Statistics, Damned Statistics, and Surveys
I am not sure who reads FX blog posts on the night before Christmas but apparently I write them. The recent publication of the annual StreamBase survey entitled “FX Trading and Technology in 2012” has been well covered. I picked up my copy here. While the survey seemed comprehensive and highly professional I did not find much to take away from the survey even though the topics all interest me. I suppose the reason for this is that in my job I look for practical intelligence and the survey’s scope is probably more appropriate for market pundits. Where I did have particular interest I found the survey to be too broad and high level. The best illustration of this is the following statement:
About half of the respondents trade outright forwards, FX options, futures and swaps. Only some 29% trade NDFs. Outright forwards, swaps and NDFs are more commonly traded by sell-side firms than buy-side firms.
The question is what does “trade” mean for a sell side firm? Does it mean trade with customers (either buy side or other sell side firms) or trade into the market? My experience is that “most” sell side banks of any size almost never trade outrights in the market but almost all trade outrights with their customers. To these banks, outrights are much more a buy side instrument than a sell side instrument – exactly the opposite from the results of the survey! The reason this is true is that most sell side banks (even pretty small ones by “bank” standards) have separate swaps and spot traders and the customer outright trades are usually split into spot and points components to cover in the market.
My conclusion is that either the survey is more heavily weighted to very small sell side firms (banks? Brokers? We don’t know) or the survey question about outrights did not further specify who the counterparty is. We are told in the survey that “Over half of the respondents trade over 50 million units per day.” If these units are dollars then that bar would be pretty low for the sell side firms I am familiar with. So maybe again I am not part of the intended audience for this survey. As in all statistics, the devil is in the details.
In any case – Happy Holidays to all our readers!