🔎 Google Trends REAL DATA · 6 ATTENTION INDICES

Methodology · Real Google Trends Data

Each of the six attention indices is computed from real Google Trends interest values (0-100 normalized) for a curated set of query terms. No randomization, no fabricated data — every value comes from trends.google.com directly.

Indices:

• crypto_fear: mean of "bitcoin crash", "crypto crash", "cryptocurrency scam"
• recession_fear: mean of "recession", "financial crisis", "stock market crash"
• employment_stress: mean of "layoffs", "unemployment", "looking for a job"
• melt_up_attention: mean of "buy stocks", "all time high", "trending stocks"
• fed_attention: mean of "fed rate hike", "fomc meeting", "powell speech"
• ai_hype: mean of "ai stocks", "nvidia stock", "openai"

Delta (Δpp) = current value minus 7-days-ago value. The change in retail attention is more informative than the absolute level — a +30pp spike is the alpha signal.

Regime thresholds: TRANQUIL <30 · WATCHING 30-55 · ELEVATED ≥55 OR Δ≥15pp · SPIKE ≥75 OR Δ≥30pp.

Composite regime aggregates the indices:

• CRISIS_ATTENTION — crypto_fear AND recession_fear both elevated
• MELT_UP_ATTENTION — melt_up_attention spiking with low fear
• STRESS_BUILDING — one+ fear indices elevated
• COMPLACENCY — all indices tranquil (often pre-event quiet)
• NORMAL — mixed

Source: trends.google.com/trends/api/dailytrends + widgetdata/multiline · refresh every 4 hours.