Bipins.Trading

A .NET algorithmic trading library that provides a technical indicator library (100+ indicators with batch and streaming APIs) and a broker-agnostic strategy engine for backtesting and paper trading. The engine consumes normalized market data, uses the indicators library, and emits chart-ready signals and execution-ready order intents.

Installation

Install the package from NuGet:

dotnet add package Bipins.Trading

Or via Package Manager:

Install-Package Bipins.Trading

Supported Frameworks

This library targets multiple frameworks for maximum compatibility:

• .NET Standard 2.1 - Supports:
  • .NET Framework 4.8+
  • .NET Core 3.0+
  • .NET 5+
• .NET 6.0 - Modern .NET with full feature support
• .NET 7.0 - Latest LTS with required members support
• .NET 8.0 - Long-term support version
• .NET 9.0 - Latest features and performance improvements
• .NET 10.0 - Future-ready with cutting-edge capabilities

The NuGet package will automatically select the appropriate target framework for your project.

---

What Is Implemented

1. Technical indicators library (Bipins.Trading.Indicators)

• Abstractions: IIndicator, IStreamingIndicator<T>, IBatchIndicator<T>, IStatefulIndicator; result types SingleValueResult, MultiValueResult, BandResult.
• Base: IndicatorBase<TResult> with warmup, Update(Candle), Compute(ReadOnlySpan, Span), Reset().
• Utilities: RingBufferDouble, ArrayPoolDoubleBuffer, SpanMath for low-allocation and span-based math.
• Models: Candle (OHLCV), Quote, TradeTick, OrderBookLevel, IIndicatorResult.

2. Strategy engine (broker-agnostic)

• Domain: Core models (Candle with Symbol/Timeframe, Position, OrderIntent, SignalEvent, Fill, PortfolioState, StrategyState), enums (PositionSide, OrderSide, OrderType, TimeInForce, SignalType), event envelope (ITradingEvent, SignalEventEvent, OrderIntentEvent, FillEvent, PositionChangedEvent, RiskDecisionEvent, DiagnosticEvent, ErrorEvent), strategy API (IStrategy, StrategyContext, StrategyResult, StrategyWarmup, IIndicatorProvider, IndicatorKey).
• Engine: In-process IEventBus / InMemoryEventBus, IndicatorProvider (per-key cache, mapping domain candles to indicator candles), IMarketDataFeed / HistoricalCandleFeed, IPortfolioService / PortfolioService, BacktestRunner and BacktestFillReceiver, BacktestConfig / BacktestResult.
• Strategies: StrategyBase, EmaCrossoverStrategy, RsiMeanReversionStrategy, BreakoutDonchianStrategy.
• Risk: IRiskManager, CompositeRiskManager, IRiskPolicy, MaxPositionsPolicy, MaxDailyLossPolicy; IPositionSizer, FixedQtySizer, FixedDollarSizer, PercentEquitySizer, AtrRiskSizer.
• Execution: IExecutionAdapter, IFillReceiver, PaperExecutionAdapter (simulated fills, optional slippage/fees).
• Persistence: IStateStore, InMemoryStateStore, JsonFileStateStore.
• Charting: IChartSink, JsonFileChartSink (signal export to JSON).
• CLI: Console app with backtest [config.json] (loads config, runs backtest, exports signals).

---

Projects

Project	Purpose
Bipins.Trading.Indicators	Indicator library: abstractions, base, models, utilities, and all indicator categories below.
Bipins.Trading.Indicators.Tests	xUnit tests for indicators (known-value tests for SMA, EMA, RSI, MACD, Bollinger Bands, models).
Bipins.Trading.Domain	Core domain models, enums, events, and strategy/indicator-provider contracts.
Bipins.Trading.Engine	Event bus, indicator provider, market data feed, portfolio service, backtest runner.
Bipins.Trading.Strategies	Strategy base and example strategies (EMA crossover, RSI mean reversion, Donchian breakout).
Bipins.Trading.Risk	Risk policies and position sizers.
Bipins.Trading.Execution.Abstractions	Execution and fill-receiver interfaces.
Bipins.Trading.Execution.Paper	Paper execution adapter.
Bipins.Trading.Persistence	In-memory and JSON file state stores.
Bipins.Trading.Charting.Abstractions	Chart sink interface and JSON file export.
Bipins.Trading.Cli	Console application for backtest (and config-driven runs).
Bipins.Trading.Tests	Engine, backtest runner, indicator provider, and risk tests.

---

Indicator Categories and Names

Indicators are grouped by category. Each supports streaming (Update(Candle)) and batch (Compute(IReadOnlyList<Candle>) or Compute(ReadOnlySpan<Candle>, Span<TResult>)).

Moving Averages (14)

SMA, EMA, WMA, SMMA, RMA, HMA, DEMA, TEMA, KAMA, VWMA, VMA, ZLMA, ALMA, FRAMA.

Momentum (14)

RSI, Stochastic, StochRSI, CCI, ROC, Momentum, Williams %R, TSI, Ultimate Oscillator, RVI, CMO, Connors RSI, Fisher Transform, DPO.

Trend (14)

MACD, MACD Histogram, ADX, Directional Movement, Aroon, Aroon Oscillator, Parabolic SAR, SuperTrend, Ichimoku Cloud, TRIX, Vortex Indicator, Linear Regression, Linear Regression Intercept, Linear Regression Slope.

Volatility (10)

ATR, Bollinger Bands, Bollinger Band Width, Keltner Channels, Donchian Channels, Standard Deviation, Chaikin Volatility, Historical Volatility, Mass Index, Volatility Ratio, Envelope.

Volume (11)

OBV, ADL, CMF, MFI, Volume Oscillator, EMV, Force Index, NVI, PVI, VWAP, Rolling VWAP, Volume Profile.

Support & Resistance (9)

Pivot Points: Classic, Fibonacci, Camarilla, Woodie, DeMark; Fibonacci Retracement; Fibonacci Extension; Linear Regression Channel; Price Channel.

Oscillators (5)

Awesome Oscillator, Accelerator Oscillator, Price Oscillator, PPO, Detrended Oscillator.

Market Structure (5)

Fractals, ZigZag, Swing High/Low, Trend Strength Index, Market Facilitation Index.

Advanced (6)

Kaufman Efficiency Ratio, Elder Ray Index, Elder Impulse System, Choppiness Index, TTM Squeeze, Z-Score Indicator.

Statistical (9)

Mean, Variance, Standard Deviation, Z-Score, Skewness, Kurtosis, Correlation, Covariance, Beta.

Price Transform (5)

Heikin Ashi, Median Price, Typical Price, Weighted Close, Log Returns.

---

Strategies (Examples)

• EmaCrossoverStrategy – Fast/slow EMA crossover; entry/exit long and short; uses IndicatorKey.Ema(period) with optional previous-bar key for crossover detection.
• RsiMeanReversionStrategy – RSI oversold/overbought thresholds (configurable); long on oversold, short on overbought; exit on opposite extreme.
• BreakoutDonchianStrategy – Donchian channel breakout (upper/lower); entries and exits based on channel penetration; uses GetMulti for upper/middle/lower.

All use IIndicatorProvider with caching (e.g. ctx.Indicators.Get(key, () => Compute(...))).

---

Techniques and Design

• Broker-agnostic: No broker or exchange dependency; domain uses symbols, timeframes, and generic order/signal/fill models.
• Batch and streaming: Indicators implement both Compute(span, span) (zero-allocation where possible) and Update(Candle).
• Stateful indicators: Reset() for reusing on a new series; warmup period before valid results.
• Performance: Span<T>, ArrayPool, ring buffers; no LINQ in hot paths in the indicator library.
• Deterministic backtest: Single-threaded bar processing, fixed symbol/timeframe order; fill receiver applies fills and collects trades; equity curve and trade list produced.
• Event-driven pipeline: Event bus for ITradingEvent; sinks (e.g. chart export) subscribe to signals and optional other events.

---

Testing Strategy

• Indicators (Bipins.Trading.Indicators.Tests): Known-value or reference tests for core indicators and models: SmaTests, EmaTests, RsiTests, MacdTests, BollingerBandsTests, ModelsTests; shared TestData where needed.
• Engine and strategies (Bipins.Trading.Tests):
  • IndicatorProviderTests: Cache behavior (e.g. compute once per key, ClearCache).
  • BacktestRunnerTests: Run backtest with synthetic candles; assert equity curve and optional determinism (same config/data ⇒ same result).
  • RiskTests: Risk policies (e.g. MaxPositionsPolicy reject when at limit) and position sizers (FixedQtySizer, PercentEquitySizer).

Run all tests: dotnet test.

---

Build and Run

Requirements: .NET 8 SDK.

dotnet restore
dotnet build
dotnet test

Publishing to NuGet

To publish the package to NuGet.org:

# Build and pack in Release mode
dotnet build src/Bipins.Trading/Bipins.Trading.csproj -c Release
dotnet pack src/Bipins.Trading/Bipins.Trading.csproj -c Release

# Publish to NuGet.org (requires API key)
dotnet nuget push src/Bipins.Trading/bin/Release/Bipins.Trading.*.nupkg --api-key YOUR_API_KEY --source https://api.nuget.org/v3/index.json

Note: Before publishing, update the repository URL in Bipins.Trading.csproj to point to your actual GitHub repository.

CLI backtest (example project in Tests folder):

cd Tests/Bipins.Trading.Cli
dotnet run -- backtest config.json

Config (e.g. config.json) can include: Symbols, Timeframes, Start, End, InitialCash, WarmupBars, DataPath (optional CSV), ChartExportPath (default signals.json). If DataPath is missing, synthetic daily data is generated for the configured range.

---

NuGet Package

This library is available as a NuGet package: Bipins.Trading

Quick Start

using Bipins.Trading.Domain;
using Bipins.Trading.Indicators.MovingAverages;
using Bipins.Trading.Indicators.Momentum;

// Create indicators
var sma = new Sma(20);
var rsi = new Rsi(14);

// Use with candles
var candle = new Candle(DateTime.UtcNow, 100m, 105m, 99m, 103m, 1000m, "BTCUSD", "1d");
sma.Update(candle);
rsi.Update(candle);

// Get results
var smaResult = sma.GetResult();
var rsiResult = rsi.GetResult();

---

Architecture (High Level)

• Indicators: Standalone library; input is OHLCV Candle (no symbol/timeframe in the indicator type); output is SingleValueResult, MultiValueResult, or BandResult.
• Domain: Defines Candle with Symbol/Timeframe for the engine; strategy context exposes IIndicatorProvider; engine implements provider and maps domain candles to indicator candles when calling the indicator library.
• Pipeline (backtest): Feed → candles per symbol/timeframe → strategy runner (with warmup) → risk → sizer → execution (paper) → fill receiver → portfolio update and event bus → chart/other sinks.