Quote from braincell:
Now my question would be, regarding spread trading:
- How many of you have your own proprietary software for statistical anylsis (correlation, arbitrage, lag, etc)? Is this rare, or only used by the bigger trading companies?
- If you do use something like this, how useful is it proving in finding good spreads in general?
Well, I show my clients how I do it - including finding spreads, checking correlations, calculating hedge ratios, etc.
But we have about 400 + + spread combinations just using various iterations of the exchange supported implied spreads which you can find listed in the exchange SPAN calculation tables for spread margin credits - including pairs, butterflys, condors, strips vs. strips, packs, bundles...etc. etc.. Seriously, think about all the possible combinations in those alone. { 400 x 400 = 160,000 }
Point is, you can make a very good living without bothering yourself to find something original and revolutionary, but I do show my clients how I at least go about doing it - without a Bloomberg.
I have seen Gennady Gertsman and Harris Brumfield make tens of millions of dollars per year trading their own personal accounts in the interest rate spreads - all very plain vanilla missionary position stuff all up and down the yield curve. I spent three years sharing an office with Gennady and I could see his P&L on his TT a few times per day.
And I personally have done really well doing some unique oddball quasi-arbitrage stuff like Copper versus the S&P 500, or a particular Asian currency spread versus a certain Eurodollar spread ( STIR cross vs. a currency cross, volatility adjusted )
The problem with someone's infatuation with a unique arbitrage or spread combination is that they simply do not hold up over time - and you don't know if it will be 2 days, 2 weeks, 2 months, or 2 years before it goes to hell and leaves a big mark on your forehead. The cointegration just eats you alive, and you have no advanced warning metric to tell you that the fundamental and statistical glue holding the relationships together are coming apart, and it is a sick feeling. In essense, it morphs into a pure divergence trade. Case in point ( just a few of many examples I personally know of and have traded in the past ): Nat Gas vs. PJM-W Power, Bund vs. US TY Note, Euro vs. Crude Oil, Euro vs. S&P 500, etc. etc.. I do that work on a case-by-case basis for HFs and some CTA's - there's alot to it, it requires alot of maintenance work, and it is not just a static ATM machine at your beckon call. The other point is that you probably will not get a SPAN margin credit from the exchange in terms of cap requirements.
But you are spot-on correct about the knowledge base regarding spread trading and arbitrage: I've built up a nice little business working with a very limited number of clients on a consulting basis, and swing trading futures spreads for my own personal account. In fact, the client work provides a welcome diversion away from the screen and in my mind at least might keep me from doing something impulsive and stupid out of boredom.
