How to Align Earnings Dates, Fiscal Quarters, Estimates, and Reported Results

An earnings event rarely exists as one clean database row. The reported period may end in April, the earnings announcement may arrive in May, the associated 8-K and 10-Q may be accepted at different times, and the analyst estimate may have been observed days or weeks earlier. Each record describes part of the same event, but each uses a different date for a different purpose.

Misalignment occurs when a pipeline treats those dates as interchangeable. A model may attach an estimate for the next quarter to the current result, assign financial data to the announcement date instead of the reporting period, or use information in a backtest before it was publicly available. A ticker-and-quarter string alone cannot prevent those errors.

Financial Modeling Prep provides structured company profiles, earnings records, analyst estimates, financial statements, and filing data that can support this integration. The pipeline still needs an explicit matching hierarchy, point-in-time controls, and metric-basis validation before those datasets become one reliable earnings record.

Key Takeaways

• No single field is a universal earnings-event key: A ticker, fiscal-year label, calendar quarter, or announcement date may help identify a candidate match, but none is sufficient across every dataset.
• Entity resolution comes first: Use a stable company identifier, such as CIK for an SEC reporting entity, while preserving the ticker as a time-dependent security identifier.
• Fiscal periods must remain separate from event dates: Store the fiscal year, fiscal period, period-end date, announcement date, filing acceptance time, and data-observation time independently.
• Estimate matching requires point-in-time evidence: A current consensus response should not be described as the pre-announcement estimate unless the record was actually observed before the announcement or the dataset explicitly provides historical vintages.
• GAAP and non-GAAP EPS are different metrics: An estimated EPS value and an actual EPS value should not be compared until the accounting basis, dilution basis, currency, and period are aligned.
• Source documents and standardized data serve different roles: Earnings releases and filings establish the disclosed event, while standardized statement APIs make the results easier to integrate across periods and companies.

Why Earnings Records Misalign Across Financial Datasets

Every earnings workflow contains several related but distinct records. The financial statements describe a reporting period. The earnings calendar describes an announcement event. Analyst estimates describe expectations for a target period. Regulatory filings document the company's disclosures, while an internal ingestion timestamp establishes when a particular system received the data.

Collapsing those records into one generic quarter field removes information the model needs later. For example, NVIDIA's first quarter of fiscal 2027 ended on April 26, 2026, even though the company announced the results on May 20. Assigning the reported figures to a calendar second-quarter bucket based on the announcement date would change the period being analyzed.

The correct structure preserves both the accounting period and the information-availability timeline.

Record Component	Date or Identifier to Preserve	What It Establishes	Common Error
Company identity	CIK or another stable entity identifier	Which reporting entity produced the data	Joining only on a ticker that may change
Fiscal period	Fiscal year, fiscal period, period end	Which operating period the result measures	Replacing the issuer's period with a calendar quarter
Earnings event	Announcement date and available event-time metadata	When the result was announced	Treating the period end as the announcement date
Estimate	Target period and observation time	What analysts expected for that period at a particular time	Using a later consensus as the pre-event estimate
Filing	Filing type, accession, filing date, acceptance time	Which regulatory document supports the record	Treating every filing timestamp as the announcement time
Internal observation	observed_at or ingestion time	When the system received a particular version	Reconstructing historical knowledge from a current response

Resolve the Company Before Matching the Period

A financial-data pipeline should establish the reporting entity before attempting to align quarters. FMP's Company Profile Data API provides company and security information that includes global identifiers such as CIK, ISIN, and CUSIP alongside the current ticker and company details.

For SEC reporting companies, CIK can provide a more stable connection between the company profile, regulatory filings, and financial statements than a ticker alone. The ticker should still be stored because it is the identifier used by many market and earnings endpoints, but it should be treated as an alias with an effective period rather than a permanent company identity.

CIK is not a universal solution for every international issuer, subsidiary, or security. A production system may need its own entity master that distinguishes the reporting company from individual share classes and maps each identifier over time.

A practical entity record can preserve:

• Internal entity ID
• CIK, where applicable
• Current ticker
• Ticker effective dates
• Exchange
• ISIN or CUSIP, where relevant
• Source and observation time

This makes Company Profile useful as the entity-resolution layer without asking it to supply fiscal-period details that belong to the financial statements.

Preserve Issuer-Reported Fiscal Periods

Companies do not all report on calendar quarters. Some use fiscal years that end in January, others use 52- or 53-week calendars, and some reporting periods contain a different number of weeks than the comparable calendar period.

FMP's standardized Income Statement, Balance Sheet Statement, and Cash Flow Statement APIs provide structured records for reported financial periods. Depending on the response, useful alignment fields can include the symbol, CIK, period-end date, fiscal year, fiscal period, filing date, and filing acceptance time.

These statement records are the standardized alignment layer. They should not be called the sole statutory authority for the reporting period. The associated 10-Q, 10-K, or other source filing remains the document used to validate the period and reported disclosures.

A fiscal-period record should preserve at least:

• entity_id
• fiscal_year
• fiscal_period
• period_end
• period_type
• reported_currency
• filing_type
• filing_accession
• filing_date
• filing_accepted_at
• observed_at

If the workflow also needs calendar-quarter comparisons, add a separate derived calendar field. Do not overwrite the issuer's fiscal-year and period labels.

A Candidate Key Is Not a Universal Join Key

A string such as NVDA-2027-Q1 can be useful as an internal label after the underlying records have been validated. It is not an exact join key across every FMP endpoint.

The earnings or estimates record may not contain the same fiscal-year and fiscal-period fields as the statement record. Symbols can change, annual and quarterly records can share the same fiscal year, and different companies can use similar quarter labels for periods ending months apart.

A safer internal grouping key combines the stable entity, issuer period, and period end:

entity_id | fiscal_year | fiscal_period | period_end

For NVIDIA's first quarter of fiscal 2027, an internal grouping key could be:

0001045810 | FY2027 | Q1 | 2026-04-26

This key groups records that belong to the same reporting period. It does not replace the original source keys. A separate version key should include the source, filing accession or record identifier, and observation time so amended or refreshed records do not overwrite earlier versions.

Use a Matching Hierarchy

The pipeline should evaluate candidate records in a defined order instead of forcing a match from one convenient field.

Matching Level	Required Test	Reject the Candidate When
1. Entity	Stable entity identifier matches, or the symbol is valid for the relevant date	The records belong to different reporting entities or securities
2. Period type	Both records are quarterly, annual, or otherwise comparable	One record is annual and the other is quarterly
3. Period end	Target period end matches exactly or according to a documented issuer calendar rule	The candidate points to a different reporting period
4. Fiscal labels	Fiscal year and fiscal period agree where both datasets provide them	The labels conflict after accounting for the issuer's calendar
5. Event timing	Announcement follows the period end within a plausible reporting window	The event clearly belongs to another period
6. Source confirmation	Filing, release, or issuer disclosure confirms the period	The source document identifies a different quarter or year
7. Metric basis	Estimate and actual use the same GAAP or non-GAAP, basic or diluted basis	The basis is different or cannot be established

Entity, period type, and target period should be treated as hard constraints. Announcement timing can help rank candidates, but proximity alone should not override a period mismatch.

Keep the Earnings Event's Clocks Separate

An earnings-event model normally needs at least four different clocks:

1. Period end: The final date of the operating period measured by the financial statements.
2. Announcement time: When the company publicly released the earnings information.
3. Filing acceptance time: When a particular regulatory filing was accepted.
4. Observation time: When the data pipeline retrieved and stored the record.

A fifth field may be needed for the scheduled earnings date. The scheduled date can change and should not automatically replace the actual announcement date.

The Earnings Calendar API provides upcoming and past earnings-event data, including announcement dates and estimated and actual EPS where available. Filing endpoints, including the Latest 8-K SEC Filings API, connect the event to regulatory records and underlying source documents.

Filing acceptance is not always the same as the first public announcement. A company may publish a release, furnish it as an 8-K exhibit, and submit its 10-Q at different times. Store those events independently and document the timezone before classifying an event as pre-market, intraday, or after-market.

Worked Example: NVIDIA Q1 Fiscal 2027

NVIDIA reported first-quarter fiscal 2027 results on May 20, 2026, for the quarter ended April 26, 2026. The company's earnings release filed as an 8-K exhibit reported revenue of $81.615 billion, GAAP diluted EPS of $2.39, and non-GAAP diluted EPS of $1.87.

The standardized cash flow statement record supplied for the example contained the following period metadata:

{

"date": "2026-04-26",

"symbol": "NVDA",

"reportedCurrency": "USD",

"cik": "0001045810",

"filingDate": "2026-05-20",

"acceptedDate": "2026-05-20 16:35:52",

"fiscalYear": "2027",

"period": "Q1"

}

This response establishes a candidate statement record for NVIDIA's first quarter of fiscal 2027. Its acceptedDate corresponds to the 10-Q acceptance time, not the earliest earnings-event filing.

The SEC filing records show two separate documents:

• NVIDIA's earnings 8-K was accepted on May 20, 2026, at 16:21:19.
• NVIDIA's quarterly 10-Q was accepted on May 20, 2026, at 16:35:52.

The 8-K acceptance preceded the 10-Q acceptance by approximately 15 minutes. Using the statement record's acceptedDate as the sole earnings-announcement timestamp would therefore attach the event to the wrong filing.

The Estimate Record That Must Be Rejected

The Financial Estimates response supplied in the original example included this record:

{

"symbol": "NVDA",

"date": "2027-01-25",

"estimatedRevenueLow": 129074989800,

"estimatedRevenueHigh": 131001033000,

"estimatedRevenueAvg": 129426091400

}

This estimate targets a period ending January 25, 2027. NVIDIA's first quarter of fiscal 2027 ended on April 26, 2026. The shared ticker and fiscal year are not enough to overcome that period-end mismatch.

Test	Q1 FY2027 Statement	Estimate Candidate	Result
Entity	NVIDIA / NVDA	NVIDIA / NVDA	Pass
Fiscal year	FY2027	Falls within FY2027	Candidate only
Target period end	April 26, 2026	January 25, 2027	Fail
Metric	Reported quarterly results	Forward revenue estimate	Compatible only after period match
Final decision	Q1 FY2027	Different future period	Reject

This is a useful rejected match because it demonstrates why symbol + fiscalYear or symbol + fiscalYear + assumed quarter cannot serve as a universal join rule.

The Validated Temporal Record

The records that can be aligned with confidence produce the following event structure:

Field	Validated Value	Source Role	Status
Entity CIK	0001045810	Company and filing identity	Pass
Symbol	NVDA	Security identifier at observation	Pass
Fiscal year	2027	Issuer-reported fiscal year	Pass
Fiscal period	Q1	Issuer-reported quarter	Pass
Period end	2026-04-26	Statement and 10-Q period	Pass
Announcement date	2026-05-20	Earnings event and company release	Pass
Earnings 8-K acceptance	2026-05-20 16:21:19	Earnings release filing event	Pass
10-Q acceptance	2026-05-20 16:35:52	Quarterly filing event	Pass
GAAP diluted EPS	$2.39	Issuer earnings release	Pass
Non-GAAP diluted EPS	$1.87	Issuer earnings release	Pass
FMP event-record EPS actual	$1.87	Structured earnings result	Matches issuer non-GAAP value
FMP event-record EPS estimate	$1.76	Structured consensus input	Basis validation required
January 25 estimate record	Different target period	Financial Estimates response	Rejected

The period and event can be joined. The EPS surprise should remain uncalculated until the basis of the $1.76 estimate is confirmed as comparable with the $1.87 non-GAAP diluted actual.

A normalized record can express that status explicitly:

{

"record_id": "0001045810|FY2027|Q1|2026-04-26",

"entity": {

"cik": "0001045810",

"symbol": "NVDA"

},

"period": {

"fiscal_year": 2027,

"fiscal_period": "Q1",

"period_end": "2026-04-26"

},

"event": {

"announcement_date": "2026-05-20",

"earnings_8k_accepted_at": "2026-05-20 16:21:19",

"form_10q_accepted_at": "2026-05-20 16:35:52"

},

"eps": {

"gaap_diluted_actual": 2.39,

"non_gaap_diluted_actual": 1.87,

"consensus_estimate_candidate": 1.76,

"estimate_basis": "validation_required",

"surprise_percent": null

},

"validation": {

"entity_match": "pass",

"period_match": "pass",

"event_match": "pass",

"estimate_basis_match": "pending"

}

}

This is safer than producing a precise surprise percentage from two numbers whose comparability has not been established.

Match Estimates to Their Target Period

The Financial Estimates API provides analyst forecasts for revenue, EPS, and other financial metrics. A pipeline should preserve the requested period type, the returned target date, the symbol or entity mapping, and the time at which the response was observed.

The public documentation describes consensus forecast data but does not establish that every response represents a historical as-of vintage. Unless the response, product documentation, or data agreement explicitly supplies historical consensus vintages, the safest approach is to treat the result as a current-state observation.

A point-in-time estimate table should therefore store:

• Target entity
• Quarterly or annual period type
• Target period end
• Metric name
• Estimate value
• Metric basis
• Number of contributing observations, if supplied
• Source
• observed_at
• Valid-from and valid-to timestamps, if the system versions records

A consensus estimate can be labeled “pre-announcement” only when its observation time precedes the earnings announcement. Pulling the endpoint after the result and assigning the returned value to an earlier date can introduce look-ahead bias because the consensus may already reflect post-event revisions.

Calculate Surprise Only After Validating the Metric Basis

FMP's Earnings Surprises Bulk API provides actual and estimated EPS values for bulk analysis. The endpoint supplies the inputs for earnings-surprise analysis, but the user remains responsible for validating the records and calculating any custom difference or percentage.

A common calculation is:

EPS surprise = Actual EPS - Estimated EPS

EPS surprise percentage =

(Actual EPS - Estimated EPS) / |Estimated EPS| × 100

If the estimate is zero, the percentage is undefined and the workflow should use the absolute difference or another documented rule. The calculation should also be withheld when the estimate and actual use different definitions.

Before calculating a surprise, validate:

• The company and security
• The target reporting period
• Quarterly versus annual periodicity
• GAAP versus non-GAAP basis
• Basic versus diluted EPS
• Continuing operations versus total reported earnings
• Currency and per-share units
• Split adjustment
• Estimate observation time
• Actual-result source

In the NVIDIA example, the issuer reported both $2.39 in GAAP diluted EPS and $1.87 in non-GAAP diluted EPS. Because the structured event record contains an actual EPS of $1.87, the value aligns numerically with the non-GAAP result. The estimate must use that same non-GAAP diluted basis before the two values can support a valid surprise calculation.

Once individual event records have been aligned, teams can use the same controls when tracking historical earnings surprises and beat-or-miss patterns. This article owns the record-matching problem, while the historical-surprise workflow owns analysis across many completed events.

Use Each Disclosure for the Role It Actually Serves

An earnings release, 8-K, 10-Q, and standardized financial statement are related sources, but they are not interchangeable.

Earnings Release

The earnings release provides the company's initial presentation of results and may include both GAAP and non-GAAP metrics, management commentary, guidance, and reconciliations. It can be the most direct source for determining what the company announced, but adjusted metrics must remain clearly labeled.

Form 8-K and Exhibits

An earnings release is often furnished as an exhibit to an 8-K. The filing establishes a regulatory record, acceptance time, accession number, and link to the disclosed materials. It does not mean every figure in the exhibit is audited or that the filing acceptance time is necessarily the first instant the information became public.

Form 10-Q or 10-K

The 10-Q or 10-K provides the period's financial statements, notes, risk disclosures, and other required information. Quarterly 10-Q statements are generally unaudited, so the filing should not be described as automatically converting every number into an audited result.

Standardized Financial Statements

FMP's statement APIs normalize reported financial information into consistent fields that are easier to retrieve and compare. These records support integration, screening, and modeling, while the filing remains available for source-level validation.

This source hierarchy should not be framed as management narrative versus verified truth. Each record answers a different question. The workflow becomes reliable when it preserves those roles and reconciles high-impact differences rather than assuming one source makes every other source unnecessary.

Broader systems that need to connect guidance, transcripts, and actual reported results can extend this structure into a cross-dataset earnings-data architecture. This article remains focused on the narrower alignment problem involving identity, periods, estimates, actuals, and filings.

FMP Endpoint Map for Earnings Alignment

FMP Dataset	Primary Role	Useful Alignment Information	Important Limitation
Company Profile Data	Resolve company and security identity	Symbol, CIK, ISIN, CUSIP, exchange, company details	Current profile metadata does not establish the target financial period
Income Statement	Retrieve standardized performance data	Period end, fiscal labels, filing metadata, revenue, income, EPS fields	Metric definitions and source periods still require validation
Balance Sheet Statement	Retrieve standardized financial-position data	Period end, cash, debt, assets, liabilities, equity	A balance sheet is a point-in-time statement and should not be treated like a period-flow statement
Cash Flow Statement	Retrieve standardized cash-flow data	Period end, fiscal labels, filing metadata, operating and investing cash flows	It is an alignment layer, not the sole authority for the statutory period
Earnings Calendar	Identify earnings events	Announcement date, estimated EPS, actual EPS where available	The event date should not be substituted for the statement period end
Financial Estimates	Retrieve forward-looking consensus estimates	Target date and estimated financial metrics	Do not assume a current response represents a historical pre-event vintage
Earnings Surprises Bulk	Retrieve actual and estimated EPS across companies	Actual and estimated EPS inputs	Validate basis and calculate custom surprise measures after alignment
Latest 8-K SEC Filings	Identify recent material-event filings	Filing records and links to source documents	Filing acceptance does not automatically equal announcement time
SEC Filings by Symbol or CIK	Retrieve filing history for the company	Filing type, company identity, filing links, dates	The underlying filing still needs to be reviewed for period and metric context

A Practical Earnings-Alignment Workflow

1. Create the Entity Record

Query Company Profile and store the company's stable identifiers, current symbol, exchange, and observation time. Maintain effective dates for ticker changes when the workflow covers long histories.

2. Retrieve the Reported Period

Query the quarterly or annual statement endpoints and preserve the issuer's fiscal year, fiscal period, period end, currency, filing date, and acceptance time. Confirm the reporting period against the source filing for high-impact workflows.

3. Identify the Earnings Event

Retrieve the earnings-calendar record for the company and search within a plausible window after the period end. Treat the event date as a candidate until the company release or filing confirms which period was announced.

4. Retrieve Filing Records

Identify the relevant 8-K, 10-Q, or 10-K and preserve each filing's type, accession number, filing date, acceptance time, and source link. Do not collapse several filings into one generic timestamp.

5. Retrieve Estimate Candidates

Request the correct quarterly or annual Financial Estimates records. Compare their target dates with the statement period end and reject records for earlier or later reporting periods.

6. Preserve the Estimate Vintage

Store the estimate response with an observed_at timestamp. If the workflow requires pre-announcement consensus, confirm that the observation occurred before the event or use a dataset that explicitly supplies the necessary historical vintage.

7. Validate Metric Comparability

Confirm that estimated and actual EPS use the same GAAP or non-GAAP, basic or diluted, currency, and split-adjustment basis. Store GAAP and non-GAAP results as separate metrics rather than overwriting one with the other.

8. Create the Normalized Event Record

Group the validated records using the entity, fiscal year, fiscal period, and period end. Preserve the source records, version information, validation results, and any unresolved fields.

9. Quarantine Failed Matches

A failed candidate should remain visible with a rejection reason. This supports auditing and prevents the same incorrect match from being reintroduced during a later refresh.

These controls can be incorporated into a broader set of pre-model financial-data validation checks, particularly when the output will feed valuation, forecasting, or recurring reporting systems.

Common Alignment Failures

Failure	Why It Breaks the Record	Required Control
Joining only on ticker	Tickers can change and may represent different securities	Maintain a stable entity identifier and dated symbol mapping
Joining on announcement quarter	The announcement can occur in a later calendar quarter than the reported period	Preserve the issuer's period end and fiscal labels
Using symbol + fiscalYear + period universally	Not every endpoint supplies or defines those fields consistently	Use a matching hierarchy and validated internal grouping key
Treating statement acceptedDate as the announcement time	The timestamp may belong to the 10-Q rather than the earnings 8-K or release	Store each document's acceptance time separately
Calling a second-level timestamp a millisecond timestamp	The claimed precision exceeds the supplied value	Describe and store only the precision actually provided
Matching the January 2027 estimate to April 2026 results	The records target different fiscal periods	Require target-period validation
Using a current estimate as historical consensus	The estimate may contain post-event revisions	Preserve point-in-time snapshots
Comparing GAAP EPS with non-GAAP EPS	The estimate and actual measure different definitions of earnings	Add a required metric-basis field
Assuming a 10-Q is audited	Quarterly statements are not equivalent to audited annual statements	Label filing and assurance status accurately
Automatically calculating a surprise	A numerical pair can still contain period or basis mismatches	Calculate only after validation passes

Validation Checklist

Before releasing an aligned earnings record to a model, confirm that:

• The reporting entity is resolved with a stable identifier.
• The ticker was valid on the relevant date.
• Quarterly and annual records are not mixed.
• The fiscal year and fiscal period remain issuer-reported.
• The exact target period end matches.
• The earnings event is connected to the correct reported period.
• Announcement, filing, and observation times are stored separately.
• The timestamp timezone and precision are documented.
• The estimate was observed before the announcement when labeled pre-event.
• GAAP and non-GAAP metrics remain separate.
• Estimated and actual EPS use the same basis.
• Filing accessions and source links are preserved.
• Mismatched candidates are rejected with a reason.
• Amended or refreshed records create new versions instead of silently replacing history.

Reliable Earnings Data Requires an Explicit Time Model

Earnings alignment is not solved by finding one perfect quarter field. It requires a structured relationship among the reporting entity, issuer-defined fiscal period, earnings event, estimate vintage, actual metric, and source filing.

FMP supplies the structured data layers needed to build that relationship. Company Profile supports entity resolution, financial statements establish standardized periods and results, Earnings Calendar identifies events, Financial Estimates provides forward expectations, filing endpoints connect the records to source disclosures, and Earnings Surprises Bulk supports broader actual-versus-estimate analysis.

The system using those datasets must preserve the dates, identifiers, metric definitions, and validation decisions that make the final join defensible. Researchers and developers can begin with a small company set in the FMP API documentation, inspect the actual endpoint responses available to their account, and expand only after the matching and rejection rules work across several fiscal-calendar structures.

Frequently Asked Questions

Why do fiscal quarters differ from calendar quarters?

Companies define fiscal calendars around their reporting and operating cycles. A fiscal quarter can therefore end in a different month from the corresponding calendar quarter, and 52- or 53-week calendars can shift period-end dates across years. Pipelines should preserve issuer-reported fiscal fields and derive calendar classifications separately.

What is the difference between the period-end date and earnings date?

The period-end date closes the accounting period measured by the financial statements. The earnings date identifies when the company announces information about that completed period. A model can associate the result with the accounting period while restricting its historical availability until the announcement or relevant disclosure.

Why is a ticker, fiscal year, and quarter not always enough?

Not every dataset provides the same fiscal labels, and the same fiscal year contains several quarterly and annual targets. Tickers can also change. A defensible match uses entity identity, period type, exact period end, event timing, and source confirmation.

How should an earnings estimate be stored?

Store the target entity, period type, target period end, metric, value, basis, source, and observation time. If the workflow needs historical consensus, retain repeated snapshots or use a source that explicitly supports point-in-time estimate vintages.

Can GAAP EPS be compared with a non-GAAP estimate?

No. GAAP and non-GAAP EPS can differ because they include different adjustments. The estimate and actual should use the same definition, dilution basis, currency, and period before a surprise is calculated.

Does a filing acceptance timestamp show when the market first knew the result?

Not necessarily. It shows when that filing was accepted by the regulatory system. A press release or another disclosure channel may have made the information public earlier, so the workflow should preserve each available timestamp and its source.

Does the Earnings Surprises Bulk API return a complete calculated surprise ratio?

The endpoint provides actual and estimated EPS inputs for earnings-surprise analysis. Users can calculate an absolute difference or percentage after confirming that the company, period, metric basis, and estimate vintage are comparable.

Which FMP APIs are most useful for earnings alignment?

The core set includes Company Profile Data, Earnings Calendar, Financial Estimates, Earnings Surprises Bulk, Income Statement, Balance Sheet Statement, Cash Flow Statement, Latest 8-K SEC Filings, and SEC filing search by symbol or CIK. The exact combination depends on whether the workflow needs event monitoring, point-in-time estimates, standardized fundamentals, filing validation, or large-scale historical analysis.