Debt Capital Markets interview preparation

Questions

100

Traced to a firm

45

Firms

26

Updated

September 2026

Asked at

All firmsTSTruist Securities5PIMCO4TD Securities4Apollo Global Management3Nomura3Scotiabank3Bain Capital2Houlihan Lokey2Mizuho2Neuberger Berman2Oaktree Capital Management2RCRBC Capital Markets2Carlyle Group1Deutsche Bank1Golub Capital1HPS Investment Partners1Invesco1KKR1Lazard1Moelis & Company1Moody's1Northern Trust1NUNuveen1Rothschild & Co1S&P Global1Wells Fargo Securities1

Topic

All topicsBond mechanics11Duration and convexity6Yield curve and rates5Credit spreads5Credit analysis and ratings13Credit modelling9Primary issuance9Syndication and loans10Structured credit7Covenants and documentation5Liability management5Indian debt markets7Fit8

Level

AnyCoreIntermediateHard

Type

AnyTechnicalBrainteaserMarket viewCaseFit

1. 005A bond has a 6 percent coupon, 5 years to maturity, and trades at 95. Roughly what is the yield? Do it in your head.Bond mechanicsIntermediatetechnicalSyndicate desks

   Say this

   About 7.2 percent. Coupon income is 6 on 95, so 6.3 percent, plus 5 points of pull to par spread over 5 years, which is about 1 point a year or another 1.05 percent on a 95 price. Add them and you get roughly 7.3, and the true answer is 7.2.

   Then walk it

   1. The approximation is the coupon plus the annualised capital gain, all over the average of price and par. That is the standard back-of-envelope yield.
   2. Formally: 6 plus 5 over 5, so 7, divided by the average of 95 and 100, which is 97.5. That gives 7.18 percent.
   3. The reason the shortcut slightly overstates is discounting: the pull to par arrives at the end, so its present value is less than a point a year.
   4. Sanity check the direction first. It trades below par, so the yield must be above the 6 percent coupon. Getting the direction wrong is fatal; being 10 basis points out is not.
   5. The desk habit worth showing: quote it as roughly 7.2 and say you would price it exactly on the calculator. Interviewers want the instinct plus the discipline, not false precision.

   Where candidates lose it

   Reaching for a calculator, or freezing. This is a mental arithmetic test dressed as a bond question. State the direction first, then the approximation, then say you would confirm it precisely. Never quote a number to three decimals from a mental estimate.

   Expect next

   • Now do it for a bond trading at 105.
   • What if it had 20 years to maturity instead of 5?
   • What is the duration of that bond, roughly?
2. 015Two bonds have the same maturity but different coupons. Which has the longer duration, and why?Duration and convexityIntermediatetechnicalFixed income asset management

   Say this

   The lower coupon bond. Duration is the present-value-weighted average time to the cash flows, and a low coupon puts proportionally more of the value in the final principal payment, which is the most distant cash flow.

   Then walk it

   1. Think of the weights. A 10-year 8 percent bond returns a lot of cash early, so the weighted average time is pulled forward. A 10-year 2 percent bond has almost all its value in the redemption at year 10.
   2. The limiting case proves it: a zero-coupon bond has no early cash flows at all, so its duration equals its maturity, which is the maximum possible for that tenor.
   3. Numbers: at a 5 percent yield, a 10-year 8 percent coupon bond has Macaulay duration around 7.1, a 4 percent coupon around 7.9, and the zero is 10.0.
   4. The same logic explains the other two drivers. Longer maturity extends the weighting, and a lower yield reduces the discounting of distant cash flows, so both lengthen duration.
   5. The practical consequence: low-coupon long-dated bonds issued in the zero-rate era carry enormous duration, which is why the 2022 rate move produced 40 percent-plus drawdowns on some sovereign long bonds.
   6. One qualification: this holds for bullets. A callable low-coupon bond may have shorter effective duration than the maths suggests, because the option truncates it.

   Where candidates lose it

   Guessing. It is a two-way question and half of candidates answer higher coupon because they think more cash flow means more sensitivity. Go back to the weighted-average-time definition and the zero-coupon limiting case, and the answer is forced.

   Expect next

   • So what is the maximum duration a 10-year bond can have?
   • What if one of them is callable?
   • Why did long sovereign bonds fall so hard in 2022?
3. 016A bond has modified duration of 7 and convexity of 60. Rates rise 150 basis points. What happens to the price, and how wrong is the duration-only answer?Duration and convexityHardtechnicalFixed income asset managementSyndicate desks

   Say this

   Duration alone says minus 10.5 percent. Convexity gives back 0.675 percent, so the estimate is about minus 9.8 percent. Duration on its own overstates the loss by roughly 68 basis points of price, and the error grows with the square of the move.

   Then walk it

   1. First term: minus duration times the yield change, so minus 7 times 0.015, which is minus 10.5 percent.
   2. Second term: half times convexity times the change squared, so 0.5 times 60 times 0.015 squared. That is 0.5 times 60 times 0.000225, which is 0.00675, or plus 0.675 percent.
   3. Net estimate about minus 9.83 percent. On 100 million of face at par that is a 9.8 million loss rather than 10.5 million — a 675,000 dollar difference from one term.
   4. Note the asymmetry: if rates had fallen 150 basis points, you would gain 10.5 plus 0.675, so 11.17 percent. Convexity helps in both directions, which is why it has value.
   5. Scale it: at a 50 basis point move the convexity term is worth only 7.5 basis points and you can ignore it. At 300 basis points it is 2.7 percent and you cannot. The error is quadratic in the move.
   6. The caveat to state: this is still a two-term Taylor expansion off a single yield. It assumes a parallel shift in the curve, and for a real portfolio you would run key-rate durations instead, because curves twist rather than shift.

   Where candidates lose it

   Dropping the one-half, or forgetting to square the yield move. Both are common and both produce an answer that is wildly wrong. Write the formula out loud before you compute, and state the units — decimals, not percentages — before you multiply.

   Expect next

   • Now do it for a 300 basis point move.
   • What if the curve steepens rather than shifts?
   • What are key rate durations?
4. 047How much would you pay for 2x your money on a 12 percent PIK security with no compounding?Credit modellingHardsuperdayApollo Global ManagementGeneralist · New York · 2019

   Say this

   You need the holding period. With simple 12 percent accrual, the instrument is worth 100 plus 12 per year, so it reaches 200 at a bit over 8.3 years. If you want 2x in five years, the accrued value is only 160, so you must buy at 80.

   Then walk it

   1. Simple accrual means the balance is 100 plus 12 times the number of years. No compounding, so it is linear, not exponential.
   2. For 2x with no discount, solve 100 plus 12t equals 200. That gives t equals 8.33 years. So if you pay par and hold to maturity, you double in a shade over eight years.
   3. If the hold is fixed, you solve for price instead. Five-year hold: terminal value is 160, and you want 2x, so entry price is 80. Three-year hold: terminal value 136, entry price 68.
   4. Sanity-check the implied return, because that is what the interviewer wants. 2x over five years is a 14.9 percent IRR; over three years it is 26 percent. State which one you are quoting.
   5. Now the real-world qualifications, which is where the marks are. PIK usually compounds, and at 12 percent compounding you double in 6.1 years by the rule of 72, not 8.3. So confirm the accrual convention before you answer.
   6. And the credit qualification: doubling requires the borrower to repay a balance that has grown 60 to 100 percent while never paying you cash. So the recovery question is whether enterprise value grows faster than the accrual. If it does not, the accrued claim is above the value and your recovery is capped well below the accreted number.

   Where candidates lose it

   Assuming compounding when the question explicitly says none — that turns 8.3 years into 6.1 and you have answered a different question. And answering without asking for the holding period, since price and horizon are two unknowns in one equation.

   Expect next

   • Now assume it compounds. How does the answer change?
   • What IRR is 2x over five years?
   • What has to be true about enterprise value for you to get repaid?

   Reported by candidates at Apollo Global Management (Generalist, New York, 2019). Source: Wall Street Oasis.

5. 092Roughly how large is India's outstanding corporate bond market? Reason it out.Indian debt marketsIntermediatetechnicalIndian debt capital marketsCredit research

   Say this

   Around 50 to 55 lakh crore rupees, so roughly 600 to 650 billion dollars, which is a bit under 20 percent of GDP. Get there from GDP: India's GDP is about 330 lakh crore rupees, and the corporate bond market has been running in the high teens as a share of it.

   Then walk it

   1. Anchor on GDP first, because it is the number you are most likely to know: about 330 lakh crore rupees, or roughly 4 trillion dollars.
   2. Then the ratio. Bank credit to the commercial sector in India is roughly 50 to 55 percent of GDP, and the corporate bond market is well under half that. High teens as a percentage of GDP gets you to 50 to 60 lakh crore.
   3. Cross-check with the issuance flow. Annual private placement issuance has been running around 8 to 10 lakh crore rupees, and with an average tenor of roughly five years, the steady-state outstanding stock is five times annual issuance, which lands you in the same 40 to 50 lakh crore region. Two independent routes agreeing is the point of the exercise.
   4. Comparison for scale: the US corporate bond market is over 100 percent of GDP, and even Malaysia and Korea run far higher ratios than India. So India is an outlier low, which is the substance behind the depth problem.
   5. Composition matters as much as size: the large majority is private placement rather than public issue, and financial sector issuers are a very large share of it. So the genuine non-financial corporate bond market is a good deal smaller than the headline.
   6. And be honest about the range. I would quote it as roughly 50 lakh crore, say which year, and say I would confirm from the SEBI or RBI data rather than pretend to a precise figure.

   Where candidates lose it

   Guessing a number with no derivation. The estimation route is what is being marked, so anchor on GDP, apply a ratio, then cross-check against issuance flow times average tenor. And use consistent units — mixing lakh crore and billions of dollars mid-answer loses the interviewer.

   Expect next

   • How does that compare to bank credit in India?
   • Why is it so much smaller than the US market?
   • How much of it is financial sector issuance?

Firm tags come from public, anonymous candidate reports on Wall Street Oasis: strong signal, not sworn testimony. Firms are named as the places a question was reported, not as partners of Fin Maverick. Answers are written for this page to show how to think out loud; they are not scripts to recite.