Discussions on Order Flow/Volume

Trader Tim...I greatly appreciate your comments and kind words. I truly hope that this thread as contributed something of value to you and wish the best of luck in your trading.

I am always available to answer your questions either here on this thread or via pm or email. As far as the software programming goes, I'll let you judge the benefits or risks of publication of your hard work.

Thank you again

Order Flow Theory and Trading Tools (Indicators)

I have recently received a number of Private Message(PM) enquiries from people looking for a delta volume indicator for the MetaTrader platform which then leads onto questions and discussions about a whole lot of other related things. I hope this post will answer most of the questions that I keep getting asked.

From the very beginning of Don Pato’s “Discussions On Order Flow/Volume”, thread he says that the point of the thread is to discuss the basis of market interactions from a structural point of view.

As I see it, it is essentially a theory of how the Forex market place works based on the indisputable fact that every participant in the Forex interbank market does so though market or limit orders placed into the market mostly through the EBS interbank network. (Some currencies use Reuters and other minor networks). You can’t get access to the EBS system. The company that produces the EBS software is owned by the major commercial banks. Access to this platform is closely guarded but even if you could get access to it you would still have to interpret what the data means. It shows orders in the market, not peoples ideas and motivations for placing these orders.

I have been looking at the use of volume in Forex trading for a number of years. Nothing I have seen or experimented with including Market/Volume Profile, Volume Spread Analysis and various other volume based approaches to trading came even close to explaining my questions about market structure and volume. There is quite a lot of information on the FF forum and elsewhere which appears to me to have been put together from bits and pieces of old systems, with added jargon to try and create a good story. A number of these stories seem to have been spun off to commercial sites who are trying to sell custom indicators for the MetaTrader platforms to retail traders.

Don Pato’s thesis is the only thing I have come across so far that makes coherent sense. I am not saying that this is the only valid information available. It is just the only source that I have come across to date. Don also says quite clearly on a number of occasions that that the information presented on this thread is not a strategy or a trading system. He also refuses to be drawn into discussions about trading platforms and indicators etc saying it is irrelevant to the principle that he is trying to present.

To me there are a number of steps involved in making use of his theory,
• Step 1. Understand the theoretical structure of Order Flow/Volume.
• Step 2. Figure out how you are going to use this model to generate trade ideas.
• Step 3. Design and test your strategy or system based on your ideas in step 2.
• Step 4. Source or construct the tools you need (if any) to execute you strategy.

Looking at tools (i.e. indicators etc) before you have a firm understanding of step 1 is to my mind putting the cart well before the horse. Don constantly advises people who ask these sort of questions to please read the whole thread from the beginning. I couldn’t agree more.

Since we cannot get access to the EBS platform and/or data we try to simulate it using what I call proxies. There are two options commonly in use. One is Forex futures data which is usually sourced from the Chicago Mercantile Exchange (CME). The other is tick data which we get from our MetaTrader (or other) trading platforms.

How good are these proxies? It has been claimed elsewhere on the FF forum that CME data is nearly 100% accurate. Like just about everything else in the Forex world, my tests show this claim to be wrong. As for Forex tick data, there are many people on the FF forum running around claiming that Fx tick volume data is absolutely useless as it only has volume and does not have buy and sell order volume. These claims are also invalid. Buy and sell volume can quite easily be calculated using the bid and ask price, the spread and the tick volume. Tick volume, provided it comes from a centralised liquidity source has also been shown to have a very high correlation with that from the EBS system. Please see the following links to my previous posts which include further links to this research information.

https://www.forexfactory.com/thread/...2#post13585192

https://www.forexfactory.com/thread/...0#post13576760

Which data makes the best EBS proxy? Fx futures or tick data?

It is difficult to say at this stage since we cannot get access to the EBS system to check and compare. The data tests which I have done to date and posted on the forum (see above links) suggest that FX tick data may have an advantage. It appears that either will work but neither are anywhere near perfect or 100%.

There are a few problems with using tick volume on the MT4 or MT5 platform.

1. It is not stored anywhere on the MetaTrader platform so it is necessary to collect history tick data and store it somewhere or import it. Alternatively you can use one minute data or simulate tick data in the same way that MetaTrader does for their strategy tester. My tests to date are inconclusive as to which if any methods produce the best or reasonable results.
2. Delta Volume = Buy Volume – Sell Volume. This seems to be the basis for all order flow volume indicators. For tick data this has to be calculated. There are various possible algorithms for doing this. All will give different results, some dramatically so and no one as yet that I am aware of seems to have published information about the best or optimum solution. (I am working on it)
3. You have to ensure that your MetaTrader data feed comes from a major liquidity pool or it may not be reasonably representative of the EBS data. This means that you may have to switch brokers if the data feed they are supplying doesn’t measure up. One method for testing your data feed is included in the links mentioned above.
4. When I have some conclusive test results on MetaTrader tick data indicators I will post the full results of these tests, along with the algorithms, on this thread. Due to the above mentioned problems together with the significant amount of abuse that I have received from numerous people on this forum resulting from previous work that I submitted, I have decided not to personally release any MT4/MT5 indicators or other software on this forum. You will therefore need to program your own or hire a programmer. Due to the above mentioned problems, obtaining a satisfactory outcome may not be straightforward.

You can buy Fx futures data directly from CME. The cost is about $US110/ month. A cheaper option seems to be to buy or rent a trading platform with a combined data package. Don Pato seems to like CTrader. NinjaTrader is another often mentioned one but there are plenty more. At the moment I am using a combination of my own custom written software and a Russian MetaTrader based system called ClusterDelta. This is a cheaper solution but the documentation is in Russian or often difficult to read English. Some of their system still seems to be under development and has bugs. I had to write my own software to extract the CME Fx futures data as the software didn’t come with a data interface. I found this to be quite a complex task and I am an experienced programmer. Numerous people have reported that they are unable to get some of the indicators provided with this system to work. I have reported bugs to the developer which have not yet been fixed. It is probably safest to regard ClusterDelta as experimental at this stage.

Don Pato appears to be using both Fx Futures and tick data the same as I am with custom written indicators for his CTrader platform.

It is worth noting that a number of people on the FF forum appear to be receiving commissions or kick backs from suppliers selling commercial trading platforms like NinjaTrader packages with CME fx futures data. I have been approached with such offers.

Disclaimer – I have no commercial or other interests in anything mentioned in this post.

There are also a number of commercial MetaTrader indicator offerings that use tick data. The problem is that all of these are “black boxes”. You have no idea how they work or what they do and some of the information provided by their vendors is to say the least suspicious. Personally I am not willing to waste my time or money playing around with any of this stuff.

There is an MT4 delta indicator with source code readily available from a number of locations on this forum. In its present form it is difficult and impractical to use. Everyone I have referred it to has been unable to get it working. In my experience this software does work but I would regard it as experimental, i.e. not suitable in its present form for trading purposes. The coding is also of poor quality. In my opinion, rather than try to fix or complete this indicator, it would be quicker and easier to start again and write a new one from scratch.

Depending on your strategy, extra data and indicators may not even be necessary. I have been thinking and trading along these lines for years just using price action. When looking for a trade idea I simply ask myself the question “What are the big guys thinking and planning to do and when are they likely to do it”? Not even the EBS system is going to tell you that but it may help you figure it out as may these other tools that we are currently looking at and working on. What Don’s structural model does for me is to update my internal mental model with something that is inherently simple and actually makes logical sense. This in turns provides me with a better basis for developing my trade ideas.

I therefore highly recommend Don Pato’s model as a solid structural foundation on which to base your thinking about the forex market, your trading strategy and how best to trade it.

At this stage I am only using these data feeds for research purposes in R/RStudio. The purpose of this research is to try and discover patterns relating to big banks and their market makers if any exist based on Don’s structural model. Further down the track this may result in new tools but I am not in a position to even think about things like that at this stage until I know what it is I want to do and have figured out a way to go about doing it.

Regards

Tim

just skimmed this thread, sorry if i've missed the obvious.

assuming you guys are talking about cummulative delta aka order flow and not some other order flow. why bother looking at the spot market and move a homogeneous market like the futures market? even if you trade the cash market, you can still get the pips with very little translation.

this is where i play all day long.

looking forward to the discussion.

We are talking about volume delta at each and every price level. This is to ascertain the prevailing dominance in the order flow and it's effect on price. Cumulative Delta is nothing more than another moving average that will always (by it's very nature,) lag behind the market.

The emphasis here in this thread is how to apply these concepts and structure to every market, as indeed it is the structure of every market down to your neighborhood supermarket. It is always your choice as to which market you wish to use. I do trade both futures (e-mini) and cash FX, as well as metals. The same structure exists in all markets....may I suggest a more in-depth reading of this thread to truly appreciate the value it contains... skimming leads to incomplete information.

Trader Tim...I truly appreciate your endorsement of what I have presented here. In fact you last post was so thorough as to leave with nothing else to add. I wish you the best of luck and hope I can answer any questions that come up. Thank you again for your kind words.

"Cumulative Delta is nothing more than another moving average that will always (by it's very nature,) lag behind the market."

well to each his own but it lags very little for me and is way way more then just another ma. my guess you don't use it correctly. i doubt it can be used on mt4/5 and really only comes into play in an complete market like futures (stocks etc) and not fx spot. i have traded spot via the futures market, even with a delayed feed. i wouldn't trade without it.

i'll have a look but if folk here think CD is just another ma then i'lll probably wasting my time. shame as not enough people are using CD or using it correctly.

thanks for your response.

i'll add that maybe there are different CD variants out there but my understanding sierracharts and ninja trader have it.

Below is a log entry from back in March of this year from a futures trade. These are examples of documentation of my trades...but also illustrates the advantage of looking at Volume Delta at each price level. The comments on this chart are self explanatory. There is a "Cumulative Delta" running on the left chart (450 ticks) that I designed and used to use (almost) exclusively, UNTIL, I learned to read the order flow at each price level. Note how I was able to distinguish between two different ranging conditions...one a continuation...the other a reversal. And able to get very FAST limits with very little risk.

The first trade (according to the cumulative delta) showed divergence and "should have" been a reversal...but wasn't. The second trade (according to cumulative delta) showed growing buying , which should've indicated a continuation...but didn't. BUT

By watching the order flow at each price level I was able to clearly "see" the liquidity supporting the bottom of the range in the first trade...and also the liquidity that was absorbing the buying order flow at the high of the day. This allowed two quick and seemingly "flawless" trades...but what it was really just being alert to the changing conditions and NOT relying on the indicator but "seeing" the order flow in real time, and how price was responding to it...again in real time.

Hi DonPato!

I did not understand how to use cumulative delta because of that reason. But just delta is extremely good. As you pointed out there are no direct dependency if there are more buy trades open buy and so on. I have tested a lot but there is no positive expectation, it looks like random *max 47% as everything). That's why I have decided to take it "as is". Huge delta = imbalance = great movements (and always big volume). In this case I'm trading reversals or trend, and reverse the trade (hedge) if something goes wrong.
May be I'm wrong with that but I don't have orderflow in mt5, so may be if add it to cumulative data there will be some sense...

Best of luck... again I don't use a "system" but my approach is "methodical". When it looks like the buying is exhausting, I look for selling opportunities...and conversely when it looks like the selling is exhausting I look for buying opportunities. It really doesn't need to be any more complicated than that

Absolutely agree. Very good approach!

Could you please tell how you determine exhaustion? What is your definition? And what you do if you trade goes wrong? Hedge/reverse? Close by SL? Thanks!

I refer you back to the beginning of this thread in which I explained in detail the concept of "exhaustion". Here are a few of these posts: #91; #102; #104; #112...there are more, but I don't have time to go back through them all...But you can do that when you have the time. Frankly, my entire purpose for learning and understanding order flow was precisely so I could "see" the waning of the dominant side of the market to get ready for the other side to take over. This almost always PRECEEDS the price movement and is indeed what causes it. There is no "technical signal", it is just learning to read the market.

I loose...nothing is perfect...as much as I would like it to be. There are days, when I just can seem to do anything right, and rather than continue misreading the market I just stop trading (no matter how upset I am at losing) and do something else. Biggest point here is...learn when NOT to trade. If there is something (usually in my attitude or my mental state) that is interfering with my ability to read the market, it is just better if I don't trade. Thus I have a 3 loss limit...If I lose 3 trades...I stop trading...something is wrong.

Does it really matter? A loss is a loss. Stops are there to keep a loss from becoming catastrophic. Hedging is just another way of saying "I'm not sure what I see." So If I'm not sure...then I don't trade. I do sometimes Stop and Reverse...but that is often a cascading loss scenario when price gets "whippy". Better to just do something else.

On the subject of hedging...I do engage in hedges in the spot market, under the existing criteria.

1. I am currently in a winning trade with a stop that is locking in significant profits (greater than 1xR)
2. I see some exhaustion occurring in my direction but nothing that is a clear reversal. This is usually due to just simple order flow "pooping out"...think the day after a big news announcement and everyone has done what they're going to do the day before...LOW VOLUME DAYS
3. There is enough profit in the current trade to cover the potential loss of the anticipated hedge (if it fails), AND still have some profit.

This being said, I don't do it very often and it usually results in just smaller profits...

I have practised hedging for years and find now is the goden period and a time to exercise extreme patience. Largely because of ATH happening with a frequency never seen before.

I always travel with 50 pips profit locked upwards towards known ATH leves, add a factor or multiplier to it to have a target peak price to liquidate longs and wait for the shorts to go down like ski skating . It is worth a few days patience . I won't say it is a superior trading method. It just suits my funful temperment for trading equity indices only.

I have no idea whether it will work with ATHs of stocks or other instruments where ATHs are hit with a known 1 to 3 weeks frequency .Forex ATHs cant be traded this way which requires existencev of a known ATH as a constant value.

I am a Big picture man and use volume in conjunction with ADR extremes to trade in a relaxed way. Micro trading of volume is also profitable but does not suit my 80:20 Pareto mentality.

In both cases guessing is involved. Take your pick from Bands, boxes, channels, envelopes to throw your dice though I have not seen even a single volume indicator using any of them.

Thanks, I'll assume this is for me.

"the other a reversal. And able to get very FAST limits with very little risk."

this is something I'd like to see more of. Again I have not read the thread and will probably tomorrow.

Here is quick synopsis of how i use CD. CD is a measure of limit order vs market orders (at least that is my version of CD). If we know that institutions rarely use market orders (again depends on the market) and retail almost always do market orders, it is possible discern institutional from retail. as we know retail does not move the market(again depends on the market). my main focus is to understand what the institutions are doing.

here is a copper trade i took on wednesday (based on CD) and closed on friday using just volume profile. i'm looking for disortions between price action and CD, as shown (second from the top is CD). i'm only really interested in large distrotions and move from market to market looking for them. this trade was posted on the copper thread.

i could go on but that's my process in a nutshell.

Just read the thread from start to finish and loved the info, Regarding manipulation

A european company has 2 trillion EUROS to be exchanged for USD to start their next project in the US which has a time limit but the bank cannot supply enough USD to fulfill the order, where does this bank get the extra USD?

Hi Don

One of the things I study every night on my Motive Wave order flow charts is the battle between passive limit orders and aggressive market orders.
I see two conditions which I believe need to be traded in two separate methods.

Without doubt, buyers don’t create bullish markets and sellers don’t create bearish markets because that’s not what allows the market to move. The passive limit orders sitting on a chart I believe are only to allow movement in one direction to encourage a reaction from traders. I see the real direction is caused by the failure of market makers and bankers to absorb high volume market orders.

I don’t agree with the idea the bankers and market makers are always in control of the markets as breakouts are too expensive to absorb. If they choose to halt direction of the markets, the big players will increase market density to absorb price, therefore repelling price backwards in the opposite direction. But I believe we have a problem here. The breakout traders have more volume on breakout than the liquidity the big players have to absorb price and change direction.
Therefore in my eyes, agressive market orders are stronger than passive limit orders in HIGH volume periods, weaker in LOW volume periods.

This is a condition I trade for every night.
Looking at EURUSD, the normal volume on a 1 minute bar is around 100 units, when we see volume increase bar by bar, we know either the smaller players are crushing the big players or the big players are running stops or allowing price to move by failure to absorb price. It’s absorption or lack of absorption which is the key to allowing the market to move.
This leads to two strategies which are very easy for me to trade.
1.. If the liquidity is showing high Sellers numbers on the DOM, the delta is strongly bearish, we see strong BID imbalance AND price is moving bearish with purpose-

SELL till the delta flips strongly bullish. The reason is it’s an uncontainable breakout.

2.. If the liquidity is showing sights of weakens in selling numbers on the DOM, the delta has become extreme, I see very high imbalance BID imbalance numbers with extreme volume AND then price stalls-

BUY when the BID imbalance flips to ASK imbalance, ensure volume is collapsing.

That is my strategy in a nutshell.
Just my opinion and please excuse my input here.

Cheers Rick

Thanks Rick,
Well done, the way it has been and will continue to be.
Enjoying your weekend,
Don

Do you use your intuition in identifying distortions?
or use any volume indicators as such?
I like your relaxed method of trading.
Any links to the volume indicators will be appreciated.
Thank you

Hi simnz,

intuition: no not really, there are exceptions (like most tools) but if price remains fix in tight range and OF\CD is moving in a clear direction, i or anyone can understand what is going on.

or any: OF/CD is my primary indicator for bias. then i move onto more retail tooling for some confirmation on what is going on. or whatever da boyz are working to.

links: as mentioned i use sierracharts with cummulative delta. they do 2 week trials (from memory)

ok i do have to read the thread.....