About This Event
“Strip out the calendar noise and see what the economy is really doing.”
A five-day virtual regional workshop on seasonal adjustment and trend-cycle estimation of high-frequency data. You will identify seasonal and calendar effects, adjust monthly and quarterly series with X-13ARIMA-SEATS and JDemetra+, estimate trend-cycles and set up policies for revisions and publication.
What You'll Explore
Seasonally adjusted series produced from your own monthly or quarterly data
JDemetra+ specifications you can reuse in regular production
Calendar regressors for moving holidays relevant to your country
A draft seasonal adjustment and revision policy for your department
A release note template that explains adjusted figures to users
Who Should Attend
Open to all qualifying staff, particularly: Statistics Compilers, Price & National Accounts Statisticians, Research Economists, Monetary Policy Analysts, Data Quality Officers.
Why This Course Matters
Clearer Economic Signals
Raw monthly data mixes real movements with harvests, holidays and Ramadan effects. Seasonal adjustment removes that noise, so policy teams can read turning points instead of calendar patterns.
Consistent Methods
A documented seasonal adjustment policy means every series is treated in the same, defensible way. Consistency avoids awkward questions when two departments publish conflicting adjusted figures.
Managed Revisions
Seasonally adjusted figures revise as new data arrives, which can unsettle users. A clear revision policy explains those changes in advance and protects the credibility of your releases.
Programme
Day 1
Time-series components & why seasonal adjustment matters
You will decompose time series into trend-cycle, seasonal, calendar and irregular components, and see how unadjusted data can mislead policy analysis of inflation, credit and activity.
Day 2
Pre-treatment: outliers, calendar & moving holidays
You will use RegARIMA models to detect outliers and level shifts, and build regressors for trading-day effects and moving holidays such as Easter, Ramadan and Eid in African data.
Day 3
X-13ARIMA-SEATS & TRAMO-SEATS in JDemetra+
You will adjust monthly and quarterly series with X-13ARIMA-SEATS and TRAMO-SEATS in JDemetra+, reading the diagnostics that tell you whether the adjustment is sound or needs revisiting.
Day 4
Trend-cycle estimation & short, volatile series
You will estimate trend-cycles with Henderson filters and model-based methods, and handle the short, volatile series common in developing economies where standard defaults often fail.
Day 5
Revision policy, aggregation & publication
You will set a revision policy, choose between direct and indirect adjustment of aggregates, and draft metadata and a release note that explains your adjusted series clearly to users.
Standards & Faculty Benchmark
ESS Guidelines on Seasonal Adjustment
Eurostat's European Statistical System guidelines setting best practice for seasonal adjustment.
X-13ARIMA-SEATS (US Census Bureau)
The widely used seasonal adjustment program combining X-11 filters with SEATS model-based methods.
IMF Quarterly National Accounts Manual
IMF guidance including a dedicated chapter on seasonal adjustment of quarterly national accounts.
The ESS guidelines and IMF manual are the benchmarks statistical agencies worldwide use to judge seasonal adjustment practice. Following them means your adjusted series are comparable with peers and your methods hold up when users or reviewers probe the numbers.
Is This Right for You?
- ☑You compile or analyse monthly or quarterly economic series
- ☑Your institution publishes or plans to publish adjusted data
- ☑You want to read seasonal adjustment diagnostics with confidence
Good to Know
Suited to staff who work with monthly or quarterly series; basic statistics is enough and the software is taught hands-on. Delivered live online to a regional cohort. You leave with adjusted series from your own data, JDemetra+ specifications and a draft revision policy.
The Bottom Line
Go back with cleaner series, a documented method and a revision policy that make your high-frequency data easier to trust.
Recommended For
Open to all qualifying staff, particularly: Statistics Compilers, Price & National Accounts Statisticians, Research Economists, Monetary Policy Analysts, Data Quality Officers.