How Fresh Is FMP Data? Understanding Update Timing Across Prices, Filings, Calendars, and Fundamentals.

Financial data does not update on one universal clock. A stock quote may change during the trading session, an SEC filing appears only after an issuer submits it, and financial statements reflect new reports or revisions. The right freshness standard therefore depends on the dataset and the decision the data supports.

This guide explains how to interpret FMP update timing across market prices, financial statements, filings, calendars, macroeconomic releases, transcripts, and company reference data. It helps analysts and developers determine whether each dataset is current enough for a specific workflow, including the spreadsheets and applications that use those inputs.

Key Takeaways

• Data freshness should be evaluated by dataset and use case, rather than through one platform-wide label.
• Real-time market prices can support intraday monitoring, while many research, valuation, and reporting workflows can use completed daily observations.
• Filing-driven datasets depend on source events; they do not become fresher simply because an application polls them more often.
• Record relevant source dates, your application's retrieval time, and internal processing time separately. Each describes a different part of the data path.

Freshness Is a Dataset Property

“How fresh is the data?” sounds like a single question, but it contains several different questions:

• How often can the underlying value change?
• When does the source publish or revise it?
• How frequently does FMP refresh the relevant endpoint?
• When did your application last request and store the response?

Those clocks are not interchangeable. A market price can move throughout an active session, while quarterly financial results depend on company disclosures. An expected earnings date may change before the event; actual results become available after they are announced.

FMP's Stable API Cycle Times page publishes endpoint-specific timing labels. Some endpoints are listed as real-time, others use intervals such as minutes or hours, and some are tied to filing dates, daily schedules, weekly schedules, or ad hoc updates.

Each endpoint should be matched to the rate at which its source information changes and the consequences of using an older observation. A recent API request establishes when you checked the dataset. It does not establish when the source published the information or when FMP first made that particular record available.

How to Read FMP Cycle-Time Labels

The Stable cycle-times page uses several types of update descriptions. The practical interpretations below help you apply those labels without assuming an availability guarantee.

Cycle-time label	Practical interpretation
Real-Time	Check the relevant observation timestamp and endpoint documentation. The label alone does not specify end-to-end delivery latency.
Minutes or Hours	Use the stated recurring cycle as refresh guidance, then check whether the expected record is present.
Daily or Weekly	Match routine requests to the documented cadence and your workflow's tolerance for older observations.
Filing Date	Treat filing events as a reason to check for updated records. The label does not specify a maximum processing delay.
Ad-Hoc	Do not infer a fixed interval or an automatic change-detection mechanism.
N/A	A standard recurring cycle is not stated on the page. Confirm the behavior relevant to your use case.

These labels should be treated as dataset guidance, not as a promise that every symbol, exchange, filing, or event will arrive at an identical second. Before deploying a production workflow, inspect the endpoint documentation, test representative symbols, and confirm that the required dataset is included in your plan.

Use the cycle-times page for the API version you call. Legacy and Stable documentation should not be treated as interchangeable.

Market Prices Follow the Trading Use Case

FMP lists stock quote endpoints and intraday stock-chart endpoints as Real-Time on its Stable cycle-times page. Stock Price and Volume Data is listed on a 5-15 minute cycle. These categories serve different purposes even when they describe the same security.

A live quote widget or intraday alert may require current during-session observations. A valuation model refreshed after the close or a daily research dashboard can often use completed end-of-day observations instead.

The Stock Price and Volume Data API provides daily open, high, low, close, volume, change, percentage change, and VWAP fields. Before treating its latest row as a completed closing observation, confirm that the relevant session has ended and that the record is suitable for your cutoff.

Chart interval and update cycle are also different concepts. A one-minute interval describes the observations in a series; it does not establish how quickly each observation reaches your application.

Increasing the refresh rate of a daily model does not make its quarterly fundamentals more current, and a real-time price does not make a filing-based valuation input real-time. Decide whether your project needs live or completed-session prices before choosing the access and refresh approach.

Financial Statements Update Around Filing Events

Income statements, balance sheets, cash flow statements, TTM statements, and several related fundamental datasets carry the Filing Date label in the Stable documentation. That is fundamentally different from a quote feed.

Quarterly revenue does not need to refresh every minute between reports. A saved figure may change when another period is reported, a filing is amended, or standardized data is corrected. Polling an income statement every few seconds cannot create a new filing.

A fundamental-data workflow should preserve the period end, fiscal year, fiscal period, filing date, acceptance time, and reported currency where available. Record your own retrieval time as well. These fields identify the reporting period and relevant filing events; they do not establish when FMP first supplied that value or version.

A newly detected filing should trigger a check for the corresponding statement records. Confirm the expected period is present before replacing your existing inputs. Earnings results and standardized statements may become available separately, so a model may need a staged refresh after earnings.

Filing-driven timing also matters for backtests. A current response can describe an older accounting period while containing later revisions. Matching the period end to a historical decision date is therefore insufficient.

SEC Filings Have Their Own Availability Clock

FMP lists Latest SEC Filings and Latest 8-K SEC Filings on a one-hour cycle. Filing-search endpoints by symbol, CIK, form type, or company name are labeled N/A; that label does not establish a recurring refresh interval.

A filing date identifies the date associated with the submission. An acceptance timestamp provides more precise information about when the SEC accepted it. Neither should automatically be treated as the first moment the market learned the underlying news. A company may publish a press release, furnish it as an 8-K exhibit, and submit a 10-Q at different times.

For monitoring, choose a request cadence consistent with the documented cycle and your tolerance for delayed detection. For historical research, retain the filing identifier, acceptance time, retrieved values, and their connection to the source document. Those records support chronological checks, but timestamps alone do not reconstruct FMP's historical database or establish which version of every value was available at a past cutoff.

Earnings Calendars Are Dynamic Event Data

FMP lists Earnings Calendar and Earnings Report on a 1-2 hour cycle. Companies can confirm, reschedule, or refine an expected report date, and actual results become available after the event.

Applications should distinguish the fiscal period end, scheduled earnings date, actual announcement date, and filing acceptance time where those details are available. The event date should not be substituted for the financial statement's period-end date.

That structure allows a calendar to change without rewriting the underlying fiscal period. When combining datasets, validate the relationship between the records so an estimate for one quarter is not matched to another quarter's results. Preserving dates helps, but does not establish the match by itself.

Use an explicit process to align earnings dates, fiscal quarters, estimates, and reported results when the workflow requires that comparison.

Economic Data Combines Schedules and Releases

Macroeconomic data has two timing layers. The calendar describes when an indicator is expected to be released, while the indicator record contains the published value and may later be revised.

FMP lists the Economic Data Releases Calendar on a five-minute cycle, Economics Indicators on a four-hour cycle, and Treasury Rates on a two-hour cycle. The different schedules mean a calendar refresh should not be treated as proof that the corresponding indicator endpoint already contains a new actual value.

Preserve the event time, country, indicator, previous value, estimate, and actual value where available, alongside your retrieval time. Keep estimates distinct from actual observations and confirm timezone context before displaying event times.

Revision risk is also important. A macroeconomic release can be current when first published and still be revised later. Retaining snapshots rather than overwriting earlier observations helps document changes from the time your collection begins. It does not automatically supply historical vintages for dates before you started collecting.

Earnings Transcripts Arrive After the Call

Transcripts depend on the underlying call and subsequent publication. FMP lists Latest Earning Transcripts and Transcripts Dates By Symbol on a 4-6 hour cycle, while the Earnings Transcript retrieval endpoint is labeled N/A.

That discovery cycle should not be interpreted as a guarantee that every transcript arrives within four to six hours of a call. A transcript workflow should check for available content and separate the call date from the time your application first observed the transcript.

Sorting only by call date can place newly observed content behind records from more recent calls. For research assistants and summarization applications, the trigger should be the presence of the required transcript, rather than the scheduled start of the earnings call.

Store the retrieval time and avoid presenting a missing transcript as evidence that no call occurred. If you record a first-observed timestamp, label it as an application observation rather than the exact first publication time.

Company Profiles and Reference Data Change Differently

FMP lists Company Profile Data as Ad-Hoc. Reference information such as company name, industry, identifiers, address, and website does not share a single market-driven refresh interval.

However, the profile endpoint also includes market-related fields such as stock price and market capitalization. A cache suitable for a business description may be unsuitable for a price displayed as current. Set expectations according to the fields you use, and use an appropriate quote endpoint when current-price monitoring is the task.

A current profile is useful for company context and entity resolution, but it should not be assumed to reproduce every historical company attribute. Refresh reference information according to the changes your workflow needs to detect, and label the record as a current profile observation unless the endpoint explicitly supplies dated history.

Match the Refresh Schedule to the Decision

The same endpoint can be fresh enough for one workflow and too slow for another. The deciding factor is what happens if the value changes before the next refresh.

Workflow	What to check	Main risk to control
Intraday quote monitor	Observation timestamp, access, and delivery through your application	Displaying an older quote as current
Daily research dashboard	Session completion and the latest usable daily record	Mixing partial and completed sessions
Valuation model	Availability of new statement records and the model's price cutoff	Combining inputs with different periods or observation dates
Earnings watchlist	Updated event records throughout the relevant window	Treating a scheduled date as permanent
Filing alert	New records returned by the filing feed	Treating acceptance as the first public disclosure
Macro dashboard	Available actuals and subsequent revisions	Confusing estimates, actuals, and revised values
Transcript summarizer	Presence of the required transcript content	Assuming the call date establishes transcript availability
Company directory	Changes to the reference fields the application uses	Treating current metadata as historical truth

This framework aligns API usage and user expectations with the actual decision being made. A faster universal refresh job may still leave the application with the wrong period, an incomplete record, or an older cached value.

Make Freshness Visible in the Output

A production application should make freshness visible rather than hiding it behind one generic “last updated” label. Preserve relevant source dates, your application's retrieval time, and internal processing time where applicable. One timestamp cannot represent all three.

Consider a dashboard checked after an earnings announcement. The earnings record may contain new results while the standardized statement still represents the prior period. Show those statuses separately, and mark the expected statement as pending until you confirm its availability.

For each refresh:

1. Check the exact endpoint and its documented cycle.
2. Inspect the source dates, reporting period, and observation context available in the response.
3. Record when your application retrieved the data.
4. Confirm that the expected new record or period is present.
5. Preserve earlier retrieved versions when changes matter.
6. Check the cache and displayed result as well as the retrieval job.

Freshness describes timing, not automatic correctness or comparability. Validate symbol identity, exchange, timezone, fiscal period, currency, units, and source context before combining datasets.

FMP's cycle time is only one part of end-to-end freshness. Network delays, retry logic, scheduled jobs, caches, databases, and front-end rendering can make an application stale even when the upstream endpoint has already updated.

A useful user-facing label is specific: “Price as of 4:00 p.m. ET,” “Financials filed on October 30,” or “Calendar checked at 14:05 UTC.” Use each only when the underlying record supports that meaning. A page-render timestamp should not stand in for a price observation or source-publication time.

Choose Freshness That Fits the Workflow

The question is whether the relevant endpoint updates at the pace required by your application. Start with the input whose delay would matter most, check its documented cycle, and decide when an older observation should be flagged or withheld.

Test the complete path from retrieval to display before promising a freshness standard to users. Confirm that you can detect the expected update, explain pending inputs, and distinguish a new record from a revised one. Where the public documentation does not establish a required availability window, confirm that requirement before building a promise around it.

Frequently Asked Questions

Is All FMP Data Real-Time?

No. FMP provides multiple dataset types with different update cycles. Quotes and intraday price endpoints may be labeled Real-Time, while statements, calendars, filings, transcripts, macro data, and profiles follow their own documented cycles.

How Often Should I Refresh Financial Statements?

Check for updated statements around new filing events rather than every few seconds. The main standardized statement endpoints carry the Filing Date label, but it does not specify an immediate-availability guarantee. Confirm the expected period is present before updating your inputs.

Are Historical Daily Prices the Same as Live Quotes?

No. Daily price records and live quotes serve different purposes. If your analysis requires a completed trading session, check session completion and the record's suitability before using the latest daily row as the close.

Does a Filing Timestamp Equal the Earnings Announcement Time?

Not necessarily. A press release, an 8-K, and a 10-Q can appear at different times. Store announcement and filing events separately, and review the source document when exact sequencing matters. Filing metadata also does not establish when FMP first made a particular value available.

How Can I Show Users Whether Data Is Current?

Display labels that distinguish the measured period or source event from the time your application checked. Use source or price timestamps only when their meaning is established, show pending inputs separately, and monitor your own cache and processing delays.