ETF (mostly) trading system development

[QuantpTrader]

Hi guys.

I will post here info about Algo Trading Strategies and Trading System Development. This is for educational purposes only. You shouldn't follow any of the signals of the trading systems that i will disclosure and that i usually post also on my twitter.

Any feedback (good or bad) will be appreciated. Also any new cool ideas that can be tested (good or bad) are also appreciated.

 

[QuantpTrader]

The first system that i will present is based on the Internal Bar Strenght (IBS) concept. I called it #QQQIBSM1.

It's a long only system apllied to QQQ. Its shows historical performance of 15.41% (Annual Return). Here is the equity line.

 

[QuantpTrader]

#QQQIBSM1 system is on daily data. Entries are made on the open of next day after the signal. Exits are also on the next open day after the exit signal.

 

[traider]

The first system that i will present is based on the Internal Bar Strenght (IBS) concept. I called it #QQQIBSM1.

It's a long only system apllied to QQQ. Its shows historical performance of 15.41% (Annual Return). Here is the equity line.

How many trades does the system make per year?
Is there any optimization involved?
Just IBS shouldn't give you such a smooth equity curve.
Does the system average down?

 

[QuantpTrader]

How many trades does the system make per year?
Is there any optimization involved?
Just IBS shouldn't give you such a smooth equity curve.
Does the system average down?

Will post full details soon.... but:

How many trades does the system make per year?

Average 21 trades.

Is there any optimization involved?

Yes, between 2000-2005. I always divide my data in periods of 5 years: 2000-2005, 2005-2010 and 2010-2015. I do optimization in one of those periods.

Just IBS shouldn't give you such a smooth equity curve.

True.

Does the system average down?

Nop

 

[QuantpTrader]

Some stats before the code!

 

[QuantpTrader]

Before I put the rules, some explanation. My signals are short term by nature (swing trading), but if I can ride a trend I don't mind.

My solution to this is using a simple (not so) proprietary indicator that permits that. The indicator is named (because is that what it is

) a ATR_STOP. See an example.

 

[QuantpTrader]

In my research, at least in QQQ i found that the last trading day activity isn't so much important and i allways check if the signal the trading day before get better overall results.

For this system, the IBS pattern is found in the previous day, not in the last trading.

The entry signal is:

If yesterday IBS<20 and today not sellday (i will define this better when i talk about the exit signal) we buy tomorrow at the open.

IBS pattern is define as follows: (Close - low)/(High - Low) * 100

 

[QuantpTrader]

There are two possibilities to exit the trade. It will depends if the yesterday the Close is greater than ATR_STOP.

If thats the case the Exit Signal is: today Close less than ATR_STOP.

If NOT the Exit Signal is Close equals the higgest High Value of Closes of last 3 days.

If you like to know the formula for ATR_STOP contact me. I will provide it.

This rule permits ride trends if the price action is above ATR_STOP and exit trades faster in a down market (bellow ATR_STOP)

 

[QuantpTrader]

Here is the complete rules for the #QQQIBSM1 system:

• Entry/Exit price: Allways at the next open daily bar.

• Its not possible entry and exit at the same bar (only executes on of the signal - Flat - Long Signal, Long - Exit Signal).

• It only opens a position at a time.

• MAIN RULES

IBS = (Close - low)/(High - Low) * 100​

Sell_1 = (Yesterday close greater than ATR_STOP AND today close less than ATR_STOP) OR (Yesterday close less than ATR_STOP AND close equal to highest high value of the closes of last 3 days).​

Sell_2 = (Yesterday close less than ATR_STOP AND close equal to highest high value of the closes of last 3 days)​

• BUY & SELL SIGNAL

Buy = IF previous day IBS<=20 AND today NOT Sell_2
Sell = Sell_1​