Quote from eusdaiki:
Hedge Funds need to move huge volumes, it takes them several days to step in and out of positions. The only way it could work is if you ran it like a prop shop where you have hundredths of traders scalping different instruments simultaneously... but then your operation costs might significantly reduce the advantage of such a hedge fund.
Yes. Size is an issue. There's an upside limit to how much you can trade.
In terms of backtesting scalping strategies, there's not a whole lotta of ways to actually test the liquidity / fillability factors. If you exclude them, sure it looks great on tests.
In a simple backtesting environment placing a limit order on previous bar's high/low price, you get a 75-85% profitability. In real life, you'll end up losing money through commissions and unfilled orders.
Another issue with intraday scalping is, the higher frequency of market patterns changing. I've verified chart patterns(technical analysis based stuff) changing from profitable to losing from 1 day to 40 days, randomly.
Taking some of the past studies I've done, the tests the original poster as done must be tested correctly. Another is his filters must be constantly updated. My advice is to start placing 1 lots into the market and get real life feedback on the strategy. Once you get a realistic feedback on how the orders are filled and continues to be profitable, I'll move to a 2 lot, 5, and 10.
Also, Even if you have the resources to test the market, as realistic as you can, you'll still under-perform.
The important factor is whether the under-performance is acceptable(profitable) enough.
