Quote from amswak:
So in summary I am looking for a software/broker/service that:
a) allows setup of powerful strategies in automated trading, down to at least 1-min intervals
b) allows backtesting
c) allows "test-driving" of strategies where you can test it in real time and view results
d) has either a per-share commission schedule or a very low per-trade rate
Sorry amswak I have misread you. When I saw the word "new" I automatically associated that with "beginner".
Let me try to be useful to the questions you asked:
I am a TradeStation user. I have taken their Easy Language class and had tried to do play with some programmed trading. But my results were horrible and I had given up on it.
I am doing day trading (equity mostly) for a living. All descretionary trades. I have my "systems" (methodologies) to trade but not that I am good enough to write down as mathematical formula. I only know of TradeStation and no others.
1) Yes you can write strategies to trade down to 1-min level. You can get down to tick charts if you want to (e.g. 144 tick, 288 tick). A bar is a bar to TradeStation whether it is weekly, daily, 1-min or 144 tick. The lower the time-frame, the more your trades are likely to get whip-sawed. There are many headfakes intraday. Very easy to get whipped in lower time-frames. Also, occassionally you get out-of-sync ticker data. These will render your strategies useless (rules fired based on false data). A lot of times, market makers like to throw these things out (because they can) to trigger some stop orders.
2) Yes you can certainly back-test. Using end-of-day data or intraday data. TS is very flexible. Intraday data goes back to at least 10 years. But the more you "back-test" your strategies, the more likely you will be confused. Because usually a given strategy tends to work well either in a trending market or in a trading market, but rarely both. The market often changes from trending to trading without warning. On an intraday basis, the price moves typically trend ups and downs and stay in range a few times. You probably should develop some logics to tell you whether at the moment you are in a trending market or in a trading market (e.g. Higher-highs/lower-lows versus equal-highs/equal-lows) then decide which strategies to employ.
3) IMO forward-testing is much better than back-testing. Test your strategies with minimum shares and evaluate with real trading results. You can see the trade markers on the charts once a trade is filled. You can see your open P&L and closed P&L in real-time.
4) TS charges on a per-share basis. $1.00 for 100 shares. If over 500 shares, $0.60 for 100 shares. Bench-mark 1000 shares: $8.00.
I used to follow my set of methodologies and traded the 1-min charts. They worked fairly well. I had recently lowered my time-frame and traded on the 42 Tick charts. My experience:
- When I lower my time-frame, I get a lot more trading signals. But this is a two-edge sword. I can easily get whip-sawed.
- Tradable price patterns are fractal. They manifest in the highest time-frames (e.g. weekly, daily) and the lowest time-frames (e.g. 42 tick, 144 tick). For example, you will see Double Bottoms/Double Tops all the time in all time-frames. The lower time-frames serve as a magnifier for price moves in higher time-frames.
- I use a higher time-frame as my guide (which direction) to trade, but a lower time-frame as a guard and timing device. e.g. On a 1-min chart I should buy. I plan to buy. But on a 42-tick chart it looks bearish. I will hold. When the 42-tick chart turns bullish, that's when I will buy.
Again, good luck!