Trading software development and sales I guess I left banking and institutional trading a few years ago and I am not trading (static) arbitrage myself, simply because it requires institutional infrastructure (guys here 100% correct about commissions, etc.). I think the trend headed towards ADR and then ETF arbitrage but I am not sure if it is still possible to profit from it, especially when guys like AMEX/EURONEXT offer to (auto) trade it for you Statarb can be profitable as always but it's model (brain) dependent and it's not really riskless. Cheers, Anton
Theres a problem with their example you need to do the offsetting currency transaction. So Buy IBM(US Sell IBM(Gr) and do the offsetting USD/EUR otherwise you are basically taking a currency view. Clearly this type of "Arbitrage" is dependent on costs and execution.
Well, the point is that traders don't do currency transaction with every arbitrage transactions but only a few times a day. Some transactions get "cancelled" during the day, i.e. sometimes you need to convert USD to EUR and sometimes vice versa, depending on the spread direction. Daily Forex transaction is usually done manually, most likely through the same bank Cheers, Anton
It IS possible to arb most markets through the exchange quoted calendars - the most popular markets to arbitrage are the interest rate contracts since there are multiple quoted months that present many different intra-contract spread markets to trade against. For simplicity's sake I'll give an example of arbing the YM, the way to do it would be legging the outrights against the exchange quoted spread. For example, if the front month is quoted as follows as it is right now with my cursory glance at the prices: ......Aug 06 .............. Dec 06 .... 10988 offer ...........10910 offer bid 10987..............bid 10909...... and the exchange quoted spread market was quoted as so: Aug6/Dec6 spread ........81 offer bid....80...... You could try working the offer in the back month - in the example above if you managed to sell the offer at 10910 and got a fill, you could then sell the spread at 80 bid and buy the front month at 10988 offer to go flat. The maths of this trade would be as so: short Dec 06 at 10910 minus your long at 10988 = a difference (spread) of 79. Since you bought the front month and sold the back month, this would put you long the spread market at 79 and since the spread market is bid at 80 you could sell the spread in the exchange quoted spread market - so your were long the spread via the outright legs at 79, sold the spread market at 80 = flat with 1 tick profit with relatively "limited" risk since you are hedged agains big moves by being both long and short. A lot of the depth of market you see in the DOM ladder is derived in exactly this way since the market makers's can quote the market against the spreads and back month. The downside is that your 1 tick will cost you four lots (front month + back month and spread counts as + two) so you can see why uber cheap commissions are an absolute necessity for any type of arbitrage - impossible to do with a retail broker such as IB. Easier said than done however, and I do not suggest you even consider trying this unless you understand exactly what you are doing and have the right tools to do it (ie: TT autospreader or similar)