Spain-STING — Banco de España's Real-Time Nowcasting Model for Spanish GDP Growth
Spain
Banco de España's Spain-STING model nowcasts quarterly Spanish GDP growth in real time from early indicators; a 2024 respecification (contemporaneous …
Austria · Vienna · See the Austria profile · See the Vienna profile
Evidence: Observational / pre–post Top 66% 53/100 · Ask Evidence Copilot about this practice
Austria's central bank built a weekly, real-time GDP tracker using payment-card and cash-to-noncash spending data. It captured a ~26% activity drop two weeks into the March 2020 lockdown, though the OeNB cautions it is built for crises, not normal-time precision.
ATOfficial Austrian GDP figures are published quarterly with a substantial lag, leaving policymakers without a real-time read on the economy during fast-moving shocks such as the COVID-19 pandemic.
To close that gap, Oesterreichische Nationalbank (OeNB) economists Gerhard Fenz and Helmut Stix set out to build a weekly indicator that could track aggregate economic activity in near real time during a crisis.
The OeNB Weekly GDP Indicator (GDP-I), published from May 2020, takes a demand-side, data-driven approach rather than a conventional time-series model, drawing on weekly high-frequency data including payment-card spending and the shift between cash and non-cash consumer transactions — a behavioural signal that became especially informative once lockdowns changed how Austrians paid for goods and services. Results were formally documented in the OeNB's 'Monetary Policy & the Economy' journal, Q4/20-Q1/21 issue.
The authors report that the weekly indicator produced "rather accurate results for aggregate economic activity" across the first two quarters after Austria's outbreak, and it captured a roughly 26% year-on-year contraction in activity within two weeks of the country's full lockdown in March 2020 — far faster than the multi-month wait for official statistics.
Fenz and Stix are explicit about the tool's limits: they describe the GDP-I as "an instrument to be used in times of crisis and not for observing economic developments in 'normal' times", and the paper does not claim its accuracy holds outside acute shocks. That self-declared boundary, published alongside the results, is itself a marker of methodological transparency rather than a weakness to hide.
National / regional programmes
Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.
Reusable artefacts from this practice — as published by their sources.
Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.
Where this practice's information was retrieved from, and when.
Spain
Banco de España's Spain-STING model nowcasts quarterly Spanish GDP growth in real time from early indicators; a 2024 respecification (contemporaneous …
Sweden
In an April 2025 Staff Memo, Riksbank economists found random-forest and neural-network models measurably outperformed Sweden's traditional time-series forecasts for …
Switzerland
SNB researchers tested eight machine-learning methods against 1,100+ time series to nowcast Swiss GDP. For the post-2008 period, every ML …
Netherlands
DNB's DFROG nowcasting model has tracked Dutch GDP growth in real time since 2024, publishing monthly estimates; a decade-long evaluation …
Open full copilot Grounded in cited practices — always check the sources.