Pine Script Algorithmic Trading: Signals, Backtests and Execution

Pine Script can express trading rules, display signals and simulate strategy orders on TradingView. Live order execution is a separate integration. A chart label, a backtest fill and a broker fill are three different events.

Start with an indicator

This moving-average example uses 9- and 21-bar SMAs. Despite “Strategy” in its display title, it declares indicator() and does not submit strategy orders. The plotting calls have been moved out of local scope; their series conditions now control whether labels appear.

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

//======================================================================//
//                    ____  _            _  __                          //
//                   |  _ \(_)_ __   ___(_)/ _|_   _                    //
//                   | |_) | | '_ \ / _ \ | |_| | | |                   //
//                   |  __/| | | | |  __/ |  _| |_| |                   //
//                   |_|   |_|_| |_|\___|_|_|  \__, |                   //
//                                             |___/                    //
//======================================================================//

//@version=6
indicator("[Pineify - Best Pine Script Generator] Enhanced Moving Average Strategy", overlay=true)

// Input parameters for customization
fastLength = input.int(9, title="Fast MA Length", minval=1)
slowLength = input.int(21, title="Slow MA Length", minval=1)
showSignals = input.bool(true, title="Show Buy/Sell Signals")

// Calculate moving averages
fastMA = ta.sma(close, fastLength)
slowMA = ta.sma(close, slowLength)

// Plot moving averages
plot(fastMA, color=color.blue, linewidth=2, title="Fast MA")
plot(slowMA, color=color.red, linewidth=2, title="Slow MA")

// Generate trading signals
buySignal = ta.crossover(fastMA, slowMA)
sellSignal = ta.crossunder(fastMA, slowMA)

// Display signals on chart
plotshape(series=showSignals and buySignal, location=location.belowbar, color=color.green, 
              style=shape.labelup, text="BUY", size=size.small)
plotshape(series=showSignals and sellSignal, location=location.abovebar, color=color.red, 
              style=shape.labeldown, text="SELL", size=size.small)

// Background color changes for trend identification
bgcolor(fastMA > slowMA ? color.new(color.green, 95) : color.new(color.red, 95))

The labels show crossovers and crossunders, while the background shows which average is higher. An open bar can change the crossover before it closes. There is no alertcondition() or alert() call in this example, so it should not be presented as a complete alert-routing setup.

The two lengths are demonstration inputs. The earlier claim that 200 days of SPY testing selected them had no reproducible result, and no win rate is assigned to this example.

Build a strategy as a separate step

A strategy needs a strategy() declaration and explicit order commands, plus sizing and exit rules. The Strategy Tester evaluates those simulated orders under its execution assumptions. Replacing the declaration alone does not turn the displayed labels into trades.

Specify entry timing, initial protection, exit conditions, commission, slippage and available capital. Compare trades against the underlying conditions. Testing a fixed number of bars is not a universal validation threshold; the sample should include enough independent trades and relevant market conditions.

Connect alerts deliberately

Pine alert code defines events that can be selected in TradingView’s alert interface. A webhook sends an HTTP message to the chosen endpoint. That receiver must interpret the message, authenticate appropriately and route orders to the intended broker if live automation is required.

A notification is not proof of an accepted order. Log alert time, receiver acknowledgement, broker order status, rejection and fill. Design duplicate-message handling and a way to stop routing before relying on automation. The strategy’s simulated equity is not automatically your broker account balance.

Understand the environment

Pine executes on TradingView’s infrastructure; the chart and editor are interfaces to it. It does not provide arbitrary outbound HTTP requests or general Python-style package installation. It can use supported datasets and calculations within Pine’s documented limits, including some numerical or machine-learning-style methods that fit those limits. Saying no machine-learning calculation is possible is too broad.

For higher-timeframe requests, distinguish developing data from confirmed data. Combining indicators into one Pineify-generated script can reduce separate chart instances, but does not remove TradingView plot, request, execution or subscription limits.

The former six-month trading diary, 60% trend capture and halved TSLA false entries had no supporting records. The practical value here is a reproducible development sequence, not a claim of live trading results.

Sources checked October 20, 2026: Pine strategies, alerts and limitations.