Think they are pairing the QQQs against a basket of Nas stocks? This would account for the large Q volume on their ECN, and the one stock mentioned where they did alot of volume is MSFT. Another thing about their "edge" or "secret". There really is nothing new under the sun and alot of big moneymakers - they have an edge, but it has little to do with predicting stock price movements. If they are adding liq at the current rebate rates, which is higher per share than clearing fees that they get doing such huge volume, they could very easily gross 0 from their trading and still make a profit. Breaking even on trades: 100 million shares * .0005 (.002 rebate - .0015 clearing, etc..) = $50,000/day The whole "what's their moneymaking secret" so misses the point. There are no secrets, a fact that is obscured by all those frauds who are trying to sell you a way to predict the future.
Maybe they need those ...whose money would they skim? The whole "what's their moneymaking secret" so misses the point. There are no secrets, a fact that is obscured by all those frauds who are trying to sell you a way to predict the future.
True. I think about where the money comes from though - maybe more from the ECN liquidity take fee than from other traders' accounts. The fact they have their own ECN is sick, becuase they can pocket the .002 they would have to pay out for liquidity. They are essentially up .003 on every share from the get-go. They put this crazy pricing in for January and everyone's convinced their losing money. The possibilities here are facinating - my guess is they're putting up most of the liq on the ECN anyway, so they are pocketing .002 instead of .003 but doing 200 million shares/day instead of 100 million. If they need to take liq (bail out of a bad trade), they go to another market where they can get a better fee than they would have to pay to the liq provider on THEIR OWN ECN at .003! Man, I could extrapolate this out so many ways.
Sorry, that's not what the one article said...they appear to be using bid/ask ticks and volume to determine their entries/exits. This is similar to what the Penn-Lehman project had done successfully....and AFAIK, it's been a winner for Lehman. I asked the professor about a year ago, and he said they were trading it (CBR-SOBI)....see below: http://www.cis.upenn.edu/~mkearns/papers/plat.pdf I believe "SOBI" stands for "static order balance-imbalance".
GOOD CATCH....Wow - you picked-up on a really subtle point....and this is definitely a possibility for one of their strats...an arbitrage of sorts of the major components of the QQQs vs. the QQQs themselves.
You could play the highly correlated pairs like this too. Nice low bar if you can shoot for just gross positive. I imagine places like tradebot, their concern is not "how do we get a more profitable system", but instead "how can we do more and more volume". Like a casino they just need to get more people in the door.
I'll really be surprised even if 1/10 of ET knows what an edge really is... And 1/100 of ET actually trades based on even one.