Welcome to the "Shortcut" of Financial Math!
In the world of actuarial science, time is money—literally. Imagine you are managing a massive portfolio of bonds and the central bank suddenly changes interest rates. You need to know, right now, how much the value of your portfolio just dropped or climbed. Re-calculating every single cash flow's present value from scratch would take too long.
In this chapter, we learn how to use Duration and Convexity as "shortcuts" to estimate these changes instantly. Think of these methods as using a quick mental map instead of unfolding a giant paper map every time you take a turn.
The Core Concept: Sensitivity
Before we dive into the math, let’s get the intuition down. When interest rates (\(i\)) go up, the present value (\(PV\) or \(P\)) of future cash flows goes down. But by how much?
Analogy: Imagine you are holding a long pole. If you tilt your hand just a little bit, the end of the pole moves a lot. The length of that pole is like Duration. The longer the duration, the more sensitive your portfolio is to changes in interest rates.
Level 1: The Linear Approximation (Modified Duration)
The simplest way to estimate a change in price is to assume the relationship between interest rates and price is a straight line. We use Modified Duration (\(D_{mod}\)) for this.
The formula for the approximate change in price (\(\Delta P\)) is:
\(\Delta P \approx -P(i) \cdot D_{mod} \cdot \Delta i\)
Or, if you want to find the new price (\(P_{new}\)) directly:
\(P(i + \Delta i) \approx P(i) \cdot [1 - D_{mod} \cdot \Delta i]\)
Why the negative sign? Remember, price and interest rates move in opposite directions. If rates go up (\(\Delta i\) is positive), the price must go down!
Quick Review: What was \(D_{mod}\) again?
Don't worry if you forgot! Just remember that Modified Duration is related to Macaulay Duration (\(D_{mac}\)) by this simple link:
\(D_{mod} = \frac{D_{mac}}{1+i}\)
If the exam gives you \(D_{mac}\), make sure to divide by \((1+i)\) before using the linear approximation formula above!
Level 2: The Macaulay Approximation
Sometimes, the exam might ask you to estimate the price change specifically using Macaulay Duration. This isn't just a straight line; it's a slightly "curved" version that often works better for volatile rates.
The formula for the new price using Macaulay Duration is:
\(P(i_{new}) \approx P(i_{old}) \cdot \left( \frac{1+i_{old}}{1+i_{new}} \right)^{D_{mac}}\)
Memory Trick: Notice how the \((1+i)\) terms are "flipped"? Since \(i_{new}\) is on the bottom, if the new interest rate is higher, the fraction becomes smaller, which correctly shows the price dropping.
Level 3: Adding the "Bend" (Convexity)
In reality, the relationship between price and interest rates isn't a straight line—it’s a curve. If interest rates change by a large amount, our linear "shortcut" from Level 1 will be wrong because it misses the curve.
To fix this, we add Convexity (\(C\)). This is called a second-order approximation (or a Taylor Series expansion, for the math fans).
The more accurate formula is:
\(P(i + \Delta i) \approx P(i) \cdot [1 - D_{mod} \cdot \Delta i + \frac{C}{2} \cdot (\Delta i)^2]\)
Did you know? Convexity is almost always good for a bondholder. Notice the plus sign before the convexity term: \(+ \frac{C}{2} \cdot (\Delta i)^2\). Because \((\Delta i)^2\) is always positive (even if rates go down!), convexity always adds value to your estimate compared to the simple linear version. It makes the price drop less when rates rise and makes the price jump more when rates fall!
Step-by-Step Example
The Problem: A portfolio has a value of \$1,000 at an interest rate of 5%. Its Modified Duration is 8 and its Convexity is 100. Estimate the new value if the interest rate increases to 6%.
\nStep 1: Identify your variables.
\n\(P = 1,000\)
\n\(D_{mod} = 8\)
\n\(C = 100\)
\n\(\Delta i = 0.06 - 0.05 = 0.01\)
\nStep 2: Calculate the Duration effect (the "drop").
\n\(-D_{mod} \cdot \Delta i = -8 \cdot 0.01 = -0.08\) (An 8% drop)
\nStep 3: Calculate the Convexity effect (the "bonus").
\n\(\frac{C}{2} \cdot (\Delta i)^2 = \frac{100}{2} \cdot (0.01)^2 = 50 \cdot 0.0001 = 0.005\) (A 0.5% gain)
\nStep 4: Combine them for the final answer.
\n\(P_{new} \approx 1,000 \cdot [1 - 0.08 + 0.005] = 1,000 \cdot [0.925] = \$925\)
Common Mistakes to Avoid
• Mixing up \(D_{mod}\) and \(D_{mac}\): Always check which one the question gives you. If you have \(D_{mac}\) but the formula requires \(D_{mod}\), you must divide by \((1+i)\).
• The \(C/2\) Trap: In the second-order formula, many students forget to divide the convexity by 2. Don't let the exam writers catch you! It is always \(\frac{1}{2} C (\Delta i)^2\).
• Sign Errors: Remember that the duration term is subtracted and the convexity term is added. Rates up = Price down (mostly).
• Units: When \(\Delta i\) is "100 basis points," that means 0.01. Using "1" instead of "0.01" will lead to a very wrong answer!
Key Takeaways Summary
1. First-Order Approximation: Uses only Modified Duration. It's a straight-line estimate: \(P \cdot [1 - D_{mod}\Delta i]\).
2. Second-Order Approximation: Uses both Duration and Convexity for a "curved" and more accurate estimate: \(P \cdot [1 - D_{mod}\Delta i + \frac{C}{2}(\Delta i)^2]\).
3. Macaulay Estimate: A specific alternative using Macaulay Duration: \(P \cdot (\frac{1+i}{1+i+\Delta i})^{D_{mac}}\).
4. Interest Rate Relationship: When rates increase, duration pulls the price down, but convexity pushes it slightly back up. Convexity is the "cushion" that helps bondholders.