Use Massive equity data for charts and research

What Massive provides#

Massive, formerly Polygon, is AMM's US-equity market-data connection. It provides historical acquisition and live or delayed market data; it does not execute orders. Select a real execution broker or Simulation independently. The public venue identifier is massive; symbols can appear as AAPL.Massive while the provider ticker remains AAPL.

A registered connection or generic capability flag does not turn Massive into a brokerage account. Its order methods are unavailable. The current covered asset family is US equities; the presence of a general-purpose configuration field does not establish futures, options or crypto support.

Open Data › Brokers & feeds › Massive. Inspect key status, key source, REST rate, live transport, WebSocket entitlement, delayed settings and flat-file settings. Enter a new API key only in the write-only credential field. Save changes to the server, then check the returned status. The Usage view helps identify consumption and throttling; a key-file status of Found establishes file presence, not paid data entitlement.

Choose transport and timing#

AMM supports REST polling and WebSocket transport for implemented live data. REST is the default transport; WebSocket must be selected explicitly and requires entitlement. Delayed WebSocket mode is a separate option. Inspect the chosen feed's actual timing rather than interpreting Connected as Real time.

Delayed Massive observations cannot drive broker-backed live execution. For a safe research workflow, pair them with Simulation. If you intend to trade against live market prices, obtain the required real-time entitlement and verify the effective source. This timing gate applies even when historical data is complete and the broker account is authenticated.

Flat files are a separate entitlement and credential path. Turn on flat-file use only when the account has access. Available REST, WebSocket and S3 credentials are not interchangeable. Provider plans and exchange agreements can change; use the provider's current account portal for entitlement terms rather than assuming an old screenshot describes your subscription.

Download a bounded research sample#

1. Open Data Download and select Massive.
2. Choose an exact equity symbol or reviewed symbol group and the required resolution.
3. Set a bounded historical range and review adjustment settings.
4. Start the download and open its job details. Track per-symbol/resolution completion, counts and errors.
5. Inspect the resulting Data Library entry, first/last timestamps and coverage before starting a backtest.
6. Open a chart from that same data source and verify the price scale, timestamps and selected adjustment basis.

Massive acquires daily/minute aggregate bars and rich trade ticks; supported lower-resolution sourcing also serves consolidated chart periods. Data received as JSON or flat-file CSV is converted to AMM's .ammbin storage. Backtests and optimizations read those local archives; they do not fetch missing history from Massive while the strategy is running.

A completed job can contain individual failures. A file's first and last record do not prove that every intervening market interval is complete. Review coverage and query evidence, especially around splits, ticker changes, quiet extended hours and sparse symbols.

Corporate actions and price basis#

AMM keeps a canonical raw dataset and applies the requested split/dividend adjustment on read. Choose the same historical adjustment basis for charts, algorithms and comparisons when a price-level comparison matters. A split-adjusted chart and an unadjusted broker quote can have different historical levels without either being corrupt.

Volumes retain exact supported fractional quantities. A reverse split can change the required quantity scale; AMM preserves that precision or fails explicitly if it cannot represent the requested values. Do not convert fractional volume into whole-share integers in an external export script.

The rich trade-tick archive retains conditions, exchange and timing metadata used to decide which prints update OHLC and volume. A volume-only odd-lot print can add volume without changing the candle's high, low or close. Therefore a count of raw prints and a count of price-changing events need not match. Use the official aggregate comparison and the recorded eligibility rules when investigating bar differences.

Combine Massive with another venue#

A stock workflow might select Massive history and a separate live feed/execution account. Check the exact stock, currency, share unit, adjustment basis and data delay for each role. Keep the stable strategy symbol and provider aliases explicit. AMM does not automatically rescale prices or infer that two same-named tickers always represent the same underlying across dates.

For index research, point-in-time constituent membership is a separate input from today's ticker list. Downloading current constituents alone does not remove survivorship bias. For factor and liquidity analysis, preserve the analysis date, source coverage and corporate-action basis with exported results.

Recover from failures#

Authentication errors need a valid key; entitlement errors need the matching provider permission. Repeated throttling calls for checking the configured request budget and other consumers. A flat-file access error is not repaired by enabling WebSocket. If history is missing, retry the specific failed symbol/range and retain the existing archive; do not replace it with another provider's file under the Massive identity.

See Historical data for coverage and downloads, Brokers and feeds for role routing, and Simulation and replay for using data without broker execution.