I need help and hopefully someone here can provide it. I’m not a hedge fund nor do I have access to a lot of the technology the big boys have but I have several high frequency trading models that trade both listed and Nasdaq securities. The stocks it trades generally trade anywhere from 500K to 5M shares a day with most being in the 1-3 million shares a day area. The models have been running for a while and while they remain robust slippage has picked up quite a bit lately. The models do not scalp but rather attempt to hold the positions anywhere from a few minutes to potentially the end of day. Generally they trade 2-3 thousand shares and I use a combination of limit and market orders to get in rather than work the order because the orders are fairly small and the executions need to be taken when I get the signal because I base the exits off where I get the signal not necessarily the execution.My questions to anyone here that can help are:1. If you were doing this how would you do the order routing? Would you go market, limit + delta amount), midspread, would you route everything through an ECN like Arca that can then hit other ecn's and market makers but charge you a bit for doing it or would you attempt to hit the best bid/offer regardless of the market center?2. How would you attempt to find the proper balance between aggressive orders and passive ones? Occasionally My 2,000 lot will needlessly eat into the book of a stock, how would you protect against this?3. These “dark pools” of liquidity on the crossing networks and such, is there anyway to know or quantify how much it would help slippage if I could hit them? And how would I do that. (I only route through ECN’s, super montage, and superdot). 4. The order algo’s that are out there, I realize they are mainly for moving around large blocks and beating VWAPS but do they have any application for systems/methods trading smaller positions (2-5K shares) but needing an arrival price type of benchmark.5. With regard to listed. Does anyone still route solely through super dot or has so much liquidity moved away from the specialist these days that its better to go through an ECN there as well? And with regard to the uptick rules if the stock requires an uptick to get short is there anyway to quantify whether its better or not to trust the specialist to get you in the best or what is the best way to handle that senario?6. I guess what I’m really asking is how can I bring my self up to speed on execution, other than Google it and read nasdaqtrader.com, etc. Clearly everyone that is doing risk-arb, program trading, basket trading, and high frequency financial modeling is not using market orders but is everything just “cancel-replace, cancel-replace”, or is there a better model to the madness. Are algo’s just for the big (>25K share) orders or can they help me get a better fill on my smaller lots that need a timely execution but not necessarily a get in at all costs (market) one?All suggestions greatly appreciated. Thanks everyone.

Your question is multi-faceted and I think the short answer is "it depends".In my view, there is no one killer strategy for executing equity orders. Much depends on the liqudiity available in a particular stock and that should not be viewed as a constant- depends on market conditions, time of day, etc. I'm not sure if you are trading 5 names or 500. But, in general, don't consider it wise to throw 2,000 shares at the book with a market order. Better to use a limit order and grab what's available within a few cents of the midpoint.How aggressive you are depends on how alpha there is in your strategy. In general, I have found it better to get a partial fill than throw big orders at the book.