|
Ok Habib - well you may have heard of fixings but you clearly haven't had them explained to you as they relate to FX, so here goes.
Background:
Many Investment managers and corporate treasuries are increasingly attempting to quantify the performance of their dealing operations (as well as that of the banks that they trade through). For a proprietary trader it is of course pretty simple - P+L high = good, P+L low (or negative) = bad.
But how do you rate the performance of a dealer for a fund manager. Typically the dealers are largely operating on an execution basis, buying JPY to cover their purchases of Japanese stocks for example. They aren't (for the most part) buying JPY because they expect the JPY to appreciate, but merely because they have a payment to make. They will of course take and express a view on currency movements, but probably 80% or more of the trading they do is cash management and hedging.
So you can't really look at P+L in the same way. Equally, a large fund simply isnt going to say "what the hell, there's numbers out this afternoon - hopefully we can pick the JPY up cheaper later". They are simply much too conservative for this (and rightly so). They specialise mostly in stock picking. So increasingly many of them are using fix trades to guarantee some sort of uniform execution.
So how does this work?
OK, So say XYZ pension fund needs to sell USD 100 Million vs JPY to cover their share purchases for the day in Japan. They could call a bank and ask for a price there and then, they could call a bank and leave an order, or they could leave a fix order.
Assuming they do the latter, they will call whatever bank they use sometime in the early morning and ask them to sell USD 100Mio for say the Frankfurt Fix. This means that they will sell their Dollars to the bank in question at whatever rate it fixes at at midday in Frankfurt. That price is determined by tick data around the exact time when the fix takes place. So if at midday USD/JPY on EBS is 110.35.37 it will very likely fix at either 110.35, 36 or 37.
From the bank's point of view it's now bought USD/JPY at the fix rate. However what usually happens is that at 09:30 (or whatever time the fund manager calls), the bank starts selling (maybe 10 million to start with). They will be selling all morning and trying to make sure that when midday comes around the are short USD 100M and the price is now at the low of the day. So Say USD/JPY was at 111.27 when they were called, and it fixes at 110.36. Maybe the bank sold a few dollars initially, more all morning and the last 40M or so in the last 10 minutes to keep it low. Maybe they sold at an average of 110.76 for example. They have therefore made 40 points on USD 100Million. That's approximately USD 362,450 profit for a morning's work. Not bad.
Bizarrely, increasingly, corprate treasuries and funds are using this as a way of measuring slippage on trading. It's never made any sense to me and seems to be purely driven by the fact that they can explain this method of execution easily to trustees etc. As far as I can see the banks are the winners on these trades 99% of the time.
The main point I'm trying to make is that this stuff really happens, and I'm pretty sure that that's what's behind the morning move you were talking about, rather than anything to do with what they thought was gonna happen when the numbers came out.
FYI there are 3 main interbank fixes - Frankfurt (midday London time), ECB (13:15 London time) and a third fix at 16:00 London time. Strange moves 2 or 3 minutes before and after these times are more than likely fix related. Hope this explains it a bit - sorry it's long winded.
GJ
|