Apparently, transactions which increase liquidity provide a credit. Clearly, this is an opportunity for profit. To minimize risk, you ideally either want to use a very stable individual stock, or a basket of stocks that together cancel out each other's change in price (negatively correlated). A basket of stocks is harder to manage, but provides more ECN credits per basket transaction. What group of stocks on the NYSE, or Nasdaq, would you select that you would expect to be negatively correlated? I'd prefer fairly liquid stocks priced between $4 and $10. Also, not too many stocks; hard to manage many at the same time in day trading. Also, I am limited to $9900 of capital. 5 minute timer comes on if loss exceeds $50. I am knocked out for the day, if my loss isn't brought below $50 in the 5 minute period. I'm new to trading. Currently, I am just experimenting. My boss encourages this. When I explained this idea to him he told me about "batch" trading.
Maybe, maybe not. Tradeworx claims, if I remember correctly, that the cost of adverse selection for liq adding orders on such ECN's now equals the gain on the liquidity rebate, on average and for each such ECN individually. Find the papers on it -- it's a lot less clear than it may first appear.
It depends on the type of order and which ECN I direct it through. I have a pile of choices. Some of the guys I work with manage to finish the day having paid no fees, and have actually received a few dollars of credit trading one stock. Now, using that logic, if I traded multiple stocks simultaneously, this credit would add up very quickly. "how many dollars per 1000 shares" pennies not dollars. As I said above, there is no straight answer to your question.
What you're trying to do is going to be very, very difficult with Swift/Title's $50/day max loss. Even with very small lots, if you diversify into two stocks you'll have to cut losers at points where you should be adding more. Try SIRI/C, finding the right spots to add liquidity to make scratch trades (no gross $ gain, but net $ credit gain).
Thanks. Yes, I know it is going to be hard. I was experimenting with C on Friday using 500 shares. My usual $.03 loss limit was very painful. The trouble with C is that the price change occurs much, much too slowly. I don't think scalping C is a very profitable idea. The only way to make this idea work is with a combination of stocks that quite literally cancel out each other's price change, and create a rapid two or three tick price differential at any given time. The result is a price change that provides the ECN credit and also provides the opportunity to scalp.
How about SIRI, that doesn't seem to move much.. If it does drop tends to recover. Also if stock you choose has weekly options that might be a way to hedge..
C is known as THE high frequency stock for automatef HFT market making. You are competing with the big boys. Their orders are out way b4 you saw you had to so you are always the sucker in this game IMO. Your other idea is also a good one but it is know as correlation trading in the HFT world. If PEP move 3 cents then you can assume KO will move at least 2 so you hit and provide 2 cents away. For achieving that you will need some serious ammo to beat the algos.
with all due respect, I find you'll fail. Rebate trading these days are very difficult with HFT dominating the inside spread.......... daytraders just don't have the tools to compete imo.