Coin flip study

I like the logical approach and clear enter and exit steps.

All you can do is try and see what happens.

Start manually backtesting 1000 trades.

Why not start with 200 trades?

 

Why not just flip the coins into a wishing well?

 

Sure, Why not? Help out those homeless guys.

 

Actually I will, right here manually in this thread. I'll start with remaining October sessions, and include today since we're only 1/2 hour into session. Entry will be based on open price of second 1 hour time bar (10:30 EST). Exit price will be 1/2 Daily ATR(20) stop loss or open price of last 15 minute time bar (15:45 EST). Every following month I will just regenerate an equal amount of coin flips to trading sessions for that month at https://www.random.org/coins/ starting from top row down/left to right like so:


10/18 Heads
10/21 Heads
10/22 Tails
10/23 Heads
10/24 Tails
10/25 Tails
10/28 Heads
10/29 Heads
10/30 Heads
10/31 Tails

These are the remaining sessions for October 2019. I'll go with just NQ for now. To be continued..............

 

I do like the idea. Not sure why all the animosity.....probably because you (digitalnomad) put it out there saying you weren't going to test it. Now that you've bellied up to the bar it may tame the responses somewhat.

Anyway, I don't think the performance is as clear cut as many might think. Definitely has to be simulated over many trials to get a statistically significant result. I don't think 1000 is enough. The challenge I have in terms of simulation is that it requires intraday data......all of my strategies use EOD data so it would require some work to get the required data.

 

The intraday/minute OHLC data is free in popular sites like this:

https://finance.yahoo.com/quote/NQZ19.CME?p=NQZ19.CME

 

Hey, you had a nice journal going, did you give up on it?

 

No still doing it. Just needed some entertainment. Frozen trader was ragging me about my journal being a comedy. So I was just ragging him back about his coin toss.

 

Gary I set up a spreadsheet to assist with this study and will post results at end of month for the above coin flips, followed by a re-flip before start of next month. Basically running an open forum walk-forward simulation month-to-month. Should be a fun experience.

Are you proficient in data manipulation using specific tools/languages? If so, PM me if you want to play with minute level OHLC data, specifically toward this study. I can provide years back for you in txt or csv.

Just to clarify with the moderators here, this isn’t a journal, but a strategy development process and/or foundation for a more sophisticated strategy, if desired.

 

All SPY profits in the last 10 years came outside regular hours, that's maybe another way to think about the profitability (without simulating).