Market data types

Every strategy is only as good as the data feeding it. Knowing the data types — and their limits — is part of the craft.

The main kinds

• OHLCV bars — Open, High, Low, Close, Volume aggregated over a timeframe (M1, H1, D1...). Compact and the most common input for backtests and most strategies.
• Tick data — every individual price update (bid/ask change). Huge in volume but essential for scalping, accurate spread modelling and realistic backtests.
• Order book / depth of market (DOM) — the resting buy and sell orders at each level. Used by microstructure and high-frequency strategies.
• Fundamentals — company financials (earnings, balance sheet), and for macro: rates, inflation, employment. Slow-moving, used for equity and macro models.
• Alternative data — news sentiment, satellite imagery, web traffic. Niche and expensive; rarely where a retail trader should start.

Bars are summaries — respect what they hide

A single H1 candle hides the path price took within the hour. Two very different intrabar journeys can produce the identical OHLC. This is exactly why bar-only backtests mislead on intrabar strategies, and why tick data matters for them.

Quality questions to always ask

• Source and adjustment — is it the same feed you will trade on? Are equities split/dividend adjusted?
• Timezone and session — when does the "day" start? Mismatched sessions silently corrupt indicators.
• Survivorship — does the historical universe include delisted names? (The bias from the backtesting chapter lives here.)
• Gaps and holidays — handled, or silently dropped?

Cheap or careless data is the most common hidden cause of a backtest that cannot be reproduced live.