VWAP On Forex

[Zunaa]

Below I Have Included some Screen shots of My Tradestation VWAP_syn indicator that will work on FORX or any Market that does not report Volume. Yet it can be used on Markets That do report volume as well.

Is anyone using VWAP to Trade Forex?



Regards,
zunaa

[Trader333]

In view of the fact that vwap needs volume to work at all, how have you gone about substituting it ?


Paul

[Zunaa]

Originally Posted by Trader333 In view of the fact that vwap needs volume to work at all, how have you gone about substituting it ? Paul

What you state regarding VWAP on the surface is true, after all even the name of it iplies Volume.
That said many, Volume infact is not needed to get an indicator to plot same as what a Traditional VWAP Does, Regardless if volume is or is not available. However continuing to Think that volume is needed, will keep one from finding or even thinking the possibility exists.

Last of all, what I mentioned above is supported by the posted screen shots and a video on VWAP_syn or synthentic Volume, so it is not confused as just an opinion.

The video is available here on TL under the video tab...
if you cant find it look in Forex or volume channel.

Rgards,
zunaa

[BlowFish]

Originally Posted by Trader333 In view of the fact that vwap needs volume to work at all, how have you gone about substituting it ? Paul

Guess it must use 'ticks' as a substitute. Certainly doesn't look bad but it is worth mentioning ES ticks are somewhat different to FX 'ticks'. It's probably alsow worth mentioning that the TS functions VWAP_H incorrectly calculates VWAP. You need to re weigh the whole sample when new volume 'arrives'. There is a discussion about this in the trading with market stats thread. I am no saying that this isn't useful btw .

[Zunaa]

Originally Posted by BlowFish Guess it must use 'ticks' as a substitute. Certainly doesn't look bad but it is worth mentioning ES ticks are somewhat different to FX 'ticks'.

A logical guess... but thats not it.
I made a comparison with ES only because many traders are familiar with ES.
It will work the same with any Instrument.... Stocks, Futures of any kind, Forex, simply anything you can plot on chart.

Bye the way... for anyone interested.
Looking at this riddle from a programmers point of view, one will never solve it... instead, look at it from a traders point of view, offcourse it is implied that one has enough trading experience.

[BlowFish]

Originally Posted by Zunaa A logical guess... but thats not it. I made a comparison with ES only because many traders are familiar with ES. It will work the same with any Instrument.... Stocks, Futures of any kind, Forex, simply anything you can plot on chart.

I edited my original post with something else that occurred to me about TS Vwap_h. If you like I can post an ES chart tommorow with my VWAP on. Might be interesting. Averaging across 5 minutes also introduces inacuarcaies, normally fairly inconsequential but last week i did have one day where my 5 minute chart several points off my 1 tick chart (on the DAX).

If you are not using tick's or bid ask volume changes then you have something quite novel there

[Zunaa]

Originally Posted by BlowFish I edited my original post with something else that occurred to me about TS Vwap_h. If you like I can post an ES chart tommorow with my VWAP on. Might be interesting. Averaging across 5 minutes also introduces inacuarcaies, normally fairly inconsequential but last week i did have one day where my 5 minute chart several points off my 1 tick chart (on the DAX).

Sure...
Re: comment on VWAP_h Function, yes I understand what your refereing to.

Originally Posted by BlowFish If you are not using tick's or bid ask volume changes then you have something quite novel there

------------------

[BlowFish]

Good job anyway! I am not trying to knock your hard work.

Actually if you are familiar with the error in VWAP_H you could probably fix your algorithm to mimic the actual Vwap. The thing is it is processor intensive (not over duly on 5 min bars) as every time you have a new piece of data you need to re-weigh all the previous pieces of data with the new total volume. But sounds like you knew that anyway

[Zunaa]

Originally Posted by BlowFish Good job anyway! I am not trying to knock your hard work. Actually if you are familiar with the error in VWAP_H you could probably fix your algorithm to mimic the actual Vwap. The thing is it is processor intensive (not over duly on 5 min bars) as every time you have a new piece of data you need to re-weigh all the previous pieces of data with the new total volume. But sounds like you knew that anyway

Thanks for your comments. I like to fool myself into believing that after a short conversation with a Trader I can tell at what level they are operating... For example the more defensive or bravado displayed -- is indicative of a trader engulfed with performance insecurity or lower level of trading competence.
On the other hand It is always a pleasure to converse with those (you) that display the opposite—it allows for all involved to have good time.

As you stated it is very PC intensive to calculate VWAP properly, but well worth it if your using smaller intervals.
What actually happened was that several Forex traders that I talk to were talking about the need for a VWAP indicator. It was easy for me because it was not something I really needed. Does it not seem to be like that with almost everything in life. The less attached you are to it the easier it comes to you.

In any case these Forex traders I mentioned are all using 5m or higher intervals. Most are using 15m 30m 60m and higher for their trading needs. This is why I didn’t think they need it the VWAP very precise. In contrast if you’re using strategies that are dependent on low intervals i.e. Entries/Risk is dependent on low intervals then there is value in having the precise VWAP calculation. Just my view point.

Regards,
zunaa

[BlowFish]

I still run a single tick chart with VWAP (just because I can I guess). I have larger bars too. Most of the time the difference is fairly negligible. There is a discussion in the thread I previously mentioned on what is 'good enough'. At the time I posted various charts comparing different algorithms. In the end with some maths help I have managed to come up with a 'continuous algorithm' that is also accurate.

I am interested in your FX approach. I am not sure how proprietary your method is but I'd be interested (broadly speaking) how you have achieved it. I wonder if time factors into it (a bit like MP uses time samples to build a profile). PM me if you think it is more appropriate, or ignore me if it is completely in-appropriate!

It might be quite interesting to emulate a more correct (than TS) VWAP and not too difficult if you have a proxy for volume. I did notice a slightly different shape to bands in my tests.

Not sure how closely you looked at the trading with marketstats threads, they really are pretty good. Between those and yur FX buddies maybe you might be persuaded to persue VWAP a little further.