Gekko Forum
[SHARE] Simple RSI BULL/BEAR strategy - Printable Version

+- Gekko Forum (https://forum.gekko.wizb.it)
+-- Forum: Gekko (https://forum.gekko.wizb.it/forum-13.html)
+--- Forum: Strategy Development (https://forum.gekko.wizb.it/forum-12.html)
+--- Thread: [SHARE] Simple RSI BULL/BEAR strategy (/thread-100.html)

Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38


RE: [SHARE] Simple RSI BULL/BEAR strategy - crypto49er - 05-01-2018

Ankasem,

I think your config file is screwed up. Why is it calling the STC indicator when the strategy doesn't use it?


RE: [SHARE] Simple RSI BULL/BEAR strategy - crypto49er - 05-01-2018

Hi all,

Just want to share that I made a video about this strategy. The main gist of the video is about:

1. How this strategy is insanely profitable (if backtests matches live trading)!
2. How this strategy works (explained in code and using a chart).
3. How to install this strategy (including the ADX indicator from Gab0).

https://steemit.com/gekko/@crypto49er/ajklzpuu

I will probably make another one explaining how the ADX version of the strategy works and also how to install this strategy for cli.


RE: [SHARE] Simple RSI BULL/BEAR strategy - LordBB - 05-01-2018

(04-30-2018, 08:09 PM)ankasem Wrote: Hello 
there
rsi-bull-bear AdX (new update)

 | 2018-04-30 21:21:18 (ERROR): Indicator STC could not load

cli mode error
Works properly in UI mode

Stc file in the display folder
I tried 2 different stc files and the result gives the same error

another file is needed

Which stc file works in cli mode

gekko loaded 0.5.14 ago




https://github.com/gab0/gekko-extra-indicators/tree/master/indicators
thank you



About STC, I simply replaced this Stoch with this js:
https://raw.githubusercontent.com/RJPGriffin/gekko/develop/strategies/indicators/STC.js

You have also to add the config for it (that you can find on the top of the js).

p.s.: you have to install mathjs


RE: [SHARE] Simple RSI BULL/BEAR strategy - ankasem - 05-01-2018

Thank you
@lordBB

I am investigating since morning
I deleted my STL files
rsi-bull-bear AdX
Both (Ui) and (Cli) work fine
i will try your solution
thanks aganin


RE: [SHARE] Simple RSI BULL/BEAR strategy - ants88 - 05-02-2018

Hello everyone! And Congrats to @tommiehansen for this amazing strategy!
I started normal RSI_BULL_BEAR strategy, with 15 minutes candle size and these settings:
{
"SMA_long": 1000,
"SMA_short": 50,
"BULL_RSI": 10,
"BULL_RSI_high": 90,
"BULL_RSI_low": 60,
"BEAR_RSI": 15,
"BEAR_RSI_high": 50,
"BEAR_RSI_low": 20
}
After warmup period the strategy created first buy advice, from LOGS:

Trader Received advice to go long. Buying XRP
2018-05-02 01:18:21 (DEBUG): portfolioManager : newTrade() : creating a new Trade class to BUY XRP / EUR
2018-05-02 01:18:21 (DEBUG): created new Trade class to BUY XRP/EUR
...
attempting to BUY XXX XRP at kraken price: 0.69241000
2018-05-02 01:21:36 (DEBUG): [kraken.js] (addOrder) BUY XXXX XRP @ 0.69241 EUR
2018-05-02 01:21:37 (DEBUG): [kraken.js] (addOrder) added order with txid: XXX
2018-05-02 01:22:38 (INFO): trade was successful BUYING XRP/EUR
2018-05-02 01:22:42 (DEBUG): [kraken.js] entering "setBalance" callback after kraken-api call, data: { error: [], result: { ZEUR: '0.0000', XXRP: 'XXXX' } }

I can see the order in kraken, but in Gekko UI Amount of trades is still 0, not 1!
And Profit Report says Waiting for at least one trade..
Is it normal?
Thank you


RE: [SHARE] Simple RSI BULL/BEAR strategy - Xeraphim - 05-03-2018

Hi first of all thanks for sharing this strategy!

I'm trying to get it up and running but unfortunately there is a problem with loading ADX... does anyone know how to solve this?

I've added the ADX.js to gekko/strategies/indicators/



  <-- POST /api/backtest
2018-05-03 20:31:17 (INFO):     Setting up Gekko in backtest mode
2018-05-03 20:31:17 (INFO):
2018-05-03 20:31:17 (INFO):     Setting up:
2018-05-03 20:31:17 (INFO):              Trading Advisor
2018-05-03 20:31:17 (INFO):              Calculate trading advice
2018-05-03 20:31:17 (INFO):              Using the strategy: RSI_BULL_BEAR_ADX
2018-05-03 20:31:17 (ERROR):    Failed to load indicator ADX
  xxx POST /api/backtest 500 181ms -

  Error: non-error thrown: Child process has died.
      at Object.onerror (/home/user/gekko/node_modules/koa/lib/context.js:105:40)
      at <anonymous>
      at process._tickCallback (internal/process/next_tick.js:188:7)


**EDIT:** I found the error: I only moved ADX.js but you have to move ALL indicators


RE: [SHARE] Simple RSI BULL/BEAR strategy - yesman85 - 05-04-2018

(04-25-2018, 11:26 AM)tommiehansen Wrote: You're talking about the strategy as if there is a fixed setting that "just works"?
This is not the case, quite frankly -- this isn't even remotely close to the case.

It will always fail to do what's the absolute greatest thing to do, every single time.

It's not about timing a specific point in time perfectly, it's about doing more good then bad in the long run.

No I agree fully with you Tommie, I'm just trying to understand what some of the implications are for each parameter. Unfortunately it's hard to figure out as each parameter influences eachother. I ran lots of backtesting too and found a couple of OK runs.

You are saying that you find that larger coins fare worse, is this because of it's non-volatile nature (compared to other smaller coins)? I found that USDT/ETH was a good candidate for stability, maybe I should look at more at something less stable?

Also found that a 12 min candle would yield better results than 5 minutes, I think I'll try to do more backtesting with some other coins.


RE: [SHARE] Simple RSI BULL/BEAR strategy - Kris191 - 05-04-2018

(05-04-2018, 03:29 AM)yesman85 Wrote:
(04-25-2018, 11:26 AM)tommiehansen Wrote: You're talking about the strategy as if there is a fixed setting that "just works"?
This is not the case, quite frankly -- this isn't even remotely close to the case.

It will always fail to do what's the absolute greatest thing to do, every single time.

It's not about timing a specific point in time perfectly, it's about doing more good then bad in the long run.

No I agree fully with you Tommie, I'm just trying to understand what some of the implications are for each parameter. Unfortunately it's hard to figure out as each parameter influences eachother. I ran lots of backtesting too and found a couple of OK runs.

You are saying that you find that larger coins fare worse, is this because of it's non-volatile nature (compared to other smaller coins)? I found that USDT/ETH was a good candidate for stability, maybe I should look at more at something less stable?

Also found that a 12 min candle would yield better results than 5 minutes, I think I'll try to do more backtesting with some other coins.

I'm still trying to figure this out i get the bull bear state and how it uses  over bough and sold signals, how the ADX and mods work i have no idea. if anyone wants to help me out with an explanation it would be welcomed.

However the strat is working and i'l fine tuning using average numbers from backtests.


RE: [SHARE] Simple RSI BULL/BEAR strategy - crypto49er - 05-04-2018

(05-04-2018, 05:58 AM)Kris191 Wrote:
(05-04-2018, 03:29 AM)yesman85 Wrote:
(04-25-2018, 11:26 AM)tommiehansen Wrote: You're talking about the strategy as if there is a fixed setting that "just works"?
This is not the case, quite frankly -- this isn't even remotely close to the case.

It will always fail to do what's the absolute greatest thing to do, every single time.

It's not about timing a specific point in time perfectly, it's about doing more good then bad in the long run.

No I agree fully with you Tommie, I'm just trying to understand what some of the implications are for each parameter. Unfortunately it's hard to figure out as each parameter influences eachother. I ran lots of backtesting too and found a couple of OK runs.

You are saying that you find that larger coins fare worse, is this because of it's non-volatile nature (compared to other smaller coins)? I found that USDT/ETH was a good candidate for stability, maybe I should look at more at something less stable?

Also found that a 12 min candle would yield better results than 5 minutes, I think I'll try to do more backtesting with some other coins.

I'm still trying to figure this out i get the bull bear state and how it uses  over bough and sold signals, how the ADX and mods work i have no idea. if anyone wants to help me out with an explanation it would be welcomed.

However the strat is working and i'l fine tuning using average numbers from backtests.

Kris191,

I made a video on the ADX indicator. That's definitely a starting point to understanding this strategy. It is specific to the indicator and how we can use it for day trading / swing trading. It was so surprising to me that I never heard of this indicator until I found here in the Gekko forum.

https://steemit.com/trading/@crypto49er/3nr7fq40

I will try to make another video specifically talking about how this indicator is being used for the strategy.


RE: [SHARE] Simple RSI BULL/BEAR strategy - Henry151 - 05-04-2018

I am very curious to find out if others are having a similar experience to mine. After spending quite a while figuring out how gekko works and getting the RSI BULL/BEAR ADX strategy working, I'm finding that in backtests, it seems like my best parameters work best for older data, but that it is very hard to find parameters that work well on more recent data (like the last two months). If I find parameters that work well on the last two months, it seems that they don't work as well on longer periods of data and older periods of data (like say, the last 8 months, or the last 2 years).

See for example the screenshots. Theses are all from one backtest with one set of parameters; this is an example of a set of parameters working well on the last 8 months of data, but if you look at the detail in the last month, not performing nearly as well.


My real question is, do you all think it is possible that this can be accounted for with a line of reasoning like "as more people begin to run this bot live, it results in it being  less profitable to do so"?

That's not phrased very well but hopefully my idea will get across and some of you can weigh in on whether or not you feel that is likely.

--EDIT--
this is obviously also just to show off my backtest results and see if you guys think it looks impressive or not Wink