Tick Data

Tick data is the most granular form of market data available — a complete record of every trade execution and quote update on an exchange, timestamped to the microsecond or nanosecond. Each tick contains the price, size, and timestamp of a single transaction. For a liquid stock like Apple or Tesla during market hours, thousands of ticks are generated every second. Tick data represents the raw, unprocessed stream of market activity before it is aggregated into OHLCV bars. Understanding tick data helps traders grasp how prices are formed and how order flow creates short-term price movements. This is educational content, not financial advice.

OHLCV bars aggregate all tick activity within a time period into five values: open, high, low, close, and volume. A 1-minute bar condenses potentially thousands of individual ticks into a single data point. This aggregation is efficient for storage and sufficient for most retail trading strategies. Tick data preserves information that bars lose: the exact sequence of trades, the distribution of trade sizes within a period, the behavior of the bid-ask spread, and the timing of price jumps. Tick imbalance — the ratio of buyer-initiated to seller-initiated trades — is only measurable from tick data and is a powerful short-term price predictor used by high-frequency traders.

Tick data enables several forms of analysis unavailable with bar data. Volume profile at the tick level reveals exact price zones where the most transactions occurred, producing more precise support and resistance levels than volume-by-bar analysis. Tick charts — which plot a new bar after N ticks rather than after N minutes — normalize price bars by activity rather than time, making them useful during volatile periods when many ticks are generated quickly. Order flow analysis at the tick level tracks the balance between aggressive buyers (hitting the ask) and aggressive sellers (hitting the bid), providing a real-time read on supply and demand pressure. Market microstructure research relies entirely on tick data to study how prices respond to information.

The scale of tick data is staggering: US equity markets generate roughly 50-100 billion ticks per year. Storing and querying tick data efficiently requires columnar databases (ClickHouse, Arctic, kdb+) that compress time-series data far more efficiently than row-oriented databases. For most retail algorithmic traders, the cost and complexity of managing full tick data is not justified. Providers like Polygon.io, Databento, and Refinitiv Tick History offer tick data subscriptions. Python libraries like pandas, Arctic, and the Xarray ecosystem provide tools for working with large time-series datasets. Cloud storage in Parquet format on AWS S3 or Google Cloud Storage is a common cost-effective approach for historical tick archives.

Real-time tick data feeds are essential for high-frequency trading, options market making, and any strategy that needs to react to individual trades rather than periodic price updates. WebSocket connections deliver tick data with sub-100ms latency from major data providers. In production trading systems, tick data is typically processed through a streaming pipeline (Kafka, Redis Streams, or direct WebSocket consumption) that computes aggregated statistics (VWAP, running volume, bid-ask spread) in real time without storing every individual tick. Tradewink consumes real-time trade ticks from the Alpaca WebSocket stream specifically for position management — updating Maximum Favorable Excursion (MFE) and Maximum Adverse Excursion (MAE) on every trade execution to produce accurate trade analytics.