Created
December 23, 2013 20:00
-
-
Save yadudoc/8103640 to your computer and use it in GitHub Desktop.
2-EMA vs 3-EMA comparison. I use a 3-EMA Ruby cross instead of a simple 2-EMA cross. Here's some simple stats which show performance difference on Bitcoin/USD on MtGox.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
I ran a parameter sweep backtest with 2 EMA crosses, with EMAs in range 0,70 on | |
$5000 USD on MtGox. | |
Starting date : 2012-12-31 16:00:00 | |
Ending date : 2013-12-16 05:00:00 | |
Simulation on 1 hour candles from MtGox. | |
Here's the results from running simple 2 EMA crosses, sorted by final USD balance | |
EMA1, EMA2, FINAL_USD | |
9, 29, 840301.79202999303 | |
10, 29, 804076.54585023969 | |
9, 31, 787425.3005569001 | |
9, 33, 760055.62747322221 | |
9, 32, 756677.08874431625 | |
19, 33, 752223.56637418119 | |
13, 23, 707142.71903429297 | |
9, 30, 707126.33584128285 | |
9, 34, 690735.78672113759 | |
12, 27, 678506.12357716134 | |
Ruby Strategy at 60min candles | |
Buy threshold : 1% | |
Sell threshold : 3% | |
Same duration and starting parameters: | |
EMA1, EMA2, EMA3, FINAL_USD | |
10, 27, 29, 1095964.1741853813 | |
4, 6, 10, 1038221.2753517051 | |
9, 27, 29, 1007524.4618898276 | |
9, 28, 29, 978802.05582687503 | |
1, 2, 3, 963086.42278712604 | |
9, 26, 29, 959511.73509285843 | |
10, 25, 32, 950532.6488902343 | |
10, 26, 32, 949953.92246836261 | |
9, 25, 29, 948698.65668256988 | |
10, 24, 32, 943597.0524466671 | |
Clearly, the 3 EMA ruby style cross performs much better. | |
To rule out over-optimisation/curve fitting, I'll run these on data on a different time scale later. | |
I'm also concerned that the 10-30 EMA bots at cryptotrader might be influencing the 10-27-29 EMA combination. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment