Derivatives Foundation interview preparation

Questions

100

Traced to a firm

29

Firms

19

Updated

September 2026

Asked at

All firmsMSMorgan Stanley4Nomura4Akuna Capital2Amundi2HSBC2PIMCO2Bank of America1Barclays1Citadel1DRW1Goldman Sachs1Jane Street1Millennium Management1Mizuho1Old Mission Capital1RCRBC Capital Markets1Scotiabank1UBS1Wells Fargo Securities1

Topic

All topicsForwards and futures10Options basics8Option pricing7The Greeks10Volatility7Option strategies9Swaps and rates7Credit derivatives4Market structure and clearing6Indian derivatives8Trading and markets9Brainteasers6Fit9

Level

AnyCoreIntermediateHard

Type

AnyTechnicalCaseMarket viewBrainteaserFit

1. 040What does the volatility term structure normally look like, and what does it mean when it inverts?VolatilityIntermediatetechnicalVolatility tradingHedge funds

   Say this

   Normally upward sloping — front-month implied volatility below the longer dates — because volatility mean-reverts and the near term is usually calmer than the long-run average. An inversion means the market is pricing near-term stress: an event, a crisis, or a known catalyst inside the front expiry.

   Then walk it

   1. The mean-reversion logic: if spot volatility is 12 and the long-run mean is 18, then the average over the next two years should be closer to 18 than to 12. So the curve slopes up from a low starting point and slopes down from a high one.
   2. Inversions are therefore a stress signal. In March 2020 the front month went to 80 while one-year implied was in the 30s — the market said the next month is chaos and then it normalises.
   3. There is a milder, non-crisis inversion too: a known event inside the front expiry. An earnings date, an election, a central bank meeting, a court ruling. That produces a local bump in the specific expiry that contains it, not a whole-curve inversion.
   4. Trading it: a calendar spread expresses a view on the slope. Long the back month and short the front is long the term structure and short near-term gamma — that is a bet on mean reversion, and it is the standard trade after a spike.
   5. The carry side of the same fact: an upward-sloping VIX curve means long-volatility ETPs bleed on the roll, which is why products like the short-term futures trackers lose value over time even when spot volatility is unchanged. Roll cost has historically dominated their returns.
   6. The limitation: an inverted curve is information, not a signal. It usually resolves by front-month volatility collapsing, which is why selling the spike works most of the time — and the times it does not, it goes much higher first, which is enough to end a career. Knowing the base rate is not the same as being able to hold the position.

   Where candidates lose it

   Saying inverted means fear and stopping. Add the two distinct causes — systemic stress versus a datable event inside one expiry — and add the roll-cost consequence for volatility ETPs. And be honest that selling an inverted curve is a high-base-rate, high-tail-risk trade.

   Expect next

   • How would you trade an inversion, and how would you size it?
   • Why do long-volatility ETPs lose money in a normal market?
   • What does a bump in one single expiry tell you?
2. 041What is the VIX, and how is it actually computed?VolatilityIntermediatetechnicalVolatility tradingIndian derivatives desks

   Say this

   It is the market's 30-day expected volatility on the S&P 500, expressed annualised in percentage points. It is not an average of implied volatilities — it is built from a strip of out-of-the-money option prices across all strikes, which makes it a model-free estimate of the square root of expected variance.

   Then walk it

   1. The construction: take all reasonably liquid out-of-the-money puts and calls in the two expiries that straddle 30 days, weight each by one over its strike squared, sum, interpolate to exactly 30 days, take the square root and annualise.
   2. The one-over-strike-squared weighting is the part worth knowing. It comes from the mathematics of replicating a variance swap: a portfolio of options weighted that way has a payoff equal to realised variance. So the VIX is the fair strike of a 30-day variance swap, quoted as a volatility.
   3. That is why it is called model-free. It does not use Black-Scholes at all — it reads variance straight off prices. The older VIX methodology up to 2003 did use at-the-money Black-Scholes implieds, and the change matters when you compare long histories.
   4. Because it uses every strike, it is sensitive to the wings. A bid in deep out-of-the-money puts lifts the VIX even if at-the-money implied volatility has not moved, which is why the VIX can rise on a flat day.
   5. You cannot trade the index. You trade VIX futures and options, and the futures are priced off forward variance rather than spot VIX, so they do not track it one for one. In a spike, the front future will lag spot VIX substantially.
   6. India's equivalent is the India VIX, built on Nifty options with the same methodology adapted by the NSE. It matters for the same reason: it is the reference for weekly-expiry positioning and it inverts in exactly the same way in a stress event.

   Where candidates lose it

   Calling it 'an average of option implied volatilities'. It is a variance-swap strike, and the one-over-K-squared weighting is the specific thing the interviewer is checking. Also expect the follow-up on why VIX futures do not track spot VIX — have the forward-variance answer ready.

   Expect next

   • Why doesn't a VIX future track spot VIX?
   • Why can VIX rise on a day when at-the-money implied is unchanged?
   • What is India VIX built on?
3. 045Straddle or strangle — how do you choose?Option strategiesIntermediatetechnicalProp trading firmsVolatility trading

   Say this

   Both are pure volatility positions with no directional view. A straddle buys the at-the-money call and put, so you pay more and get maximum gamma right where spot is. A strangle buys out-of-the-money strikes, so it costs less and needs a bigger move, but it gives you more exposure to the tails per rupee spent.

   Then walk it

   1. Straddle: highest gamma and vega concentrated at the strike, highest premium, highest theta bleed. You want it when you expect a move and you expect it soon, and when you will delta hedge to harvest the path.
   2. Strangle: cheaper, wider break-evens, lower theta per day. You want it when you expect a large move but are unsure of timing, or when you specifically believe the wings are underpriced relative to the body.
   3. The wings-versus-body choice is a skew and kurtosis view, not just a cost decision. Long strangle, short straddle is a butterfly — that is a pure bet that the distribution is fatter-tailed than the smile implies.
   4. Break-even arithmetic: a 4.5 percent straddle needs a 4.5 percent move by expiry. A strangle costing 2 percent with strikes 5 percent out needs a 7 percent move. So the strangle wins only in the big-move scenarios and loses in the moderate ones.
   5. In practice the choice is often dictated by liquidity and margin. In Indian index options the out-of-the-money weekly strikes are extremely liquid and cheap in absolute rupee terms, which is why retail gravitates to strangles — and why the margin framework treats short strangles more punitively after the 2020 peak-margin reforms.
   6. The limitation for both: if you are not delta hedging, you are betting on the terminal price, not on volatility, and a stock that swings wildly and closes flat pays you nothing. State which trade you are actually putting on.

   Where candidates lose it

   Framing it purely as 'strangle is cheaper'. The real distinction is where you want your gamma and whether your view is about the body or the tails of the distribution. And name the unhedged-versus-hedged difference, because otherwise you are describing a direction bet.

   Expect next

   • Long strangle against short straddle — what have you built and what is the view?
   • Which would you rather own into an earnings print?
   • What does margin treatment do to the choice in India?
4. 048Explain a butterfly and an iron condor, and tell me what view each expresses.Option strategiesIntermediatetechnicalProp trading firmsVolatility trading

   Say this

   Both are short-volatility, range-bound structures with capped losses. A butterfly is short the body and long the wings — sell two at-the-money options, buy one either side. An iron condor is the same idea with a gap in the middle: sell an out-of-the-money put and call, buy further-out ones as protection. The view is that the underlying stays in a range and that implied volatility is too high.

   Then walk it

   1. Butterfly payoff: maximum profit if the underlying pins the middle strike at expiry, losses limited to the width less the credit. It has the highest payoff concentration of any standard structure, which is why it is the expiry-day trade of choice.
   2. Iron condor: a wider profit plateau between the two short strikes, smaller maximum profit, higher probability of ending inside the range. It is the same trade with less precision required about where the underlying lands.
   3. Read a long butterfly as short kurtosis. You are selling the body and buying the wings, which is a statement that the distribution is thinner-tailed than the smile implies. That is why butterfly prices are how FX desks quote the curvature of the smile.
   4. Both are short gamma and short vega in the middle, so they make money from time passing and from implied volatility falling. Both have their worst outcome on a large move in either direction, which is bounded by the long wings.
   5. Where they actually get used in India: Nifty and Bank Nifty weekly expiries, because the short-dated theta is large and the structures are margin-efficient once the wings are in place. That is also where they are most frequently oversized by retail traders.
   6. The real risk is the one the payoff diagram hides: the position is fine at expiry and can be badly underwater before it. A move to the edge of the range mid-life produces a mark-to-market loss and a margin call, and traders get closed out of positions that would have been profitable if held. Capped loss is not the same as capped margin.

   Where candidates lose it

   Drawing the payoff diagram and stopping. Two things earn the answer: naming the butterfly as a curvature or kurtosis trade, and pointing out that a capped-loss structure can still force you out early through margin. That second point is where retail traders in weekly options actually lose.

   Expect next

   • Why do FX desks quote the smile using butterflies?
   • Which is safer for a retail trader, and does the margin agree with you?
   • What happens to your iron condor two weeks in with spot at the short put strike?
5. 049What is a calendar spread and what are you really trading?Option strategiesIntermediatetechnicalVolatility tradingMarket making

   Say this

   Same strike, two expiries. Long the back month and short the front is a long calendar: you are long vega, short gamma, and long the term structure. What you are really trading is the slope of the volatility curve plus the difference between short-dated and long-dated realised volatility.

   Then walk it

   1. Positioning: the front month has most of the gamma and theta, the back month most of the vega. So long the back and short the front collects theta from the front and stays long vega on the back.
   2. The classic use is after a volatility spike. The front month is at 60, the back at 30, so you sell the front and buy the back, betting on mean reversion in the near term rather than on the level of volatility.
   3. It is also an event trade in reverse. If an earnings date sits in the front expiry, the front implied is inflated by the event. Selling the front and buying the back captures the event premium if the print is quiet.
   4. Risks are asymmetric and this is the part to get right. Short front-month gamma means a large move immediately is very painful, because the front option's gamma dwarfs the back's. The position is long volatility in vega terms and short it in gamma terms, and those two can lose at the same time.
   5. Roll and pin risk at the front expiry are real operational issues. You have to manage the front leg through settlement, and if it finishes at the strike you have a pin problem on one leg of a position you intended to hold.
   6. The honest limitation: a calendar spread is a term-structure view, so you can be exactly right about the volatility level and lose because the curve moved in parallel rather than flattening. Calendars are best sized small and judged on the spread between the two implieds, not on either leg alone.

   Where candidates lose it

   Describing a calendar as 'selling time decay'. It is a term-structure trade with opposite signs on gamma and vega, and the danger is an immediate large move against short front-month gamma. Naming the post-spike mean-reversion use case shows you know why anyone puts it on.

   Expect next

   • Which leg holds your gamma and which your vega?
   • What happens if the market gaps the day after you put it on?
   • How would you use a calendar around an earnings date?
6. 053How do you price an interest rate swap?Swaps and ratesIntermediatetechnicalRates derivativesQuant trading

   Say this

   Value each leg as a bond and take the difference. The fixed leg is a set of known cash flows discounted on the curve; the floating leg is worth par at each reset, so it discounts to a simple expression. The swap rate is the fixed rate that makes the two legs equal, so the swap is worth zero at inception.

   Then walk it

   1. Fixed leg: sum the fixed coupons, each discounted by the appropriate discount factor. That is just a bond price without the principal.
   2. Floating leg: the elegant result is that a floating-rate note resets to par at every coupon date, so the whole floating leg is worth notional times the difference between two discount factors. Alternatively, project each forward rate off the curve and discount it — same answer.
   3. Set them equal and solve for the fixed rate. The swap rate is a weighted average of the forward rates over the life, with the discount factors as weights.
   4. Post-2008 the important refinement is that you use two curves, not one. You project the floating rate off the relevant forward curve — SOFR or MIBOR — and you discount on the curve that matches the collateral you actually post, which for a cleared swap is the overnight rate. That is OIS discounting, and it materially changed swap valuations in 2009 to 2010.
   5. Why that matters: before the crisis everyone discounted at LIBOR and assumed one curve served both purposes. The crisis blew out the basis between overnight and term rates, revealing that discounting must reflect the funding cost of the collateral, not the index on the floating leg.
   6. Then the valuation adjustments for an uncollateralised swap: CVA for the counterparty's default risk, FVA for the cost of funding an unhedged position, and for the dealer, capital charges. A swap with a corporate that posts no collateral prices meaningfully away from the cleared mid, and that gap is a real charge, not a spread grab.

   Where candidates lose it

   Giving the single-curve textbook answer. Any rates interviewer will follow up on OIS discounting and the multi-curve framework, because that is the actual market practice since 2010. Also be ready to say why a floating leg is worth par at reset — the argument, not the formula.

   Expect next

   • Why do you discount on a different curve from the one you project on?
   • Why is the floating leg worth par at each reset?
   • How would the price change for an uncollateralised corporate counterparty?
7. 054How does duration affect interest rate risk?Swaps and ratesIntermediatetechnicalPIMCODebt Capital Markets · San Diego · 2026

   Say this

   Duration is the sensitivity of a bond's price to a change in yield, in years. A duration of 7 means a 100 basis point rise in yields costs you roughly 7 percent of value. So duration is not a description of the bond's maturity — it is the size of your interest rate exposure, and it is what you hedge.

   Then walk it

   1. Macaulay duration is the weighted average time to receipt of the cash flows. Modified duration divides that by one plus the yield, and it is the number you use for price sensitivity.
   2. The working formula: percentage price change is approximately minus modified duration times the yield change. Add convexity for large moves — plus a half times convexity times the yield change squared — because the price-yield relationship is curved, not linear.
   3. Convexity is your friend as a bondholder: it means you lose less on a rate rise than duration alone predicts and gain more on a fall. Which is also why convexity costs something in the price.
   4. The drivers: longer maturity, lower coupon and lower yield all raise duration. A zero-coupon bond's duration equals its maturity, which is the cleanest case and the reason zeros are the sharpest rate instrument.
   5. How it gets used: dollar duration, meaning duration times market value, is what you actually hedge. If a 500 million portfolio has duration 7, its dollar duration is 35 million per 100 basis points, and you short enough bond futures or pay fixed on enough swap notional to offset it.
   6. The limitation to volunteer: duration assumes a parallel shift in the curve. Real curves steepen, flatten and twist, so a duration-matched portfolio can still lose money on a curve move. That is why desks look at key rate durations bucketed along the curve rather than one number. And for callable or mortgage-backed bonds, duration itself changes with yields — negative convexity — so the static number misleads exactly when you need it.

   Where candidates lose it

   Defining duration as average time to cash flows and stopping. The question asks about risk, so lead with the sensitivity reading and the dollar duration hedge. And name the parallel-shift assumption — a bond manager will expect key rate durations to come up.

   Expect next

   • What does convexity add?
   • How would you hedge the duration of a 500 million portfolio?
   • Where does duration break down as a risk measure?

   Reported by candidates at PIMCO (Debt Capital Markets, San Diego, 2026). Source: Wall Street Oasis.

8. 055What is effective duration, and when would you use it instead of modified duration?Swaps and ratesIntermediatetechnicalAmundiRates · London · 2018

   Say this

   Effective duration is measured rather than derived: you shock the whole yield curve up and down by a small amount, reprice the bond with its options and cash flow rules intact, and read the sensitivity off the two prices. You use it whenever the cash flows themselves depend on rates — callables, putables, mortgages, floaters — because modified duration assumes they do not.

   Then walk it

   1. Formula: price down minus price up, divided by twice the initial price times the size of the shock. It is a numerical derivative, which is the whole point — you are not assuming a closed form.
   2. Modified duration is computed from fixed, known cash flows. The moment a bond is callable, the issuer's option changes the cash flows as rates move, so the analytical number is simply wrong.
   3. Callable bonds are the classic case. Rates fall, the call becomes likely, the expected life shortens, and duration falls — so the bond's price rise is capped. That is negative convexity, and effective duration captures it while modified duration cannot.
   4. Mortgage-backed securities are the extreme version, because prepayment behaviour is the option. Effective duration on an MBS moves sharply with rates, which is why convexity hedging by mortgage portfolios amplifies rate moves in the Treasury market.
   5. It is also the right measure for a floating-rate note, where the coupon resets. A floater has a long maturity and an effective duration of months, because its price barely responds to a level shift in rates.
   6. The caveat worth adding: effective duration is model-dependent, since repricing a callable requires an assumption about volatility and about how the issuer exercises. Two houses will produce different effective durations for the same bond, and the difference is a model choice rather than a data error. And the parallel-shift assumption is still in there — key rate durations are how you get past it.

   Where candidates lose it

   Treating effective and modified duration as synonyms, or defining effective duration with a formula but no reason to prefer it. Name a bond with embedded optionality — callable or mortgage — and say that its cash flows move with rates. That is the whole distinction.

   Expect next

   • What is the effective duration of a floating-rate note?
   • Why does a callable bond have negative convexity?
   • How does MBS convexity hedging move the Treasury market?

   Reported by candidates at Amundi (Rates, London, 2018). Source: Wall Street Oasis.

9. 057How does a currency swap differ from an interest rate swap, and who uses one?Swaps and ratesIntermediatetechnicalFX derivativesCorporate treasury

   Say this

   In a currency swap the notionals are in two different currencies and they are exchanged — at the start, at the end, or both — and the interest payments are made in their own currencies. That makes it a funding instrument as much as a rate instrument, and it carries principal risk that a plain rate swap does not.

   Then walk it

   1. Structure: you pay rupee interest on a rupee notional and receive dollar interest on a dollar notional, with the two notionals exchanged at the outset at spot and re-exchanged at maturity at the same rate. So you have both rate and currency exposure managed in one trade.
   2. The classic use: an Indian company issues a dollar bond because the market is deeper and cheaper, then swaps it into rupees so its liability matches its rupee revenues. It has raised offshore and ends up with a domestic-currency obligation.
   3. The reverse flow is what drives the market: foreign issuers raising in a currency because the swap back to their home currency is cheap. That relative cheapness is the cross-currency basis, and it is a funding indicator, not just a spread.
   4. Cross-currency basis is worth naming. In theory covered interest parity should hold and the basis should be zero. Since 2008 it has been persistently negative for dollar funding, because of balance sheet constraints on dealers and structural dollar demand. That persistent deviation is one of the clearest post-crisis examples of an arbitrage that is not arbitragable.
   5. Risk profile: because principal is exchanged, the credit exposure is far larger than on a rate swap of the same notional. A currency swap can be deeply in the money on the principal leg alone, which is why they carry heavier collateral and capital treatment.
   6. The limitation to say: the hedge is only clean if the maturity matches the exposure. Companies that rolled short-dated currency hedges against long-dated dollar debt have been caught by both the basis widening and the rupee depreciating at the same moment, which is precisely the correlation a short hedge fails to capture.

   Where candidates lose it

   Saying the difference is 'two currencies' without noting that principal is exchanged, which is the reason the credit exposure is much larger. And if you are interviewing anywhere India-facing, have the Indian issuer swapping dollar debt back into rupees as your example, plus the cross-currency basis.

   Expect next

   • What is the cross-currency basis and why is it not zero?
   • Why is the credit exposure bigger than on a rate swap?
   • How would an Indian corporate hedge a ten-year dollar bond?
10. 058Walk me through the move from LIBOR to SOFR. What actually changed, and what broke?Swaps and ratesIntermediatetechnicalRates derivativesCorporate treasury

    Say this

    LIBOR was a survey of what banks said they could borrow at unsecured; SOFR is a volume-weighted average of actual overnight repo transactions secured by Treasuries. So the benchmark moved from judgement to data, and from unsecured term lending to secured overnight. The two hardest consequences were the loss of credit sensitivity and the absence of a forward-looking term rate.

    Then walk it

    1. Why LIBOR had to go: it was a submission, not a transaction. After the 2012 manipulation scandals and the collapse of real unsecured interbank lending, the rate was being set on a market that barely existed. Regulators ended it rather than patch it.
    2. What SOFR is: roughly a trillion dollars a day of actual repo transactions, secured, overnight. Essentially unmanipulable because of the volume, which was the whole design goal.
    3. First problem — no credit spread. LIBOR rose when bank funding stress rose, so a bank lending at LIBOR was naturally hedged to its own funding cost. SOFR is secured, so in a crisis it falls while bank funding costs rise. That is a real basis risk for lenders, and it is why credit-sensitive alternatives like BSBY and AMERIBOR appeared, and mostly failed to gain traction.
    4. Second problem — no term structure. LIBOR was quoted for three and six months in advance; SOFR is an overnight rate you only know in arrears. The market solved it with compounded-in-arrears conventions, plus CME Term SOFR for loan markets where borrowers need to know the coupon at the start of the period.
    5. Transition mechanics: ISDA's 2020 fallbacks protocol amended existing contracts to fall back to compounded SOFR plus a fixed credit adjustment spread — 26 basis points for three-month USD. That spread number is worth knowing, because it is the price the industry put on the credit component.
    6. India ran a parallel version. MIBOR-based products and the shift to overnight-indexed benchmarks, with the RBI pushing users off LIBOR references in external commercial borrowings by mid-2023. The same two issues appeared: no credit sensitivity and the need for a term rate. The unfinished part everywhere is the long tail of legacy contracts — loans, securitisations and bonds with fallbacks that were drafted for a temporary LIBOR outage, not a permanent end.

    Where candidates lose it

    Describing the change as 'LIBOR was manipulated so they replaced it'. The interviewer wants the two structural consequences: no credit sensitivity and no forward term rate. Knowing the credit adjustment spread exists, and roughly what it was for three-month USD, is what marks this out as real knowledge.

    Expect next

    • Why does the loss of credit sensitivity matter to a bank lender?
    • How does a compounded-in-arrears coupon work in practice for a borrower?
    • What happened to legacy contracts with no proper fallback language?

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.