Synthetic PCA index (0–100) based on Google Trends search queries, seasonally adjusted via STL

Economy:
  1. High-Frequency Big Data Ingestion:
    • Monthly Google Trends query volumes spanning from January 2004 to August 2026 for 6 major European economies: Poland (PL), Germany (DE), Spain (ES), Netherlands (NL), France (FR), and Italy (IT).
    • For each country, 4 representative query clusters are tracked covering:
      • Collective bargaining and union wage agreements (e.g. Tarifverhandlung, convenio colectivo, NAO salaire, rinnovo CCNL, CAO).
      • Direct salary raise and negotiation searches (e.g. podwyżka pensji, Gehaltserhöhung, aumento stipendio).
      • Statutory minimum wage revaluation queries (e.g. Mindestlohn, SMIC, subida SMI, minimumloon).
      • General earnings and compensation calculators.
  2. Denoising & STL Seasonal Adjustment:
    • High-frequency sampling noise in Google Trends data is filtered using a 3-month center moving average.
    • Seasonal calendar variation (e.g., recurring January/Q1 minimum wage adjustments) is decomposed using STL (Seasonal and Trend decomposition using Loess) with adaptive seasonal windows (s.window = 7).
    • The seasonally adjusted components (SA) are smoothed with a 12-month moving average (MA 12M) to extract the underlying macroeconomic wage momentum.
  3. Dimensionality Reduction via Principal Component Analysis (PCA):
    • For each economy, the standardized component matrix is decomposed using PCA.
    • The first principal component (PC1) explains between 60.8% and 93.4% of total variance across the countries, validating a coherent common wage-pressure factor.
    • PC1 is oriented positively with respect to labor cost demand and min-max scaled to an intuitive 0 to 100 index scale.
  4. Data Attribution & Cloud Architecture:
    • Underlying analytical tables are synchronized with Google BigQuery (macroeconomic-dashboards.Disinflation_dashboard).
    • Developed as part of the ALK Macroeconomic Dashboards initiative.