Fixed Income Analytics
CDS Cointegration
Khalid Khan
—
Data as of: —
Implied CDS · CreditGrades model · derived from equity prices

Lol 😂

Do you know how expensive real CDS data is? But if you really want it, please send me a message — I'll see what I can do.

What is this dashboard?

Two CDS instruments are cointegrated when their credit spreads tend to move together over time — not perfectly in lockstep, but anchored by a long-run relationship. When one spread drifts too far from that relationship, it tends to snap back. This dashboard identifies those moments and shows the historical track record of trading that reversion.

The heatmap shows which pairs are currently cointegrated — green means the relationship is statistically significant. Click any cell to load the full analysis for that pair: price history, spread dynamics, trade signals, a simulated backtest, and a live trade recommendation if conditions are met right now.

⚠ Spreads shown are implied CDS derived from equity price history using the CreditGrades structural model (λ=0.50, R=40%, 252-day hist. vol). These are not quoted market spreads. Universe restricted to CDX.NA.IG names with mean spread 5–500 bps, post-2021 data only.

Parameters
Sector
i
Filter to a single credit sector. Pairs within the same sector share common risk factors, making cointegration more economically meaningful.
Window
i
Number of trading days used for each rolling calculation. 126 days (6 months) balances stability and responsiveness.
P-Value Threshold
i
Only pairs below this threshold are highlighted as cointegrated. 0.05 means a 5% chance the relationship is random. Lower = more selective.
Entry Trigger (σ)
i
How far the spread must deviate (in σ) before a trade is entered. A second position is added automatically at entry + 1σ.
Profit Target (σ per leg)
i
Each leg exits when z-score reverts by this amount from its entry. e.g. entered at -1σ with 0.5σ target exits at -0.5σ.
Beta Stability (CV)
i
Coefficient of variation of rolling beta. Lower = more stable hedge ratio. Trades are blocked during unstable periods.
Heatmap As-Of Date
i
Show the heatmap as it looked on a specific date. Leave blank for latest available data.
P-Value Heatmap — click a cell to load pair
Loading…
Loading pair data
No Signal
Backtest PnL ($)
—
Win Rate
—
Total Trades
—
Current P-Value
—
Current Z-Score
—
Half-Life (days)
—
Beta (current)
—
Beta Stability CV
—
Proposed Trade
Cointegrated (p-value below threshold)
Cointegrated + Beta stable — trades allowed
Price Levels, Spread, Signal & Performance
Diagnostics
i
Rolling beta measures how much T2 spread moves per unit of T1 spread. The CV (coefficient of variation) measures beta stability — lower is more stable. Trades are blocked when CV exceeds the threshold.
i
Half-life estimates how many days it takes for the spread to revert halfway to its mean. Shorter is better for trading — values above ~60 days suggest slow or unreliable reversion. Spikes indicate instability.
Trade Blotter
Trade Type Date Days Held Dir T1 Size T2 Size T1 T2 T1 Px T2 Px Spread Z Trade Beta Rolling Beta Spread PnL ($) Carry ($) Total PnL ($)