Configurable Multi MA Crossover Voting System

This strategy allows flexible configuration of multiple fast/slow moving average pairs. It goes long when all fast MAs crossover above slow MAs, and exits when any fast MA crosses below slow MA. The voting mechanism…

Pine ScriptTrend following Machine learning Volatility SMA ATR

Author: ChaoZhang
Complexity: Starter · 65 lines
Parameters: 6
Backtest bar: 1d · swing
Tested on: BTC/USDT · futures

Jump to code View in vault Original on FMZ

Overview

Strategy Logic

The key components and rules are:

Multiple fast/slow MAs: Using SMA, WMA, VWMA etc.
Long signal: All fast MAs crossing above slow MAs.
Exit signal: Any fast MA crossing below slow MA.
Profit/loss points: Fixed points based on ATR.
Configurable: Flexible configuration of multiple MA pairs.

The voting-based entry with multiple MAs improves signal reliability. Custom configurations provide flexibility.

Advantages

Compared to single MA strategies, the advantages are:

Multiple MAs provide more comprehensive trend assessment.
Voting avoids false signals from noise.
Large tuning space from custom MA configurations.
Support for different MA types enhances adaptability.
Defined profit/loss points control per trade risk/reward.
Works better on longer timeframes, less curve whipsaws.
Simple and intuitive logic, easy to implement and operate.
Overall more stable with greater longevity versus single MA.

Risks

However, some risks exist:

Increased complexity with multiple MAs.
Risks of over-optimization.
Fundamental lagging in identifying trend changes.
No volume considered, risks being trapped.
Profit/loss points may cause unnecessary exits.
Performance subject to changing market regimes.
Need to monitor reward/risk ratios and curve smoothness.
Robustness across instruments requires validation.

Enhancements

Based on the analysis, enhancements may involve:

Testing parameter robustness across different instruments.
Adding volume or volatility confirmation.
Optimizing profit/loss points.
Setting maximum tolerable drawdown limit.
Constructing dynamic position sizing models.
Evaluating effect from introducing machine learning.
Monitoring maximum drawdown and curve smoothness.
Continual iterations to avoid overfitting.

Conclusion

The configurable multi-MA approach forms a robust position holding mechanism. But preventing overfitting and dynamic adaptations to changing markets are key for any strategy's longevity. Only through rigorous ongoing optimizations and testing can a quant strategy sustain success.

Parameters

Name	Default	What it controls
v_input_string_1	2x4,3x5,4x6	Crossover Config
v_input_source_1_high	0	source: high \| close \| low \| open \| hl2 \| hlc3 \| hlcc4 \| ohlc4
v_input_string_2	0	Moving Average Type: WMA \| SMA \| VWMA
v_input_1	14	ATR Period
v_input_2	10	Profit ATR x
v_input_3	5	Loss ATR x

Source Pine Script

/*backtest
start: 2022-09-16 00:00:00
end: 2023-09-22 00:00:00
period: 1d
basePeriod: 1h
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/
// © levieux

//@version=5
strategy(title='Configurable Multi MA Crossover Voting System', shorttitle='Configurable Multi MA Crossover Voting System', initial_capital=1000, overlay=true, default_qty_type=strategy.percent_of_equity, default_qty_value=100, commission_type=strategy.commission.percent, commission_value=0.1)
crossoverConfig= input.string(defval="2x4,3x5,4x6", title="Crossover Config")
source= input.source(high)
maType= input.string("WMA", title="Moving Average Type", options=["WMA","SMA","VWMA"])
atrPeriod= input(14, title="ATR Period")
profitAtr = input(10, title="Profit ATR x")
lossAtr = input(5, title="Loss ATR x")


ma(src,length,type) => 
    float ma = switch type
	    "SMA" => ta.sma(src, length)
	    "WMA" => ta.wma(src, length)
	    "VWMA" => ta.vwma(src, length)

crossoverGroups= str.split(crossoverConfig, ",")
crossoverCount= array.size(crossoverGroups)
crossovers= array.new_string(crossoverCount)
positions= array.new_int(crossoverCount, 0)
longVotes= 0
for i= 0 to crossoverCount-1
    crossover= str.tostring(array.get(crossoverGroups, i))
    crossoverBoundaries= str.split(crossover, "x")
    int fastLength= math.round(str.tonumber(array.get(crossoverBoundaries, 0)))
    int slowLength= math.round(str.tonumber(array.get(crossoverBoundaries, 1)))
    wmaFast= ma(source,fastLength,maType)
    wmaSlow= ma(source,slowLength,maType)
    if wmaFast>wmaSlow
        longVotes:= longVotes + 1
        array.set(positions, i, 1)

longCondition= longVotes==crossoverCount and strategy.position_size==0


//profitTicks = profitAtr*ta.atr(atrPeriod)/syminfo.mintick
//lossTicks = lossAtr*ta.atr(atrPeriod)/syminfo.mintick
profitPrice= close+profitAtr*ta.atr(atrPeriod)
lossPrice= close-lossAtr*ta.atr(atrPeriod)

if strategy.position_size>0
    profitPrice:= profitPrice[1]
    lossPrice:= lossPrice[1]

plot(profitPrice, color=color.green)
plot(lossPrice, color=color.red)

if longCondition and profitPrice>0
    strategy.entry("Long", strategy.long)

if longVotes<crossoverCount and strategy.position_size>0 and (high>profitPrice or low<lossPrice)
    strategy.close("Long")
    
longVotes:= 0

Published by ChaoZhang on the FMZ strategy square · last modified 2023-09-23. 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/Configurable-Multi-MA-Crossover-Voting-System.md

npx agenkit install engineering-kit in your project
Paste the command into your agent and approve each phase
Get tests, backtest, risk controls and a paper-trading switch

Get AgenKit How /agenkit works

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.