I came across this interesting article today, which contains some relevent information about arbitration trades:
(Unfortunately, there was no link included on the original forum, so I do not know the source.)
----------------------------------------------------------------------------------
1. I can't speak for the industry as a whole, but the ratio of winners to losers amongst spread betting clients with my company is not quite as bad as 10:90. It really depends on the product in question and how you define "active"; taking the community of daily Dow traders dealing over any given week I'd say 70% are losing money over their entire trading history. But looking at a sample like that inevitably introduces "survivorship bias".
If you look at all clients ever to have traded, then the proportion of losers must be 95% plus, purely because so many people open accounts, blow their trading reserves in a hurry and then stop dealing. Newcomers tend to get overconfident after initial success and deal in an inappropriate size. Then a small move against them puts them out of the game.
At the level of individual trades, if you correct for transaction charges (i.e. spread) you see a cumulative distribution of results that is almost a perfect bell-curve, centred on zero. As a number of academics like to claim, day traders as a whole neither make nor lose in the long run, once one corrects for dealing charges.
A cumulative distribution hides a lot of individual stories of course. Some clients make a lot of money, month in, month out. A slightly larger number lose far more than one would expect from the spread they've paid, month in month out. Whether there's skill (for the winners) or "anti-skill" at work (for the losers), or whether it's all down to chance is a subject over which I'm agnostic. There's plenty of serious academic theory on both sides of that particular debate.
The one thing that's pretty indisputable is that, the more transaction charges you rack up, the harder it is to beat the bell curve. So daytraders who take just a couple of intraday views will always end up doing better than those who jump in and out of the market like maniacs.
2. The most successful clients are arbitrageurs who, by definition, always win. These people open up spread betting accounts with 4-5 online bookmakers and then spend their time trading on arbs when daily prices get out of line. For example, at 7.15pm on the night of a chaotic FOMC announcement, you might see one company with a daily FTSE price of 4200-08, and another with a price of 4212-20. The arber buys the first, sells the second and locks in a certain profit.
Clients who do this are very obvious to the dealers on duty, as arb trades stick out like a sore thumb on a client's trading record (all trades are opened out-of-hours and are on daily products, all trades are left to expiry). After a brief honeymoon period they find that an unusual number of online trades are being rejected. Arbers hate this, as it typically leaves them with an exposed position on the non-rejected leg of the trade, which they then have to either take a chance on or close off, incurring spread.
3. If I had to spread bet for a living, I would trade housing markets. A couple of bookies do these; the key point is that they're impossible to hedge, so prices get forced way out of line by herd behaviour in the client base. For instance, a few months back everyone in the UK was betting that London house prices were going to crash. They all sold the bookies' quotes, and both companies involved had to move their futures prices on the London market through the floor, in a desperate attempt to attract buyers to balance out their risk. For a number of weeks it was possible to go long of London housing at a level that could only have lost you money in the event of the most catastrophic market move in history. Needless to say, the market didn't collapse and anyone who got long made a killing.
Apart from that, I'd go after any market with the same characteristic (i.e. unhedgeable markets where bookies get pushed from the "right" prices by flows of unthinking client business). Off the top of my head I'd list daily FTSE (only out-of-hours, mind) daily Dow (ditto) and binaries.