Retail Quant
← All guidesTools & Data Directory
The stack a retail quant actually needs, data, backtesting, execution, with honest notes on the tradeoffs. Most of this is free. Where a tool has a catch (rate limits, survivorship, overfitting risk), it's called out.
Market & economic data
The unglamorous truth: data quality caps strategy quality. Free sources are fine to learn on, just know their gaps (survivorship, adjustments, point-in-time).
yfinance↗
PythonFreeFree Yahoo Finance wrapper. Great for prototyping equities/ETFs; adjusted OHLCV. Unofficial and rate-limited, never build production on it, and beware silent gaps and revisions.
Best for: Learning, quick equity/ETF prototypes
FRED (via fredapi / pandas-datareader)↗
RESTFreeThe Federal Reserve's economic database, rates, CPI, spreads, macro series. Authoritative, versioned (ALFRED gives point-in-time), and genuinely free with an API key.
Best for: Macro and rates data, point-in-time series
Alpha Vantage↗
RESTFreeFree tier (with a key) covering equities, FX, crypto, and some fundamentals. Tight rate limits (a few calls/min); fine for end-of-day research, not intraday scale.
Best for: EOD equities/FX with fundamentals
Tiingo↗
RESTFreeClean adjusted EOD equity data and news with a generous free tier; better data hygiene than yfinance. A common first paid upgrade.
Best for: Reliable adjusted EOD equities
Nasdaq Data Link (Quandl)↗
RESTFreeAggregator of free and paid datasets (e.g., Sharadar fundamentals as a paid add-on). Good for point-in-time fundamentals when you outgrow scraping.
Best for: Curated fundamentals & alt datasets
ccxt↗
PythonFreeUnified API across 100+ crypto exchanges (Binance, Coinbase, Kraken…) for OHLCV, order books, and trading. The de facto standard for crypto data and execution.
Best for: Crypto data and trading
Backtesting libraries
Vectorized libraries are fast for cross-sectional/EOD research; event-driven ones model fills and portfolio state realistically. Learn by writing one yourself first.
Roll your own (pandas/NumPy)
PythonFreeFor a single vectorized signal, a from-scratch backtest is ~30 lines and teaches you exactly where lookahead hides. Start here before any framework, see the Getting Started series.
Best for: Learning; simple vectorized signals
vectorbt↗
PythonFreeExtremely fast vectorized backtesting/optimization on NumPy/Numba, sweep thousands of parameter sets in seconds. Steep learning curve; easy to overfit precisely because it's so fast.
Best for: Large parameter sweeps, fast research
backtrader↗
PythonFreeMature event-driven framework with brokers, live trading, and indicators. Realistic fills and portfolio accounting; slower and more verbose than vectorized tools. Community-maintained.
Best for: Event-driven strategies, paper/live
zipline-reloaded↗
PythonFreeThe community fork of Quantopian's zipline: point-in-time, pipeline API for cross-sectional factor research, calendar-aware. Heavier setup, but strong for equity factor work.
Best for: Cross-sectional equity factor research
bt / ffn↗
PythonFreeLightweight allocation-and-rebalancing backtesting (bt) plus performance analytics (ffn). Nice for portfolio-level strategies without event-driven complexity.
Best for: Portfolio allocation & rebalancing
Broker & execution APIs
Where paper trading becomes real money. Prioritize order-type support, fill quality, and a sandbox, not just low commissions.
Interactive Brokers (ib_insync / TWS API)↗
PythonFreeThe retail-to-pro standard: global markets, all asset classes, real order types, and a paper account. The API is powerful but fiddly; ib_insync makes it tolerable.
Best for: Serious multi-asset retail algo trading
Alpaca↗
RESTFreeCommission-free US equities and crypto with a clean REST/websocket API and a proper paper-trading sandbox. The easiest first live-trading integration; US-centric.
Best for: First live US-equity/crypto automation
ccxt (crypto exchanges)↗
PythonFreeSame unified library as the data section, routes live orders to Binance/Coinbase/Kraken/etc. 24/7 markets and easy sandboxes make crypto a common first live venue.
Best for: Automated crypto execution
OANDA / broker FX APIs↗
RESTFreeRetail FX with a documented REST API and practice accounts. Reasonable for systematic FX; mind the spread-based cost model and dealer conflicts.
Best for: Systematic retail FX
Ready to use these? The Getting Started series walks through setting up the environment, pulling data, and writing your first honest backtest with them.