What You Actually Need to Know for Bond Math Interviews
Bond math interviews aren't about memorizing formulas. They're about whether you can reason through a pricing problem under pressure without panicking when the numbers get weird. I've sat on both sides of that desk, and the pattern is predictable. People who memorized Durbin or whatever prep book they bought will choke when you throw a Turkish lira fixed coupon bond at them with semi-annual compounding and a settlement date that falls mid-coupon. The core question every interviewer circles back to is how you price a bond. Not the textbook version where cash flows are evenly spaced and YTM is constant. The real version. Start with the basic present value equation: sum each cash flow discounted at the appropriate spot rate, accounting for day count convention and accrued interest. That's it. Everything else is decoration.
Common Bond Math Interview Questions That Actually Show Up
Macaulay duration, modified duration, convexity adjustment. DV01 calculations. Yield curve construction. Swap pricing basics. Forward rates. Option-adjusted spread if you're interviewing for rates quant roles. Here's what most candidates miss: they'll compute duration correctly but then fail when you ask how duration changes when the yield curve steepens non-parallel. They know the formula but not the mechanics. I once had a candidate who correctly calculated the key rate duration for a 10-year bond but couldn't explain why shifting the 2-year rate by 1 basis point affected the price at all when the bond had no cash flow near year 2. That's the distinction between someone who can run a calculation and someone who understands what the number means. Key rate duration isn't just a bucketed approximation. It captures the sensitivity to non-parallel shifts, and understanding that requires thinking about which cash flows are actually impacted by which parts of the curve. Another question that catches people out: what happens to convexity when you have embedded options? A callable bond's convexity goes negative at low yields because the option kicks in and shortens the effective duration. Candidates will tell you convexity is always positive because the formula says so. The formula assumes fixed cash flows. Options break that assumption. This isn't some advanced topic. It's basic rates intuition.
Here's a practical tip I wish more people knew: when you're doing duration calculations by hand under time pressure, use the bucketed method rather than trying to derive everything from first principles. Price at y plus delta, price at y minus delta, apply the standard formula. It's faster and less error-prone. I've seen people waste ten minutes deriving the analytical duration formula when they could have had a numerical answer in two minutes. The interview will also test whether you understand the relationship between different yield measures. Bond equivalent yield versus money market yield versus effective annual yield. Pick the wrong one and your entire price calculation is off. This sounds trivial but I've seen it cost deals. One time during a live trading conversation I noticed a junior colleague was quoting prices on a T-bill using BEY instead of the money market convention we trade. The price difference was tiny in absolute terms but it flagged that they'd never internalized how convention drives everything in fixed income. For OAS and Monte Carlo modeling questions, keep it grounded. You don't need to derive the Hull-White model from scratch. Know that OAS strips out the option component from the spread to give you a risk-adjusted measure, and that the simulation needs to be path-independent for some products but path-dependent for others like prepayment models. If they ask about lattice models versus PDE approaches, explain when each is appropriate. Lattice is intuitive and flexible for discrete barriers. PDE is more efficient for smooth payoffs in continuous time.
Get the Full Details
Let me address something bluntly: bond math as a standalone skill set has real limitations. The models break down in illiquid markets where spot rates are interpolated from sparse data. Convexity adjustments become meaningless when volatility spikes and optionality dominates price movement. Duration is a linear approximation and that linearity assumption fails badly for bonds with significant option content or when rates move more than 50 basis points in a single shift. No interviewer expects you to pretend these models are perfect. They want to hear that you know where they break and what you'd do instead. When I encounter a situation where the standard duration framework fails, I fall back to full repricing across multiple rate scenarios. It's computationally heavier but it doesn't make assumptions about linearity or constant convexity. For products with embedded options, I switch to a binomial tree or a short-rate model depending on the complexity. This usually takes longer but it's honest about the uncertainty. One more thing that separates good candidates from the rest: they can explain why bond prices move inversely to yields without just reciting the rule. The mechanism is discounting. A higher yield means each future cash flow is worth less today. That's all there is to it. But if you can connect that mechanical explanation to what happens at the margin when rates change, show you understand that it's not a symmetric relationship because of convexity, and acknowledge that this breaks down when prepayment risk enters the picture, you're already ahead of half the room.
If you want to practice, the best material isn't some generic prep book. It's actually working through problems on your own calculator until the relationships feel intuitive. Price a bullet bond. Compute its duration. Then shift the curve and reprice. Watch how the duration estimate compares to the actual price change. Do it for different coupons, different maturities, different yield levels. The pattern will emerge naturally and that's what sticks during an interview.
