will provide some strategies. All I need is a strategy and then I can go away and program it in Excel.

 

I have no suggestions as to how to overcome the problem of hidden volumes on the bid and ask. GLOBEX allows orders to be shown that do not reflect the actual size being bid or offered; e.g., a bid of 1 contract at a certain price could actually be 1000 contracts. That policy would seem to make suspect any trade decision based on what's showing on the bid and ask at any given time regardless of the flawlessness of the data itself.

 

Another thing to consider regarding the usefulness of the bid/ask volume as a trade signal is that at any given time there are potential market orders that can overwhelm the inside market without first being bid or offered. In my view, potential volume from "invisible" market orders far exceeds that of the limit orders visible on the inside market.

[joshdance 10]

I have no suggestions as to how to overcome the problem of hidden volumes on the bid and ask. GLOBEX allows orders to be shown that do not reflect the actual size being bid or offered; e.g., a bid of 1 contract at a certain price could actually be 1000 contracts. That policy would seem to make suspect any trade decision based on what's showing on the bid and ask at any given time regardless of the flawlessness of the data itself.

 

I think carlton is talking about transactions that have been executed, not orders in the book.

 

Another thing to consider regarding the usefulness of the bid/ask volume as a trade signal is that at any given time there are potential market orders that can overwhelm the inside market without first being bid or offered. In my view, potential volume from "invisible" market orders far exceeds that of the limit orders visible on the inside market.

 

Ditto the above.

[joshdance 10]

Gosu, that I would agree is a problem. One way to overcome that problem is write an excel formula to alert me to the frequency of bids/asks. I understand that when large institutions start to accumulate they do so in small lots. So, I will write a program that will watch for that kind of action. Other than that I have don't know how to deal with that issue.

 

I'm not trying to discourage you, but big institutions hire very smart, highly educated people and pay them well to write complex algorithms that are designed solely for the purpose of hiding their intentions. It's not as simple as "a bunch of 1 lots means someone's buying big." Now, we may see large volume come into the market and base a trade decision on that information--I do this a lot. However, it is pretty safe to say that it's impossible for you to write a program that will discern the intention of a large buyer or seller, whether it be accumulation or whatever you want to call it. Some algos operate on a sub-second time frame; others on a minute, or hourly basis. There are simply too many variables, and the orders far too complex, for you to "figure out" what's going on. And this has nothing to do with reading the tape by the way.

[carltonp 10]

I think carlton is talking about transactions that have been executed, not orders in the book.

 

 

 

Ditto the above.

That's is correct mate

[carltonp 10]

I think carlton is talking about transactions that have been executed, not orders in the book.

 

 

 

Ditto the above.

 

Josh, is it possible to determine if a transaction has been executed at bid or ask, realtime?

 

Cheers

[joshdance 10]

Josh, is it possible to determine if a transaction has been executed at bid or ask, realtime?

 

Cheers

 

Open up a time and sales window, and look at the preferences for that window. You should see the option to color transactions executed at the bid and at the ask separately. Most people will color the transaction executed at the ask green, as this represents a market buy, and at the bid red, as this represents a market sell. This window will show every transactions. The time and sales window is "the tape." This is what people are looking at when they talk about tape reading.

 

What software do you use?

[carltonp 10]

Open up a time and sales window, and look at the preferences for that window. You should see the option to color transactions executed at the bid and at the ask separately. Most people will color the transaction executed at the ask green, as this represents a market buy, and at the bid red, as this represents a market sell. This window will show every transactions. The time and sales window is "the tape." This is what people are looking at when they talk about tape reading.

 

What software do you use?

 

Thanks Josh,

 

I use Esignals advanced a charting software.

[carltonp 10]

Open up a time and sales window, and look at the preferences for that window. You should see the option to color transactions executed at the bid and at the ask separately. Most people will color the transaction executed at the ask green, as this represents a market buy, and at the bid red, as this represents a market sell. This window will show every transactions. The time and sales window is "the tape." This is what people are looking at when they talk about tape reading.

 

What software do you use?

 

Josh, I think I've been bit of a numpty (British term for what Americans call a Putz).

 

Do you remember when Gosu said 'That policy would seem to make suspect any trade decision based on what's showing on the bid and ask at any given time regardless of the flawlessness of the data itself.'

 

What is exactly what I was doing, when it actual fact I thought I was doint what you thought I was doing when you said I think carlton is talking about transactions that have been executed, not orders in the book.

 

It's still possible to write a formula for that, just going to require a little more thought....

 

Thanks mate....

[Tams 20]

this video might be of interest to you (post #3056)

http://www.traderslaboratory.com/forums/technical-analysis/6320-price-volume-relationship.html#post128204

[joshdance 10]

this video might be of interest to you (post #3056)

http://www.traderslaboratory.com/forums/technical-analysis/6320-price-volume-relationship.html#post128204

 

I did not watch all of this video, or the subsequent ones, but this explanation on the first part of the video where the software vendor offers "proof" is laughable. Sure, one example where heavy selling leads to price moving down--hardly a proof. If it were this simple, you sell when you see lots of red, and you will always win, right?

 

The quote should be altered to, "volume always leads price, you just don't know which direction price will go." Sure, volume is required to move price. A transaction must occur for the last traded price to update. But it's not that simple. This is basic stuff here, but in the first attached chart you will see first an example of heavy selling (that's the delta on the volume at the bottom) led to an immediate move up, and second where heavy buying led to an immediate move down. In fact, these were great trade opportunities. The other chart shows just one example in many that occur every day where selling is heavier over a period of time (in this case 1 minute), yet price was higher at the end of that period than at the beginning of that period.

 

If someone reading this wants really good and logical volume information, see Wyckoff. He was the man, and still is.

[carltonp 10]

Josh,

 

What trading platform do you use?

[rwk 10]

... is it possible to determine if a transaction has been executed at bid or ask, realtime?

I'm a little late to this thread, but I thought I would chime in. I am pretty much where carltonp is, that I am attempting to use automation to read and act on order flow. I was a professional programmer before becoming a full time trader. I am well aware, as joshdance points out, that I am competing with the smartest and best-financed traders in the world at their game. I also have many years' trading experience.

 

Thanks to gosu for pointing out that "GLOBEX allows orders to be shown that do not reflect the actual size being bid or offered; e.g., a bid of 1 contract at a certain price could actually be 1000 contracts.". I knew that was common in stocks, but I had not heard that about GLOBEX.

 

I have heard that GLOBEX is now aggregating trades. I'm not sure if ECBOT is doing that too, but I would not be surprised if they are or will be soon. The problem is bandwidth and transaction volumes.

 

That's also something I wonder if carltonp is aware of. While IB's data is aggregated, it has less lag, especially during busy times, than more complete data feeds. You can have completeness or you can have timeliness ... take your pick. I use IB's data. There are some things about IB's data that don't seem to be widely known. For example, we can now get live volume together with price in the same update, along with a timestamp. Pair that with DOM, and we have a pretty complete and timely picture of what is happening at a rate of well over 100 updates per second.

 

By way of clarification, we need two things to be able to trade off order flow: the limit order book (which shows bid/ask but not trades) and the tape (which shows trades, not bid/ask). This thread got to be pretty confusing about that.

 

Regarding carltonp's question above, it comes down to identifying whether a trade removed size from the bid or ask. I believe you can guess, and you will be right most of the time. But what I really wonder is: How useful is that information?

 

[rwk]

[joshdance 10]

Josh,

 

What trading platform do you use?

 

Investor RT for charting, and NT for order routing to my broker (using the DOM).

[gosu 10]

Josh, I think I've been bit of a numpty (British term for what Americans call a Putz).

 

Do you remember when Gosu said 'That policy would seem to make suspect any trade decision based on what's showing on the bid and ask at any given time regardless of the flawlessness of the data itself.'

 

What is exactly what I was doing, when it actual fact I thought I was doint what you thought I was doing when you said I think carlton is talking about transactions that have been executed, not orders in the book.

 

It's still possible to write a formula for that, just going to require a little more thought....

 

Thanks mate....

 

It seems to me that you are looking to design some type of a poor man's HFT algorithm to trade a retail account. My experience is that the "micro" level of the market is not the optimum place for a retail trader to operate. There is so much gamesmanship at that level that you really need to secure your bearings as to the right side of the market from less transitory elements and then zoom in when an action is required.

 

Indeed, once you know the right side, you will find that the best action points are fading the "fake outs" on the micro level and not reaching for prices after the move is underway. By "best" action point I mean an entry that can be "washed" easily if the need arises and an exit that extracts the bulk of the move--after the "swoosh," if you will.

 

You will be operating on another level when you no longer wait for "confirmation" but hit the transmit button to fade the micro spike on the stall because you KNOW that is the right side. Capturing the spread is just icing on the cake.

[Jleon 10]

The T/S and DOM is a crazy place that is for sure. To be able to see Iceberg and other institutional orders is a big help with any trading method.

 

I have created Software that works with Ninja Trader and plots these orders on a chart in real time. I have been a trader for 11 years and have always looked at the order flow on the DOM and T/S but a lot of orders have become way to fast for the human eye or are hidden in tiny lots. I have examples and screen shots to help explain. This will help anyone that is trying to understand the market by tape reading.

 

My website is Tradewithvolume

 

Have a great day!

[bobcollett 10]

Hi cabbie

Are you also a partner in this business, or just support?

Regards

bobc

[eros 10]

Hi guys,

 

I find this topic quite interesting and I should say I am one of the programmers who use Excel to automate trading decisions.

 

My excel macro simply tries to do things described here, and from this thread I got an idea as to what an always-winning (or mostly-winning) algrorithm should be.

 

Is this thread dead already before sharing my observations and experince with you guys?

 

I also wonder what developments have happended over the past 10 months.

 

I will update my excel function to create a signal after 20 seconds of recording HOD and LOD and crossing any one of them with enough volume increase within the next 10 seconds and will share the results shortly. This is what I come up with after all reading the discussion here.

 

Does it sound a good algorithm for you professional traders?

[eros 10]

Carltonp,

 

Where are you at the moment, I mean your progress on the matter, that is developing an Excel function using Tape Reading technique?

 

I don't see my posts listed here, but it might be a matter of time...

 

I will be checking for any response...

[MightyMouse 12]

Eros,

 

Last I knew, carltonp met a very nice girl, had a child, and is now running a daycare center for toddlers with his new bride. I don't think he will respond to your inquiries.

[eros 10]

Thanks @MightyMouse,

 

It does not look good for the infamous always-winning strategy then. Does it? Perhaps there is nothing such a thing...

 

What Carltonp has done is exactly what I should do, in fact. At least my mom thinks so. I got stucked with all this algorithm stuff and set apart my social, personal and professional life for over the past 1.5 years. And yet not so much success in stock markets. I feel like I am on the wrong course.

 

In this respect, well done Carltonp!

[invb 10]

Hi guys, I don't know if this thread is still active. I'm just wondering does anybody in this forum trade stocks with time & sale like Steve Kohen does?

[invb 10]

Does anyone in this forum trade stocks based on the Time & Sales?