One Tick Matters

[CRCatNinja]

I created a strategy that trades the ES market (maximium liquidity...b/a spread). It closes a position, using a Market Order, with 1 to 2 tick profits.

In backtesting, and setting the fill algorithm to "Default", it backtests very well.
When I run the strategy in real time mode on Sim101 account however, it produces VERY different results.

ENTRY RULE= if x happens, close open position enter new position (opposite direction) on close of bar (market order). On real-time testing, the Entry price is always 1-2 ticks above(below) the stated bar close price.
CLOSE RULE= if y happens, close that position and go opposite again on close of bar (market order). On real time testing, the Close price is always 1-2 ticks below(above) the stated bar close price.

So, my question relates to strategy development...how to incorporate this into strategy? Should I (or anyone) only EVER try to create strategy that requires 4-5 tick profit in a liquid market like ES just to realize 1-2 tick in actual profit? NT help indicates that the backtesting algo is sophisticated enough to account for estimated differentials in situations like this (certainly not akin to real market...I understand that)

Or am I mis-applying certain capabilities of NT?
Or, might actual market provide "better" fills than NT real time fill algo?
Or is execution hardware other than my PC, internet and Zen-Fire necessary?

[CRCatNinja]

Added Clarification

I am only using tick data for backtesting and strategy development.

I assumed that BID / ASK SPREAD would always be the biggest hurdle to overcome on a "scalping" type of strategy. Is that what the differences are here...a larger actual spread in real time testing than the backtest algo that NT employs?

[NinjaTrader_RyanM1]

Hello CRcatNinja,

There can for sure be differences between backtesting and real time trading:

http://www.ninjatrader-support.com/HelpGuideV6/DiscrepanciesReal-TimeVsBacktest.html

Using a one tick profit in a backtesting environment will work a lot differently than trading live. In backtesting: Without taking your stop loss into account, you will be filled on this almost every time except for when the open = high.

You can monitor a smaller time frame to get intrabar granularity during backtesting. This reference sample helps with this:

http://www.ninjatrader-support.com/vb/showthread.php?t=6652

To profit from this type of strategy in real trading, you would need to be right an unusually high % of the time. Even if your stop loss is only 5 ticks, then you need to be right > 80% of the time. One loss wipes the gains of 5 wins.