us-market-bubble-detector

---
name: us-market-bubble-detector
description: Use when assessing market bubble risk, valuation concerns, or profit-taking timing via quantitative Put/Call, VIX, margin debt, breadth, and IPO data (Minsky/Kindleberger, v2.1).
category: Market & Macro
version: 1.0.0
tools: []
---

# US Market Bubble Detector (Revised v2.1)

Evaluate market bubble risk with a quantitative, data-driven Minsky/Kindleberger-style framework that prioritizes objective metrics (Put/Call ratio, VIX, margin debt, breadth, IPO data) over subjective impressions, with strict qualitative-adjustment criteria and confirmation-bias prevention.

## v2.1 Key Revisions
1. Mandatory quantitative data collection — measured values, not impressions or speculation
2. Clear numerical thresholds for every indicator
3. Two-phase evaluation, strict order: quantitative scoring → qualitative adjustment
4. Stricter qualitative cap: max +3 points (reduced from +5 in v2.0), each point requires measurable evidence
5. Explicit confirmation-bias checklist before adding any qualitative point
6. New "Elevated Risk" phase (8-9 points) for more gradual risk transitions

## When to Use
- "Is the market in a bubble?" / "Are we in a bubble?"
- Profit-taking, new-entry timing, or short-selling decisions
- Social phenomena reported (non-investors entering, media frenzy, IPO flood)
- Narratives like "this time is different" or "revolutionary technology" going mainstream
- Risk management for existing positions

## Phase 1: Mandatory Quantitative Data Collection
Collect via web_search before scoring anything:
- **Put/Call Ratio** (CBOE Equity P/C) — 5-day moving average
- **VIX** — current value and its percentile over the past 3 months; 21-day realized volatility; whether VIX sits in the bottom decile historically
- **Margin Debt** (FINRA balance) — latest month and year-over-year % change
- **Breadth** — % of S&P 500 stocks above the 50-day moving average
- **IPO activity** — quarterly count and median first-day return

Do not proceed to scoring without this data.

## Phase 2: Quantitative Scoring (0-12 points, mechanical)
| Indicator | 2 pts | 1 pt | 0 pts |
|---|---|---|---|
| Put/Call Ratio | P/C < 0.70 | P/C 0.70-0.85 | P/C > 0.85 |
| Volatility Suppression + New Highs | VIX < 12 and index within 5% of 52-week high | VIX 12-15 and near highs | VIX > 15 or >10% from highs |
| Margin Debt (leverage) | YoY +20%+ and all-time high | YoY +10-20% | YoY ≤10% or negative |
| IPO Market Overheating | quarterly count > 2x 5-yr avg and median first-day return +20%+ | count > 1.5x 5-yr avg | normal levels |
| Breadth Anomaly (narrow leadership) | new high with <45% of stocks above 50DMA | 45-60% above 50DMA | >60% above 50DMA |
| Price Acceleration | 3-month return > 95th percentile of past 10 years | 85-95th percentile | below 85th percentile |

Rationale, briefly: P/C<0.7 and low VIX-near-highs signal excessive optimism/complacency; rapid margin-debt growth and a poor-quality IPO flood are late-stage-bubble precursors; narrow breadth means the rally is fragile; extreme 3-month price acceleration is unsustainable.

## Phase 3: Qualitative Adjustment (max +3 points total)
Before adding any point, confirm: concrete measurable data (not impressions), an independent observer would reach the same conclusion, and no double-counting with Phase 2.

- **A. Social Penetration** (0-1 pt): requires ALL of — direct user report of non-investor recommendations, specific named/dated examples, and at least 3 independent sources. Vague claims like "AI narrative is prevalent" score 0.
- **B. Media/Search Trends** (0-1 pt): requires BOTH — Google Trends showing a measured 5x+ YoY increase, and confirmed mainstream coverage (magazine covers, TV specials with dates). "Elevated narrative" without data scores 0.
- **C. Valuation Disconnect** (0-1 pt): requires ALL of — P/E > 25 (only if not already counted in Phase 2), fundamentals explicitly ignored in mainstream discourse, and "this time is different" documented in major media. If real earnings support the valuation, score 0.

## Phase 4: Final Judgment
Final Score = Phase 2 (0-12) + Phase 3 (0 to +3), range 0-15.

| Score | Phase | Risk Budget |
|---|---|---|
| 0-4 | Normal | 100% |
| 5-7 | Caution | 70-80% |
| 8-9 | Elevated Risk (new in v2.1) | 50-70% |
| 10-12 | Euphoria | 40-50% |
| 13-15 | Critical | 20-30% |

## Recommended Actions by Phase
- **Normal (0-4)**: continue normal strategy; ATR 2.0× trailing stop; stair-step profit-taking (+20% take 25%). No shorting.
- **Caution (5-7)**: begin partial profit-taking (20-30%); tighten ATR to 1.8×; cut new position sizing 50%. Shorting not recommended.
- **Elevated Risk (8-9)**: increase profit-taking (30-50%); tighten ATR to 1.6×; new positions highly selective/quality-only; build cash. Consider shorting only after ≥2 of 7 composite conditions confirm, small exploratory size (10-15% of normal), ATR 2.0× stop.
- **Euphoria (10-12)**: accelerate profit-taking (50-60%); tighten ATR to 1.5×; no new longs except on major pullbacks. Actively consider shorting after ≥3 of 7 conditions confirm, small size (20-25% of normal), defined-risk (options/tight stops).
- **Critical (13-15)**: major profit-taking or full hedge; ATR 1.2× or fixed stop; cash-preservation mode. Shorting recommended after ≥5 of 7 conditions confirm, scale in small and pyramid on confirmation, tight stop (ATR 1.5×+), consider puts for defined risk.

## 7 Composite Conditions for Short-Selling
Require at least 3 (scaling to 5 at Critical) confirmed before shorting:
1. Weekly chart shows lower highs
2. Volume peaks out
3. Leverage indicators drop sharply (margin debt decline)
4. Media/search trends peak out
5. Weak stocks start breaking down first
6. VIX surges (spike above 20)
7. Fed/policy shift signals

## Common Failures to Avoid
- Scoring media saturation from "many reports" without a measured Google Trends number
- Treating an expert warning as decisive without checking Put/Call, VIX, and margin-debt data
- Scoring a single sharp daily move as "price acceleration" without checking its position in the 10-year distribution
- Scoring valuation alone (e.g., "P/E 17 = disconnect") without checking narrative dependence and other quantitative indicators

## Reporting
Present: final score and phase, evaluation date, the full Phase 2 table (measured value / score / rationale per indicator), the Phase 3 qualitative adjustments with their evidence and a confirmation-bias self-check, and the recommended risk budget and actions for the phase reached.