Regressors
Moving-holiday exposure
Ramadan share, Eid al-Fitr, Eid al-Adha, Hajj and working-day controls.
Core asset 02 · schema + verified regime seed live
MENA-CAL
Open calendar-adjustment infrastructure for MENA economic data. The region’s economic behaviour is generated by Gregorian time and moving Hijri-calendar events simultaneously; the dataset is designed to make Ramadan, Eid, Hajj, working days and historically changing weekend regimes explicit model inputs.
Gregorian clock
365days in a common Gregorian year
Lunar clock
≈354days in a Hijri lunar year; moving events shift through Gregorian months over time
Country-date contract
The target is daily, country-specific, historically versioned infrastructure from 1990–2050. That target is not presented as complete today: the schema and first verified regime records are live; country-date expansion is the build.
working_dayHistorically applicable working-day flag.weekend_regimeVersioned working-week structure, including reforms.ramadanDaily Ramadan exposure and day index.eid_fitrCountry-specific coded Eid interval.eid_adhaCountry-specific coded Eid interval.hajj_periodHajj exposure where economically relevant.working_hours_regimeSpecial hours regimes where verified.fraction_ramadan_monthMonthly share usable directly as a regressor.Why historical versioning matters
A calendar variable can be wrong even when the holiday date is right if the working-week regime applied to that date is wrong.
A working-week reform changes the regressor itself.
The UAE federal working-week system moved from a Friday–Saturday weekend toward a Saturday–Sunday weekend with a Friday half-day from January 2022. MENA-CAL stores the regime as time-varying rather than pretending one weekend definition holds across history.
Official source →Official seasonal adjustment already recognizes moving religious-calendar effects.
GASTAT documents specialized treatment for Ramadan and Hajj in its seasonal-adjustment methodology. The research question is therefore not whether calendar effects exist, but how much moving-holiday structure remains across published series and statistical systems.
Official methodology →Residual Eid seasonality has been documented before.
An IMF study of Jordan reported residual seasonality associated with Eid al-Fitr and Eid al-Adha after ordinary Gregorian X-12 adjustment. MENA-CAL turns that issue into reusable regional infrastructure rather than a one-off adjustment.
IMF study →Empirical programme
Does Ramadan survive seasonal adjustment?
Take officially published seasonally adjusted monthly series and test whether statistically detectable Hijri-calendar structure remains after the official adjustment. The result is a residual-calendar audit, not an accusation that an agency “got seasonality wrong.”
Diagnostics
Lunar-frequency residuals
regARIMA, Fourier terms, dynamic harmonic regression, state-space effects and spectral diagnostics where appropriate.
Evaluation
Out-of-sample loss
Compare baseline forecasts with MENA-CAL-augmented models rather than declaring improvement from in-sample fit alone.
API contract
The commercial endpoint is deliberately simple: dates in, verified calendar exposures out.
GET /calendar?country=SAU&start=2026-01-01&end=2026-12-31
returns:
ramadan_exposure
eid_exposure
working_days
weekend_regime
hajj_exposure
country_holiday_exposure
regime_source
validation_state
Claim boundary. MENA-CAL currently publishes its schema and a verified regime seed. Full 1990–2050 country-date coverage, school holidays and private-sector exceptions remain build targets and will only be marked live after source verification.
Who can use it
The same calendar infrastructure can be reused across forecasting and operational models rather than rebuilt ad hoc by every team.
Banks
Payments, credit and liquidity patterns around moving holidays and working-day changes.
Retail
Demand forecasting, staffing and Ramadan/Eid exposure.
Tourism
Hotel, flight and visitor-demand models.
Statistical systems
Moving-holiday regressors and residual-seasonality audits.