FX realised Vol by using BBG and Reuters' data

zxem · November 27th, 2012, 5:26 pm

Hi all,I have tried to use BBG and Reuters FX spot tick data to calculate the realised Vol. For the time period, the Reuters' data do present the negative autocorrelation problem which mentioned in many papers about high frequent data. While BBG's data is quite clean, my current explanation is the spot source I used is BBG BGN. BGN is BBG cleaned data source. Could anyone share its experts in this area pls? Tks

tags · November 27th, 2012, 8:40 pm

i didn't know BBG has say q-trading extensions.

zxem · November 28th, 2012, 2:29 pm

Sorry, what do you mean?

gpop · January 4th, 2013, 3:03 pm

What about using central bank FX fixings ?Eg. the NY Fed or ECB ones. As they are supposed to be observed every day at the same time and from the same sources and filtered by the same method, these time series should be a good input for your statistical analysis.They are available in Bloomberg under specific tickers (such as EUCFUSD Index for EURUSD from ECB and USCFEURO Index for EURUSD from NY Fed), on some pages in Reuters (ECB37), or on the web (you may download some histories from there):www.federalreserve.gov/releases/h10/His ... ex.en.html

zxem · January 4th, 2013, 4:52 pm

Thanks for the reply. What I am calcuating is the intraday vol with tick spots.

yugmorf2 · January 5th, 2013, 2:14 am

how different are you vol estimates using the two sources and what bar frequency are you looking at? for anything >=1min bars, there shouldn't be too much difference, and if so, then first point to check might be that you are using the correct time stamps (eg price at close of interval, UTC time)