Convexity
Duration provides a linear approximation of bond price sensitivity to yield changes, but the actual relationship between bond prices and yields is convex (curved). Convexity measures this curvature and provides a second-order correction to duration-based price estimates. Understanding convexity is essential for accurate bond price prediction, especially for larger yield changes, and is a critical component of professional fixed income analysis.
Why Convexity Matters
The Limitation of Duration
Duration assumes a linear relationship between price and yield:
- If yield increases by 1%, price decreases by D*%
- If yield decreases by 1%, price increases by D*%
The Problem:
The actual price-yield relationship is curved, not linear. This means:
- Duration underestimates price increases when yields fall
- Duration overestimates price decreases when yields rise
- The error increases with larger yield changes
The Convex Price-Yield Relationship
The price-yield curve for bonds exhibits positive convexity:
- The curve is upward-sloping and curved upward
- As yields fall, prices rise at an increasing rate
- As yields rise, prices fall at a decreasing rate
Graphical Representation:
- At low yields: Small yield decrease → Large price increase
- At high yields: Small yield increase → Small price decrease
- The curve is steeper at low yields and flatter at high yields
Why Bonds Have Positive Convexity
Mathematical Explanation:
The bond pricing formula creates a convex relationship because:
- Present value calculations use (1+y)^t in the denominator
- As yield changes, the impact on later cash flows is non-linear
- The weighted average of these non-linear effects creates convexity
Economic Explanation:
- When yields fall, bond prices rise, making bonds more valuable
- The increase in value accelerates as yields continue to fall
- When yields rise, bond prices fall, but the rate of decline slows
Convexity Definition and Formula
Convexity measures the curvature of the price-yield relationship. It's the second derivative of the price-yield function, normalized by price.
Convexity Formula
Where:
- C = Convexity
- P = Current bond price
- y = Yield to maturity (per period)
- CF_t = Cash flow in period t
- t = Time period
- n = Number of periods until maturity
Convexity for Coupon Bonds
For a standard coupon bond, the formula becomes:
[
C = \frac{1}{P} \times \frac{1}{(1+y)^2} \times \left[ \sum_{t=1}^{n} \frac{t(t+1) \times C}{(1+y)^t} + \frac{n(n+1) \times F}{(1+y)^n} \right]
]
Where:
- C (in formula) = Coupon payment per period
- F = Face value
Understanding Convexity Units
Convexity is typically expressed as:
- A positive number (for most bonds)
- Units: Time squared (years² or periods²)
- Higher convexity = More curvature = Better price performance
Example:
- Convexity of 50 means the bond has moderate convexity
- Convexity of 100 means the bond has high convexity
- Convexity of 20 means the bond has low convexity
Convexity Properties
1. Positive Convexity (Standard Bonds)
Most bonds have positive convexity:
- Straight (non-callable) bonds
- Zero-coupon bonds
- Puttable bonds
Benefits:
- Price increases more than duration predicts when yields fall
- Price decreases less than duration predicts when yields rise
- Asymmetric benefit: Gain more when rates fall, lose less when rates rise
2. Negative Convexity (Callable Bonds)
Some bonds have negative convexity:
- Callable bonds (when call is likely)
- Mortgage-backed securities (due to prepayment risk)
Characteristics:
- Price increases less than duration predicts when yields fall
- Price decreases more than duration predicts when yields rise
- Asymmetric disadvantage: Gain less when rates fall, lose more when rates rise
Why Callable Bonds Have Negative Convexity:
- When yields fall, call probability increases
- Issuer calls bond, limiting price appreciation
- Price is capped near call price
- Creates negative convexity
3. Zero Convexity
Theoretical case:
- Linear price-yield relationship
- Rare in practice
- Would occur if duration were constant across all yields
Duration + Convexity Approximation
Combining duration and convexity provides a more accurate estimate of price changes than duration alone.
The Improved Formula
Percentage Price Change:
Dollar Price Change:
Where:
- ΔP = Change in price
- P = Current price
- D* = Modified duration
- C = Convexity
- Δy = Change in yield (as a decimal)
The Two Components
1. Duration Effect (First-Order):
- -D* × Δy
- Linear approximation
- Dominates for small yield changes
2. Convexity Effect (Second-Order):
- ½ × C × (Δy)²
- Quadratic correction
- Becomes more important for larger yield changes
- Always positive for bonds with positive convexity
Example: Duration vs. Duration + Convexity
Given:
- Bond price: $1,000
- Modified duration: 5.0
- Convexity: 50
- Yield change: +2% (0.02)
Duration-Only Estimate:
Price change = -5.0 × 0.02 = -10%
New price ≈ $900
Duration + Convexity Estimate:
Price change = -5.0 × 0.02 + ½ × 50 × (0.02)²
Price change = -0.10 + ½ × 50 × 0.0004
Price change = -0.10 + 0.01 = -9%
New price ≈ $910
Difference:
The convexity adjustment reduces the estimated price decline by 1 percentage point, providing a more accurate estimate.
When Convexity Matters Most
Large Yield Changes:
- Convexity effect = ½ × C × (Δy)²
- Effect increases with the square of yield change
- For 1% change: effect is small
- For 5% change: effect is significant
High Convexity Bonds:
- Zero-coupon bonds have highest convexity
- Long-maturity, low-coupon bonds have high convexity
- These bonds benefit most from convexity
Convexity and Bond Characteristics
1. Convexity and Maturity
Relationship:
- Longer maturity → Higher convexity
- Convexity increases with maturity at an increasing rate
- Very long-maturity bonds have very high convexity
Why:
- Longer-maturity bonds have more cash flows in the future
- Future cash flows are more sensitive to yield changes
- This creates more curvature in the price-yield relationship
2. Convexity and Coupon Rate
Relationship:
- Lower coupon → Higher convexity
- Higher coupon → Lower convexity
- Zero-coupon bonds have maximum convexity
Why:
- Lower coupons mean more weight on the principal payment
- Principal payment is furthest in the future
- This creates more curvature
3. Convexity and Yield
Relationship:
- Lower yield → Higher convexity (for given bond)
- Higher yield → Lower convexity (for given bond)
Why:
- At lower yields, future cash flows are more valuable
- This increases their weight in the convexity calculation
- Creates more curvature
4. Convexity of Zero-Coupon Bonds
For zero-coupon bonds, convexity simplifies to:
[
C = \frac{n(n+1)}{(1+y)^2}
]
Where n = maturity in periods.
Example:
- Zero-coupon bond, 10 years to maturity
- Yield: 5%
Convexity = (10 × 11) / (1.05)² = 110 / 1.1025 = 99.8
This is very high convexity, making zero-coupon bonds highly sensitive to yield changes in a beneficial way.
Portfolio Convexity
For bond portfolios, convexity can be calculated as a weighted average:
Portfolio Convexity:
[
C_{portfolio} = \sum_{i=1}^{n} w_i \times C_i
]
Where:
- w_i = Weight of bond i in portfolio
- C_i = Convexity of bond i
- n = Number of bonds in portfolio
Example:
- Bond A: $500,000, Convexity = 40
- Bond B: $300,000, Convexity = 60
- Bond C: $200,000, Convexity = 80
- Total portfolio: $1,000,000
Portfolio Convexity = (0.5 × 40) + (0.3 × 60) + (0.2 × 80)
Portfolio Convexity = 20 + 18 + 16 = 54
Convexity Adjustment
The convexity adjustment improves price change estimates:
Accuracy Improvement
For Small Yield Changes (≤1%):
- Duration alone: Reasonably accurate
- Duration + Convexity: Slightly more accurate
- Improvement: Modest
For Moderate Yield Changes (1-3%):
- Duration alone: Noticeable error
- Duration + Convexity: Significant improvement
- Improvement: Substantial
For Large Yield Changes (≥3%):
- Duration alone: Large error
- Duration + Convexity: Essential for accuracy
- Improvement: Critical
Example: Accuracy Comparison
Given:
- Bond: 10-year, 5% coupon
- Price: $1,000
- Modified duration: 7.5
- Convexity: 60
- Yield change: -3% (yields fall)
Duration-Only Estimate:
Price change = -7.5 × (-0.03) = +22.5%
New price ≈ $1,225
Duration + Convexity Estimate:
Price change = -7.5 × (-0.03) + ½ × 60 × (-0.03)²
Price change = +0.225 + ½ × 60 × 0.0009
Price change = +0.225 + 0.027 = +25.2%
New price ≈ $1,252
Actual Price (calculated):
≈ $1,255
Error Analysis:
- Duration-only error: $30 (2.4%)
- Duration + Convexity error: $3 (0.2%)
- Convexity reduces error by 90%
Negative Convexity: Callable Bonds
Callable bonds exhibit negative convexity when the call is likely to be exercised.
Characteristics of Negative Convexity
When Yields Fall:
- Bond price rises, but call probability increases
- Issuer calls bond at call price
- Price is capped, limiting appreciation
- Price rises less than duration predicts
When Yields Rise:
- Call probability decreases
- Bond behaves like non-callable bond
- Price falls more than duration predicts (normal convexity lost)
Price-Yield Curve for Callable Bonds
The price-yield curve for callable bonds:
- At high yields: Similar to non-callable bonds (positive convexity)
- As yields fall: Curve flattens (convexity decreases)
- Near call price: Curve becomes flat or downward-sloping (negative convexity)
- Price is effectively capped at call price
Managing Negative Convexity
For Investors:
- Understand call risk
- Use yield to call (YTC) when appropriate
- Consider call protection periods
- Factor negative convexity into risk assessment
For Portfolio Managers:
- Monitor call probabilities
- Adjust duration estimates for callable bonds
- Consider callable bonds' asymmetric risk profile
Practical Applications
1. Price Prediction
Use Duration + Convexity for:
- Estimating price changes from yield movements
- Scenario analysis and stress testing
- Risk assessment and limit setting
2. Portfolio Optimization
Consider Convexity for:
- Balancing duration and convexity
- Optimizing risk-return tradeoffs
- Managing portfolio curvature
3. Hedging Strategies
Use Convexity in:
- Calculating hedge ratios
- Adjusting hedges for convexity effects
- Managing basis risk
4. Performance Attribution
Analyze Convexity Contribution:
- Decompose returns into duration and convexity effects
- Understand sources of performance
- Evaluate convexity's contribution to returns
Key Takeaways
Convexity measures the curvature of the price-yield relationship and provides a second-order correction to duration-based price estimates.
Most bonds have positive convexity, meaning prices rise more (fall less) than duration predicts when yields fall (rise).
Callable bonds can have negative convexity when call is likely, limiting price appreciation and creating asymmetric risk.
The duration + convexity formula provides more accurate price change estimates, especially for larger yield changes.
Convexity increases with maturity and decreases with coupon rate; zero-coupon bonds have maximum convexity.
Convexity adjustment becomes more important for larger yield changes, as the effect increases with the square of the yield change.
Understanding convexity is essential for accurate bond pricing, risk management, and portfolio optimization.
Professional fixed income analysis requires both duration and convexity for comprehensive risk assessment.
Next: Section 6 explores advanced bond valuation models for complex bond structures.