0bps

Strategy vault / Other

Moving Average Colored EMA SMA

Amazon Prime Video presents Garaj Garaj Jugalbandi Video Song from Bandish Bandits. A folk song performed by Ritwik Bhowmik and Atul Kulkarni, sung by Farid Hasan & Mohammed Aman, music directed by Shankar Ehsaan Loy…

Author
ChaoZhang
Complexity
Intermediate · 92 lines
Parameters
9
Backtest bar
1h · intraday
Tested on
BTC/USDT · futures

Amazon Prime Video presents Garaj Garaj Jugalbandi Video Song from Bandish Bandits. A folk song performed by Ritwik Bhowmik and Atul Kulkarni, sung by Farid Hasan & Mohammed Aman, music directed by Shankar Ehsaan Loy and lyrics by Sameer Samant. In the season finale Radhe and Digvijay repeat this song as an ode to their common teacher, Panditji.

backtest

IMG

Parameters

NameDefaultWhat it controls
v_input_1true(?Moving Average Colored EMA/SMA)Show EMA on chart
v_input_int_113Ema Length
v_input_2falseShow SMA on chart
v_input_int_223Sma Length
v_input_int_350(?Holders Sentiment)length
v_input_3_close0Source: close | high | low | open | hl2 | hlc3 | hlcc4 | ohlc4
v_input_timeframe_1D(?Supertrend)Resolution
v_input_410ATR Length
v_input_float_13Factor

Source Pine Script

/*backtest
start: 2021-05-10 00:00:00
end: 2022-04-22 05:20:00
period: 1h
basePeriod: 15m
exchanges: [{"eid":"Futures_Binance","currency":"BTC_USDT"}]
*/

// This source code is subject to the terms of the Mozilla Public License 2.0 at https://mozilla.org/MPL/2.0/
// © subhajitbhattacharya


//@version=5
indicator('Moving Average Colored EMA/SMA', overlay=true, precision=2)
gr1 = "Moving Average Colored EMA/SMA"
emaplot = input(true, title='Show EMA on chart',group=gr1)
len = input.int(13, minval=1, title='Ema Length',group=gr1)
src = close
out = ta.ema(src, len)
up = out > out[1]
down = out < out[1]
mycolor = up ? color.green : down ? color.red : color.blue
plot(out and emaplot ? out : na, title='EMA', color=mycolor, linewidth=3)


smaplot = input(false, title='Show SMA on chart',group=gr1)
len2 = input.int(23, minval=1, title='Sma Length',group=gr1)
src2 = close
out2 = ta.sma(src2, len2)
up2 = out2 > out2[1]
down2 = out2 < out2[1]
mycolor2 = up2 ? color.green : down2 ? color.red : color.blue
plot(out2 and smaplot ? out2 : na, title='SMA', color=mycolor2, linewidth=1)


//===================== Holders Sentiment ==================================================

gr2 = 'Holders Sentiment'
length = input.int(50,group=gr2)
src3 = input(close, title='Source',group=gr2)

zeroline = 0
// hzplot = plot(zeroline, color=color.new(color.black, 0))
difabs = (src3[0] - src3[length]) / src3[length] * 100
plotColour = difabs > 0 ? color.green : color.red
// changeplot = plot(difabs, color=plotColour, linewidth=2)
// fill(plot1=changeplot, plot2=hzplot, color=color.new(color.blue, 90))

//================ Supertrend ================================================

gr3 = "Supertrend"
tf = input.timeframe('D', "Resolution",group=gr3)
atrPeriod = input(10, "ATR Length",group=gr3)
factor = input.float(3.0, "Factor", step = 0.01,group=gr3)

[supertrend, direction] = request.security(syminfo.tickerid, tf, ta.supertrend(factor, atrPeriod), barmerge.gaps_off, barmerge.lookahead_off)

bodyMiddle = plot((open + close) / 2, display=display.none)
upTrend = plot(direction < 0 ? supertrend : na, "Up Trend", color = color.green, style=plot.style_linebr)
downTrend = plot(direction < 0? na : supertrend, "Down Trend", color = color.red, style=plot.style_linebr)

//fill(bodyMiddle, upTrend, color.new(color.green, 90), fillgaps=false)
//fill(bodyMiddle, downTrend, color.new(color.red, 90), fillgaps=false)

//=================================================================================

bs1 = ta.crossover(out,out2) ? 1 : ta.crossunder(out,out2) ? -1 : 0
bs2 = difabs > 0 ? 1 : -1
bs3 = direction < 0 ? 1 : -1
bs = 0
bs := bs1 == 1 and bs2 == 1 and bs3 == 1 ? 1 : bs1 == -1 and bs2 == -1 and bs3 == -1 ? -1 : 0

buy = bs == 1 and nz(bs[1]) !=1
sell = bs == -1 and nz(bs[1]) != -1

//bgcolor(buy ? color.new(#00ff00, 50) : na)
//if buy
//    label.new(bar_index, low, yloc = yloc.belowbar, text="Buy", textcolor = #00ff00, size = size.large, style=label.style_none)
//
//bgcolor(sell ? color.new(#b92ad8, 80) : na)
//if sell
//    label.new(bar_index,high, yloc = yloc.abovebar, text="Sell", textcolor = #b92ad8, size = size.large, style=label.style_none)
    

alertcondition(buy, title='Buy Alert', message='Buy')
alertcondition(sell, title='sell Alert', message='Sell')


if buy
    strategy.entry("Enter Long", strategy.long)
else if sell
    strategy.entry("Enter Short", strategy.short)

Published by ChaoZhang on the FMZ strategy square · last modified 2022-05-11. Collected in the Quant Trading Vault. Unvalidated research: re-test with your own data and costs before trusting it.

Build this with AgenKit

A spec is not a trading system. Paste this into your coding agent and AgenKit runs a specialist team — architecture, test-first build, risk gates, review — with paper trading on by default.

in Claude Code, Cursor, Codex & 4 more
/agenkit build a production-grade trading bot with a backtest engine, risk controls and paper trading by default from strategies/Moving-Average-Colored-EMA-SMA.md
  1. npx agenkit install engineering-kit in your project
  2. Paste the command into your agent and approve each phase
  3. Get tests, backtest, risk controls and a paper-trading switch

Understand it first

Before you run this, read the lesson behind the idea and how to backtest it honestly. Specs like this one are the survivors of an unknown number of trials; deflate their results accordingly.

Related specs

esc

Try , , or .

↑↓ move↵ open