Even as a new trader I found backtesting to be extremely resourceful, in fact even more so. Although ICM and fxview have simple interfaces but still as a newbie if I was to start trading without looking at the past data, everything would have come as a bigger surprise to me. I know the market's unpredictable but having some figures to look at helps to an extent in developing a strategy more realistically.
As a retail scalper, you can't really scalp only 1 tick because you're always paying the bid-ask spread. For example if you bought at 3100.25 and the markets froze, you can only immediately sell at 3100. You always pay 1 tick, so the market would need to move 2 ticks in your favor to profit 1 tick. Overall for reducing slippage it is important to have the lowest latency to your broker, and broker to the exchange. This is the reduce why HFT co-locate servers with exchange servers, but the cost for something like this is in the thousands per month and not feasible for scalpers. The next best thing is to trade on a high performance, reliable, VPS. Would highly recommend checking out https://www.ninjamobiletrader.com/ Connections are 1,000mbs, so the server is always connected to the broker/data source. They have a free trial and setup servers immediately.
%% LOL/exactly. UNLESS you can figure out a way to use positive slippage......... I use automated orders a lot\ but almost always a limit order getting in. Personaly , i do stuff with average day volume >> than ES + some volume less that ES. Even with a larger spread or smaller bid\ ask than ES i almost never use a market entry order
Can you really get anywhere near competing with professional high frequency traders as a retail trader? Can you even get anything like accurate tick data for back testing? Maybe I am missing something, but it sounds as if the assumptions made must be wrong for a retailer backtest to be trying to pick up a single tick. I have always though that a retail trader would have to look at longer term trades than this to have a chance.
I traded ES (well, MES) on IB's paper trading platform using their API. For testing of algorithms, I went with market orders. The spread is very small most of the time (like 1 tick) if liquidity is high enough. ES is very liquid a lot of the time, I believe. Or if you are going long, you could do LMT and be prepared to cancel your order if it does not fill within X seconds. Then you would have to wait for another signal or resubmit at a different LMT.