Fixed Income Quantitative Strategies: Duration, Curve, and Spread
Fixed income markets, with over $130 trillion in global outstanding debt, offer a rich landscape for systematic quantitative strategies. Unlike equities, where price discovery is primarily driven by earnings expectations and sentiment, fixed income pricing is grounded in measurable quantities: interest rates, credit quality, prepayment behavior, and the term structure of risk. This mathematical tractability makes bonds particularly suited to quantitative approaches. The three primary systematic strategies -- duration timing, yield curve positioning, and credit spread trading -- each exploit a different dimension of fixed income risk and return.
Duration Strategies
Duration Risk and Return
Duration measures bond price sensitivity to interest rate changes. Modified duration gives the approximate percentage price change for a 1% change in yield:
dP/P approximately = -D_mod * dyFor a bond with modified duration of 7.0, a 50 basis point yield increase produces approximately -3.5% price decline.
Key rate duration decomposes duration into sensitivity to specific points on the yield curve. A portfolio with total duration of 6.0 might have key rate durations of 1.5 at 2-year, 2.0 at 5-year, 1.5 at 10-year, and 1.0 at 30-year, revealing its specific curve exposure.Systematic Duration Timing
Duration timing adjusts portfolio duration based on quantitative signals predicting interest rate direction:
Momentum signal: When rates have been falling (bond prices rising), maintain or extend duration. The autocorrelation of monthly yield changes is approximately 0.15-0.25, supporting momentum-based timing. Mean-reversion signal: When real yields (nominal minus inflation expectations) are more than 1 standard deviation above their 5-year average, extend duration (yields are likely to decline). When below average, shorten duration. Macro signal: When leading economic indicators are deteriorating, extend duration (anticipating rate cuts). When improving, shorten duration. Combined duration signal: D_target = D_neutral + delta_mom mom_signal + delta_value value_signal + delta_macro * macro_signalWhere D_neutral is the benchmark duration, and delta terms control the tactical tilt magnitude (typically plus or minus 1-2 years of duration).
Historical Performance
Duration timing strategies based on combined signals have historically added 30-70 basis points of annual alpha relative to a constant-duration benchmark, with information ratios of 0.3-0.5. The primary value is drawdown reduction during rate-rising environments.
Yield Curve Strategies
Curve Shape Decomposition
The yield curve can be decomposed into three principal components:
Level (PC1): Parallel shift in all yields. Explains approximately 85-90% of yield curve variation. Duration strategies capture level risk. Slope (PC2): Change in the steepness of the curve (long rates minus short rates). Explains approximately 8-12% of variation. Slope strategies are distinct from duration strategies. Curvature (PC3): Change in the curve's convexity (belly rates relative to wings). Explains approximately 2-4% of variation. Butterfly strategies capture curvature.Steepener/Flattener Trades
Steepener: Long short-term bonds, short long-term bonds. Profits when the yield curve steepens (short rates fall relative to long rates). Typical trade: long 2-year notes, short 10-year notes, duration-neutral. Flattener: Short short-term bonds, long long-term bonds. Profits when the curve flattens or inverts. Duration-neutral implementation: Match the DV01 (dollar value of a basis point) of the long and short legs: Notional_2Y / Notional_10Y = DV01_10Y / DV01_2YFor 2-year DV01 of $0.0198 per $100 and 10-year DV01 of $0.0850 per $100:
Ratio = 0.0850 / 0.0198 = 4.29Hold $4.29 of 2-year notes for every $1 of 10-year notes to achieve duration neutrality.
Systematic Curve Trading
Carry and roll-down: Calculate the carry (coupon income minus financing cost) and roll-down return (price appreciation as the bond "rolls down" the curve to shorter maturities) for each maturity point. Overweight maturities with the highest total carry. Roll-down return for a bond at maturity T: Roll_T = -(D_T (y_T - y_{T-dt})) (dt/T)Where y_T is the yield at maturity T and dt is the holding period. In a steep curve, the roll-down return is substantial for intermediate maturities (5-7 years), creating a systematic overweight signal.
Mean-reversion in curve slope: The 2s10s spread (10-year minus 2-year yield) has historically mean-reverted with a half-life of approximately 6-12 months. When the spread is more than 1 standard deviation from its 3-year average, fade the move (steepener if spread is abnormally flat, flattener if abnormally steep).Butterfly Trades
A butterfly combines three maturities: short the wings (short-term and long-term) and long the body (intermediate), or vice versa.
Cash-neutral butterfly: 2s-5s-10s butterfly with weights chosen so the position has zero duration and zero net investment:- Short 2-year, long 5-year, short 10-year (long belly butterfly)
Credit Spread Strategies
Credit Spread Decomposition
The credit spread (yield above the risk-free rate) compensates for:
- Expected default loss: The probability-weighted loss from default. For investment-grade bonds, this is typically 10-30 bps of the spread.
- Credit risk premium: Compensation for bearing default risk beyond the expected loss. This is the component that systematic strategies target.
- Liquidity premium: Compensation for the lower liquidity of corporate bonds relative to Treasuries. Typically 10-50 bps for investment-grade, 50-200 bps for high yield.
Systematic Spread Trading
Cross-sectional momentum: Bonds whose spreads have been tightening (outperforming) tend to continue tightening. Rank corporate bonds by spread change over the past 3-6 months and go long the top quintile (tightening) and short the bottom quintile (widening). Value (spread level): Bonds with wider spreads relative to their rating and sector peers are expected to tighten. This is analogous to the value factor in equities. Measure relative value as the spread deviation from the predicted spread (based on rating, duration, sector, and maturity). Quality momentum: When aggregate credit quality is improving (more upgrades than downgrades), overweight credit. When deteriorating, underweight. The ratio of upgrades to downgrades has a 6-12 month lead on credit spread movements. Carry: In credit markets, carry equals the credit spread minus the expected default loss: Carry = OAS - PD * LGDWhere OAS is the option-adjusted spread, PD is the probability of default, and LGD is the loss given default. Overweight bonds with the highest carry (spread minus expected default loss).
Credit Curve Trading
The credit term structure (spread as a function of maturity) offers additional opportunities:
- Steep credit curves: Overweight short-dated credit (high carry per unit of spread duration) and underweight long-dated credit
- Flat credit curves: Overweight long-dated credit (similar spread with more duration, implying the market is pricing low future default risk)
CDS vs. Cash Bond Basis
When the CDS spread differs from the cash bond spread (the "basis"), arbitrage opportunities arise:
Negative basis (CDS spread < bond spread): Buy the bond and buy CDS protection. Earn the positive basis as carry. Positive basis (CDS spread > bond spread): Short the bond and sell CDS protection (more complex and risky).The basis has historically averaged near zero but can widen significantly during stress (negative basis exceeded 200 bps during the 2008 GFC). Basis trading captures mean-reversion in this relationship.
Risk Management for Fixed Income Strategies
Interest Rate Risk
Hedge duration exposure using Treasury futures or interest rate swaps. Monitor key rate duration exposures to ensure the portfolio's rate sensitivity matches the intended strategy.
Credit Risk
Manage credit exposure through:
- Issuer limits: Maximum 2-3% per issuer to avoid concentration
- Rating distribution: Maintain minimum investment-grade allocation
- Sector limits: Maximum 15-20% per sector to avoid sector concentration
- CDS hedging: Use CDS index (CDX IG or HY) to hedge systematic credit exposure while retaining idiosyncratic spread positions
Liquidity Considerations
Corporate bonds are substantially less liquid than government bonds. Bid-ask spreads range from 5-20 bps for investment-grade and 25-100 bps for high-yield. Position sizes should be calibrated to daily trading volume, with maximum positions typically limited to 1-2 days of average volume.
Key Takeaways
- Fixed income quantitative strategies operate along three dimensions: duration (interest rate level), curve (term structure shape), and spread (credit risk premium), each providing independent sources of return
- Duration timing based on momentum, mean-reversion, and macro signals adds 30-70 bps annually, with the primary value in drawdown reduction during rate-rising environments
- Yield curve strategies exploit predictable patterns in curve slope and curvature, with carry and roll-down providing the most reliable systematic alpha
- Credit spread strategies combining momentum, value, and carry signals capture the credit risk premium while managing default and liquidity risk
- Combining duration, curve, and spread strategies achieves diversification benefits because the three return sources have low correlation with each other
Frequently Asked Questions
What data do I need for fixed income quant strategies?
At minimum: daily yield curves (Treasury par yields or zero-coupon yields by maturity), credit spreads by rating and sector (OAS from Bloomberg or ICE), CDS spreads for liquid issuers, and economic indicators (PMI, employment, inflation). For credit strategies, bond-level data (CUSIP-level spreads, ratings, trading volumes) is required. Data quality is critical: fixed income data is more heterogeneous and less standardized than equity data.
How do fixed income quant strategies perform during rising rate environments?
Pure duration strategies underperform during rate rises (by construction). However, duration timing strategies can limit damage by shortening duration when rates are rising. Credit strategies can provide positive returns during gradual rate increases (which often accompany economic strength and credit improvement). Yield curve strategies are largely rate-direction agnostic because they are duration-neutral. A combined approach typically delivers positive returns even during moderate rate increases.
Are fixed income quant strategies capacity-constrained?
Less so than equity strategies because the fixed income market is larger ($130T+ globally vs. $100T+ for equities). However, specific segments (high-yield, emerging market debt) have limited liquidity. Investment-grade credit strategies can manage $5-20 billion without significant capacity constraints. Duration and curve strategies using Treasury futures are highly scalable.
Can I apply equity factor models to fixed income?
Conceptually yes, but the implementation differs. The "value" factor in fixed income is spread cheapness relative to fundamentals. "Momentum" is spread change direction. "Quality" maps to credit quality metrics (leverage, interest coverage). "Carry" is the spread minus expected default loss. The same factor concepts apply, but the specific signals and risk management are tailored to fixed income characteristics.
How do I hedge a fixed income quant strategy against systematic risk?
Duration risk is hedged with Treasury futures or interest rate swaps (choose the maturity that best matches the portfolio's key rate duration profile). Credit risk is hedged with CDS indices (CDX IG for investment-grade, CDX HY for high-yield). Liquidity risk is managed through position limits and cash reserves. A well-hedged fixed income quant strategy isolates idiosyncratic opportunities while neutralizing macro rate and credit exposure.