Compare

sigc vs QuantConnect Lean

Open-source event-driven trading engine in C# / Python.

Lean is the engine behind QuantConnect: a full event-driven algorithmic trading platform written in C# with Python bindings. It has deep brokerage integrations, an institutional-grade data catalogue, and a hosted research environment. sigc is much smaller in scope: a typed DSL for signal-to-portfolio pipelines with a Rust runtime, leaving execution to lighter integrations like Alpaca.

Feature

sigc

QuantConnect Lean

Architecture

Compiled DSL + Rust runtime; signal-to-portfolio is the unit of work

Event-driven engine; algorithms react to OnData / OnOrderEvent callbacks

Language surface

Custom .sig DSL (data, params, signal, portfolio) with compile-time checks

C# or Python classes deriving from QCAlgorithm

Brokerage integrations

Alpaca (paper + live) per the README; BYO for others

Broad: Interactive Brokers, Tradier, OANDA, Coinbase, Bitfinex, and more

Data

Loads CSV, parquet, S3/GCS, Postgres; bring your own data warehouse

Built-in data catalogue (US Equities, Forex, Crypto, Futures, Options) via the QuantConnect platform

Determinism

Content-addressed cache via blake3 + sled; identical inputs = identical outputs

Reproducibility tied to the engine; deterministic if the data and engine version are fixed

Asset classes

Cross-sectional equities is the canonical use case in the repository examples

Equities, FX, crypto, futures, options across the platform

Backtester features

Rank-based long-short, vol-target, bps + square-root slippage, benchmark attribution

Full event-driven backtest with fills, slippage models, and broker fees by venue

Production mode

sigc daemon (one binary) with sigc.yaml posture + Prometheus /metrics

Lean engine deployed to a server or run on QuantConnect-hosted infrastructure

Footprint

Single Rust binary + sled cache directory

Engine + plugins + data subscriptions; heavier surface area

License

MIT

Apache 2.0 (Lean engine)

Pick QuantConnect Lean when

You need a full event-driven engine, broad brokerage integrations, a managed data catalogue, and asset classes beyond equities. The QuantConnect platform is the obvious choice if you want the hosted research environment as part of the package.

Pick sigc when

You want a small, typed, Rust-native pipeline for the signal-to-portfolio half of the problem, with deterministic caching, and you are comfortable using Alpaca (or your own integration) for execution.

Honest note

These tools overlap less than the names suggest. Lean is a trading engine; sigc is a signal compiler. A serious shop could use both: write the alpha in sigc, route the resulting weights to a Lean-managed execution layer. The comparison rows are about features, not winners.

Back to overview Read the docs