277 reputation
312
bio website
location
age
visits member for 2 years, 10 months
seen Jan 31 at 7:26

Python and R


Apr
20
awarded  Necromancer
Mar
9
awarded  Nice Question
Feb
28
awarded  Notable Question
May
22
awarded  Nice Question
Apr
11
awarded  Popular Question
Jan
18
awarded  Popular Question
Dec
3
revised Time series of PCA - Sign change in factor loadings
deleted 8 characters in body
Dec
3
answered Time series of PCA - Sign change in factor loadings
Jun
3
awarded  Yearling
Sep
1
accepted How do I eliminate developed currency funding cross rate risk in an EMFX position?
Sep
1
comment How do I eliminate developed currency funding cross rate risk in an EMFX position?
Thanks - it's what I suspected. I'll look at Stock & Watson. I think the multiple regression route that I suspected is the way to go. I am going to have to look at the multicolinearity aspects though if I include a lot of other EMFX in there. Perhaps a PCA first?
Sep
1
comment How do I eliminate developed currency funding cross rate risk in an EMFX position?
Yep - not looking at the rates - could use IRS too of course. or fund bonds in FX forwards or repo. Looking for the pureplay on ZAR so your last para kind of answers the question.
Aug
22
comment How do I eliminate developed currency funding cross rate risk in an EMFX position?
That is wrong for the following reason. If I have a view on the South African rand, for whatever reason, I do not want to trade EURUSD. Yet there is a high correlation between USDZAR and EURUSD. Thus even though I am benchmarked in USD, trading USDZAR gives me EURUSD risk. Indeed USDZAR contains a whole bunch of other risks too, including SPX, bonds, BRICS etc etc. What I want to do is isolate out the ZAR specific risk, so that I can trade a pureplay on South African fundamentals or technicals. I am seeking a generalised technique to isolate a pureplay on a currency (or indeed on any series)
Aug
19
comment How do I eliminate developed currency funding cross rate risk in an EMFX position?
nothing wrong. Bloomberg has a correlation weighted set of indices (BCWI <go>} but they average the weights on each optimal basket - which to me doesn't make a huge amount of sense - takes away from the pureplay of each currency as we deviate from its optimal weights. I was just wondering at the time if there was something more interesting being done out there. I have gone with PCA bloomberg-style, without weight averaging.
Aug
16
accepted How to derive the implied probability distribution from B-S volatilities?
Aug
16
comment How to derive the implied probability distribution from B-S volatilities?
Thank you very much. Now off to program it......
Aug
12
comment How to derive the implied probability distribution from B-S volatilities?
I am indeed looking for the implied distribution as from there it's simple enough to do the qqplot. My first priority is an empirical approach, as the target audience of this visualization will want to use it to get and idea of which parts of the surface are cheap/dear, by visual comparison to historical performance of the pair. This will not (yet) be for a trading model. Of course I would like to be able to show this in an intuitive fashion (with all the usual caveats about historical v future returns, and low-probability event risk). Thank you for your interesting starting points.
Aug
11
awarded  Promoter
Aug
9
comment What is the role of Credit Valuation Adjustment (CVA) desks in investment banks?
I'll add one final point, and that's about culture. The "culture" of the front office is taking risk, wherease the culture of the CVA desk is inherently to reduce risk. If you're asking this question in relation to career choice, you need to keep this in mind, as the training you will receive and the instincts you may develop on markets, will obviously be affected.
Aug
9
awarded  Commentator