Thoughts on Forex Volume

[GammaJammer 10]

Part of the problem is that there is no way to determine how much size was placed in (or taken out) of the market to move price those ticks. On GJ's platform, absolutely nothing went through yet price moved. Yes, somewhere, size was put into the system, but because forex is not centralized, there is no way for anyone to tell how much size went through..

 

No, not neccessarily. If you study how price behaves across ALL the ecns, you realise that in many cases EBS is still the driver, and with the way the prices were flying about my best guess is that what little did go through on the other ecns was more likely to be latency arbitrage (and therefore really just a bit of noise in terms of overall market direction) than it was genuine interest trading. People had pulled a lot of their interest out at that point waiting for it to calm down a bit.

 

THAT's what I'm talking about in a nutshell.

 

GJ

[GammaJammer 10]

Are you sure

 

Oh stop

[wasp 10]

Oh stop

 

Sorry, one of these days/moods!

[cowpip 10]

Are you sure

 

:hmmmm:.... ummm.... :hmmmm:....

 

:did I say that?:

[cowpip 10]

No, not neccessarily. If you study how price behaves across ALL the ecns, you realise that in many cases EBS is still the driver, and with the way the prices were flying about my best guess is that what little did go through on the other ecns was more likely to be latency arbitrage (and therefore really just a bit of noise in terms of overall market direction) than it was genuine interest trading. People had pulled a lot of their interest out at that point waiting for it to calm down a bit.

 

THAT's what I'm talking about in a nutshell.

 

Cool. So I've learned something new today.

[Guest forsearch]

No Walter - it isn't. It might work for you to trade this way but be assured - the statement you made here in reation to FX and tick volume is utterly false.

 

EUR/CHF trading on EBS right after the oil data was a perfect perfect perfect example of what I'm saying.

 

Price was 1.6323/4. It went 23 Given then 22/23 21/22 20/21 19/20 without a bean going through (my screen shows bid, offer, last given, last paid etc etc). Finally someone paid the 20 offer (for smalls, which was all that was there). So at that point you had a 20/21 market where last paid was 20 but last given was 23!!!!!!!!

 

In between those two actual trades you had had lots of price updates that would have registered on your tick volume, but that didn't equate to a single euro of business going through.

 

This kind of thing is why I get frustrated. I try again and again to explain how stuff works, and people just hear what they want to hear.

 

GJ has it right. All those "tick volumes" that you see is just someone adjusting the bid or ask or both. No trades have to occur for this to happen. Think of it like a bookie adjusting the betting line to get the action going.

[BlowFish 10]

 

In between those two actual trades you had had lots of price updates that would have registered on your tick volume, but that didn't equate to a single euro of business going through.

 

This kind of thing is why I get frustrated. I try again and again to explain how stuff works, and people just hear what they want to hear.

 

OK please dont get frustrated with me as you know I am a dunce when it comes to FX. (some would say that its not just FX but that's another story)

 

I do wonder if it is partly a matter of terminology. With futures (for example) you would only count actual trades as ticks not 'order book' updates or changes in best bid best ask etc. If you can't differentiate a trade from say someone pulling a limit order you are p*ssing in the wind. (I guess that's your point put rather more vulgarly) In fact if you are counting order book changes that are not the result of actual trades in any way shape or form you are going to be way way off base, however a futures trader would not refer to these as ticks.

 

Futures seem so straightforward sometimes FX and those that trade it seem quite different sometimes

[GammaJammer 10]

 

Futures seem so straightforward sometimes FX and those that trade it seem quite different sometimes

 

Hey - I'm very straightforward

 

GJ

[Flojomojo 10]

Hi,

thank you all for all this great input! I think its great for a new guy like me to discuss such stuff with the head of trading desk and other professionals!

 

So far my conclusion is: The assumption I made in my very first post (see picture) is wrong due to the way the OTC market and its microstructure works.

Since price changes in FX don’t equate with the classical view of volume I would put it in the term I read somewhere: The price “feels” its way up and down…by constantly adjusting Bid/Ask prices the market searches for the available liquidity pools.

Please correct me if I’m wrong.

 

GJ’s last post at T2W is: “Not saying they (tick volume and actual volume) are never positively correlated, just that they will just as often probably be either negatively correlated or uncorrelated.”

So any ideas out there how to determine from the market action which of the three cases applies?

 

Since this thread has attracted both the “pro tick volume” and “contra tick volume” crowd, I would like to push this thread into a new direction of insight:

 

OK…so the term “volume” is bogus in FX. Bid/Ask prices change due to transactions taking place or simply by price adjustments. Nevertheless this “tick volume”, how its usually called, seems to work for some people. Why is this the case?

Let’s drill a bit deeper and see whether we can find a new terminology that suits the pro and contra crowd!

Here’s my first go at it:

Let me pick up the example by GJ and assume it happens in an uptrend:

1.6323/4

22/23

21/22

20/21

19/20 …here someone hit the offer at 20. This is the only transaction between 24 and 20

20/21

 

Scenario1: Price falls straight from 24 to 20 then moving up to 21. Tick volume recorded = 10 (8 on the downside and 2 on the upside). Conclusion: The Composite Men (Wyckoff’s name to the total sum of more informed forces that move the market. Akin to “The market,” or “They” in other parlance.) are (a) bearish or (b) let price drift lower in order to buy at a cheaper price or © are at the sideline. If (a) is true GJ would have seen more action on the bid side. Therefore (b) or © are the case. Possible VSA equivalent: Low volume down bar with little “tick volume” in an uptrend. => probably (b)

 

Scenario 2: Price falls to 20 but wildly fluctuates between 24 and 21. Tick volume recorded is, lets say 100. Conclusion: CM are unsure if liquidity is to be found below 24, but consensus slowly favors the downside. Conclusion: CM are unsure, but let price drift lower. Possible VSA equivalent: High “tick volume” down bar in an uptrend but no real result. => “squat bar” or test in the making

 

So can “tick volume” be seen as more of a “sentiment indicator” of how market participants react to different price levels?

 

Curious about your opinions,

Flojomojo

[GammaJammer 10]

My opinion is no, it isn't necessarily due to market sentiment at all It's just noise. For a start it's a chart feed and could come from anywhere. Secondly, even if it's taking its price from a half decent source, what happens behind that is that such a source will likely comprise contributions from many banks (it won't likely be ecn feeds as most of them dont publish this data externally in real time yet).

 

Thus all these contributor banks will usually be updating their price every few seconds on a 'heartbeat' basis and sending them to Reuters etc to publish. These are indicative quotes only, NOT dealable prices (which aren't simply streamed out to the ether in that way).

 

As the banks like to see their names on Reuters / Bloomberg contributor pages regularly, they're gonna update prices regularly. So you might get 3 updates in 10 seconds from each bank say, with the result that even assuming only 10 banks contribute to the feed, it's gonna be a very busy feed.

 

And these contributions are subject to a few certain rles. Chief of which is that they AREN'T alllowed to, for example, take the EBS Best bid and offer, re-brand them as the Citibank bid / offer (just an example) and send it out to Reuters as a contributor price. They would leave themselves open to litigation by an ecn like EBS if they did that (as they're not meant to be simply white labelling data that EBS make a lot money from selling).

 

But as it's only an indication price, it can be a bit wider. Doesn't have to be one point wide in eurusd.

 

So maybe EBS is 50/51, so they have a model that takes this in, and pumps out a price a pip around it, so they're 49/52. Now it's their own price, so that's fine.

 

Say someone else has a different model, one that always shades one side only. So their price might be 50/52 for example.

 

A third bank might have one that's one point wide, but always a tick higher than the EBS best bid say, so they are 51/52

 

If they all send in price updates to this mythical price feed vendor (someone whose data feeds your charts) the price might go 49/52, 50/52, 51/52

 

To the uninitiated it might look like buying interest is stacking up, but in reality all that happened is that the contributions lined up in a certain way.

 

And if the feed driving the chart is 'cleaned' to keep for example a constant spread or whatever, these underlying updates would all still probably require a new tick update in the chart.

 

Make sense? I'm a trader, not a data vendor, so any finer detail than this is straying away from my area of expertise a bit, but I just feel that a lot of people are just purely guessing in this thread, and not from a position of being able to make their guesses educated. Unless you have actually been involved in this stuff (and I have) then it's tough to have the relevant knowledge. Which is cool, but people just can't stop themselves from trying to guess at the answer anyway, and imho missing the point.

 

Hope this clears things up a bit

 

GJ

[wasp 10]

End of the day, surely, the simple fact that, as there is no exchange, means no one chart provider will have all the figures, whether its IG Index or Reuters.

 

The only way you can have a true representation of volume is if all trades go through an exchange (ie futures). Unless this is the case, you are never going to get a full representation of volume and thus, its all b*******.

 

 

EDIT: GJ beat me to it!

[GammaJammer 10]

Even that would be fine IF one chart provider's data was even a reasonable proxy for the market. But I'm trying to demonstrate why even that just isn't so. This price update stuff is a total total red herring. end of.

[wasp 10]

Even that would be fine IF one chart provider's data was even a reasonable proxy for the market. But I'm trying to demonstrate why even that just isn't so. This price update stuff is a total total red herring. end of.

 

Precisely. Unless there is an exchange, no one chart provider could provide these stats as there is no central body taking every single piece of information. I think we are just confirming each other here so I'll sod off!

[GammaJammer 10]

Plus everyone else has stopped joining in and it just seems to be you and me now.

[Kiwi 16]

I'll add to it then.

 

GammaJammer is correct. All forms of Volume and Tick counts on fx are bogus. They can't be anything else. Why? Because they are not a measure of volume and they are not a measure of transaction counts.

 

If you wanted to call it Picks or Pricks or then you have a valid measure (PriceIcs instead of Transaction Ics because all they really show is that the average of the bid and ask price changed). But. And its a very big But. The Pricks are only valid for your data feed. And they are exceptionally easy to manipulate because no actual transaction is required.

 

So, better to accept the uncertainty and just use time based charting.

[Flojomojo 10]

Hi GJ,

did you ever attempt a study about how many price changes have been trades and how many just price adjustments?

[cowpip 10]

Thanks for your insight, GJ. Those of us with lesser knowledge appreciate it. I knew "tick volume" was a bogus concept in forex, but I didn't understand it as deeply as you have explained. Even when you've been in this gig this for years, there's still a mountain more you can learn. That's what I love about this.

[GammaJammer 10]

Hi GJ,

did you ever attempt a study about how many price changes have been trades and how many just price adjustments?

 

Yes I knocked a quick Phd thesis out on that very subject before I had my cup of tea and bowl of shreddies this morning.....

[wasp 10]

Yes I knocked a quick Phd thesis out on that very subject before I had my cup of tea and bowl of shreddies this morning.....

 

I thought you sat around talking shop all day and not actually trading much as it is (middle management bankers! pah!) so surely this should be a walk in the park for a range bound Thursday like today!

[GammaJammer 10]

My god - it's like you have a webcam pointing at my desk.

[wasp 10]

My god - it's like you have a webcam pointing at my desk.

 

Nice tie btw! :o

[GammaJammer 10]

It's not a tie at all, merely a very creative stain from my tomato soup, cunningly spilled down my front in the shape of a tie. Getting the windsor knot at the top to look right was a beeatch.

[Flojomojo 10]

Yes I knocked a quick Phd thesis out on that very subject before I had my cup of tea and bowl of shreddies this morning.....

 

:o well...it was worth a try! ...and if you need a Phd candidate to resolve this question...I might actually have the time soon...

[GammaJammer 10]

:o well...it was worth a try! ...and if you need a Phd candidate to resolve this question...I might actually have the time soon...

 

ICAP (owners of EBS) actually have a research department working on stuff like this. They do seminars on it. Don't always agree with their findings (a few of the assumptions are imho deeply flawed) but it's very enlightening stuff nevertheless.

 

GJ

[Flojomojo 10]

Let me summarize this thread in a couple of sentences:

 

- Due to the OTC microstructure of the FX market, obtaining detailed volume figures is impossible.

- Tick-"volume" as it is called only gives a snapshot of the overall activity going on.

- Tick-"volume" is a misleading expression. Since here the number of price changes are counted and the FX market has a constant stream of repricing quotes, the figure doesn't necessarily provide a volume indication.

- Tick changes and volume might me uncorrelated or correlated sometime...but to this point nobody in this forum can or wants to tell whether or when a correlation exists. Theres also no study about how the flow of price quotes/ticks gives indications about the underlying supply/demand situation.

- Usage of tick-volume has no proof yet (that I am aware of) linking ticks to volume. Scientifically speaking it yet belongs to the mumbo jumbo section. Nevertheless some traders use it and are doing well!

 

Have a nice day,

Flojomojo