Stress & Yield Curve Interaction

Core Question

How does market stress interact with yield curve regimes, and what has that historically meant for fixed-income conditions?

Why This Project Exists

Yield curve shape and market stress are rarely independent. This project quantifies their historical relationship using public FRED data — examining which curve regimes have coincided with elevated stress, how curve structure behaves conditional on stress state, and where the current environment sits relative to history.

This is the third project in a connected fixed-income research portfolio:

1. yield-curve-inflation-dashboard — macro environment: yield curve and inflation regimes
2. rate-sensitivity-regime-dashboard — bond duration risk across yield curve regimes
3. stress-curve-interaction-dashboard — stress behavior and curve interaction (this project)

Data Sources

All data from the Federal Reserve Economic Data (FRED):

Series	Description	Frequency
DGS2	2-Year Treasury Constant Maturity	Daily → monthly
DGS10	10-Year Treasury Constant Maturity	Daily → monthly
FEDFUNDS	Effective Federal Funds Rate	Monthly
BAA10Y	Moody's Baa corporate spread over 10Y Treasury	Daily → monthly
VIXCLS	CBOE VIX daily close	Daily → monthly mean

BAA10Y coverage begins ~1986. VIXCLS coverage begins January 1990. Analyses using both series are limited to their overlap period (~1990 onwards). All series resampled to month-end; VIX uses monthly mean rather than month-end to better represent realized volatility over the period.

Methodology

Yield Change & Spread

• spread_10y2y = DGS10 − DGS2
• spread_3m_chg = 3-month first difference of spread

Stress Score — Important Disclaimer

This is a simplified descriptive proxy constructed from two public market indicators. It is not a risk model, a tradable signal, or an investment recommendation.

Z-score computation:

z_i = (x_i − mean(x)) / std(x)    computed over full available history of each series

Combined stress score:

combined_stress = mean(available z-scores for that month)

If only BAA10Y is available (pre-1990), combined_stress = baa10y_z. If only VIX is available, combined_stress = vix_z. If neither is available, combined_stress = NaN and the month is classified as Unknown.

Z-scores are anchored to each series' full historical distribution, not just the overlap period. This maximises the historical context for each component.

Stress Regime Definitions

Regime	Condition
Low Stress	combined_stress < −0.5
Moderate Stress	−0.5 ≤ combined_stress ≤ +0.5
High Stress	combined_stress > +0.5

Yield Curve Regime Definitions

Applied in strict priority order (identical to Projects 1 and 2):

1. Re-steepening — spread was negative in prior 6 months AND spread rose > 0.25 ppts over 3 months AND spread is −0.25 to +0.75
2. Inverted — spread < 0
3. Flat — spread 0 to 0.50
4. Normal — spread > 0.50

Dashboard Walkthrough

Section	Content
1 — Current Snapshot	2Y, 10Y, Fed Funds, spread, BAA10Y, VIX, stress score, both regime badges
2 — Stress by Curve Regime	Avg z-score bar chart + raw values table by YC regime
3 — Stress × Curve Interaction	Co-occurrence heatmap (centerpiece) + 4 observational bullets
4 — Curve Behavior Under Stress	Spread and inversion stats by stress regime + summary bullets
5 — Stress → Portfolio Interpretation	Rule-based interpretation layer: market condition, credit risk, volatility risk, duration risk, credit exposure, convexity value, and positioning implications derived from current stress and curve regimes

Key Takeaways

• Inverted + High Stress is historically among the least common combinations, but has coincided with significant credit spread and equity dislocations.
• Normal curve regimes have historically coincided with Low or Moderate stress the vast majority of the time — High stress during a normal curve has been the exception.
• Re-steepening is rare (~4% of months) and has historically occurred during transitional macro environments; its stress profile varies depending on whether it follows a shallow or deep inversion.
• Stress regimes drive curve structure: High Stress periods have historically been associated with lower average spreads and higher inversion frequency than Low Stress periods.
• Portfolio interpretation layer (Section 5) translates the current stress and curve regime into historically observed risk characteristics — covering duration risk, credit exposure, convexity value, and positioning implications. All labels are descriptive heuristics, not investment recommendations.

How to Run

cd stress-curve-interaction-dashboard
pip install -r requirements.txt
streamlit run app.py

Optional: FRED API Key

cp .env.example .env
# Edit .env: FRED_API_KEY=your_key_here

Free API key: https://fred.stlouisfed.org/docs/api/api_key.html

Data is cached in data/processed/ as parquet files on first run (refreshed every 24 hours).

Disclaimer

This is an educational portfolio project. It is not investment advice. Stress regimes are simplified rule-based classifications using two public market indicators. The combined stress score is a descriptive proxy only — it is not a risk model output and should not be used for investment or risk management decisions. Historical patterns are not predictive of future outcomes.