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. 077How do you view the market today?Trading and marketsIntermediatetechnicalMizuhoSales and Trading · New York · 2026

   Say this

   Give a structure, not a survey. I would frame it as: here is what the market is currently pricing, here is where I think that pricing is wrong, and here is the trade. Three levels quoted from memory, one view, one thing that would change my mind.

   Then walk it

   1. Open with the pricing, because that is how a trader talks. Where the policy rate is and how many cuts or hikes the curve has in it over the next year. Where ten-year yields are. Where index implied volatility is against realised. Those three numbers tell the interviewer you look at screens.
   2. Then the regime read in one sentence. Something like: the market is pricing a soft landing with the equity risk premium near its lows and volatility subdued, which means the price of being wrong on growth is unusually high.
   3. Then your actual view, stated as a disagreement. Not 'I am cautious' — rather, 'I think the curve has too many cuts priced for the inflation prints we are getting, so I would be paid at the front end.'
   4. Then the derivatives expression, because this is a derivatives seat. If the view is that realised volatility will exceed the low implied, own gamma. If the view is directional with a date attached, use a spread rather than an outright.
   5. Then the falsifier. One data point or level that would make you abandon the view. This is the part that makes you sound like someone who has run risk rather than read commentary.
   6. And know your own numbers. If you quote a level you must be able to say where it was three months ago and what moved it. A wrong number said confidently is worse than saying 'roughly 4 and a quarter, I have not checked this morning'.

   Where candidates lose it

   Summarising the news. Everyone can say inflation is coming down and the Fed is data-dependent. The answer that gets a callback quotes three levels, names one disagreement with the market's pricing, expresses it as a trade, and says what would falsify it. And never invent a number you cannot defend.

   Expect next

   • So what trade would you put on?
   • Where were those levels three months ago?
   • What would make you change your mind?

   Reported by candidates at Mizuho (Sales and Trading, New York, 2026). Source: Wall Street Oasis.

2. 078What is the current market sentiment?Trading and marketsIntermediatetechnicalNomuraGlobal Markets · New York · 2026

   Say this

   Answer it with positioning and prices rather than adjectives. Sentiment is observable: where implied volatility sits versus realised, how steep the skew is, what the put-call ratio and futures positioning look like, and how credit spreads are behaving relative to equities. Then say whether sentiment and fundamentals are pointing the same way.

   Then walk it

   1. The measurable sentiment indicators I would name: index implied volatility and its term structure, the skew or 25-delta risk reversal, high yield credit spreads, and CFTC or equivalent positioning data on the major futures.
   2. The most useful single read is often implied versus realised volatility. A low VIX with even lower realised volatility means complacency is cheap; a low VIX against rising realised means the market has not caught up yet.
   3. Then a cross-asset check, because sentiment is only interesting when assets disagree. Equities at highs with credit spreads widening, or gold making highs with real yields also rising, tells you something is unresolved. Consistent moves across assets tell you less.
   4. Then the positioning point, which is where sentiment becomes tradeable: extreme one-sided positioning makes the market fragile to news that would otherwise be minor, because the marginal buyer is already fully invested.
   5. Say it in one line at the end: 'so the market is priced for a benign outcome with low protection demand, which means the payoff to owning tail hedges is better than usual even though nothing is obviously wrong.' That is a sentiment read that ends in a trade.
   6. And the caveat worth volunteering: sentiment is a terrible timing tool. Extremes can persist for quarters, and 'everyone is bullish' has been a losing short signal far more often than a winning one. I would use it for sizing and for hedging cost, not for entry.

   Where candidates lose it

   Answering with a feeling — 'cautiously optimistic', 'risk-on'. On a markets desk, sentiment means observable positioning and prices. Name four indicators, say what they currently show, and finish with what it implies for the cost of protection. Then admit it is not a timing signal.

   Expect next

   • Which single indicator would you rely on most, and why?
   • Where do assets currently disagree with each other?
   • Has extreme positioning ever been a good short signal?

   Reported by candidates at Nomura (Global Markets, New York, 2026). Source: Wall Street Oasis.

3. 079Why is crypto lagging gold even though both are supposed to be hedges?Trading and marketsHardsuperdayNomuraGlobal Markets · New York · 2026

   Say this

   Because they are not hedging the same thing. Gold is a hedge against monetary debasement and geopolitical risk, with central banks as a price-insensitive structural buyer. Bitcoin behaves empirically like a high-beta risk asset — it correlates with the Nasdaq and with liquidity conditions, not with fear. The 'digital gold' framing is a narrative, and the correlation data has never really supported it.

   Then walk it

   1. Look at the behaviour in stress. In March 2020, in the 2022 rate shock, and in most risk-off episodes, bitcoin fell with equities and often fell harder. Gold's drawdowns in the same episodes were smaller and shorter. That is not a hedge, that is a levered risk asset.
   2. The buyer base explains most of it. Central bank gold buying has been running at record levels since 2022, accelerated by the freezing of Russian reserves, which gave every non-aligned reserve manager a reason to hold an asset no one can sanction. That flow is price-insensitive and persistent.
   3. Crypto's marginal buyer is discretionary risk capital, plus ETF flows that are themselves procyclical. When liquidity tightens, that buyer disappears — which is precisely when a hedge is supposed to work.
   4. There is a real overlap in the thesis: both are non-sovereign stores of value with no yield. But gold has four thousand years of institutional acceptance, a central bank bid, and jewellery demand as a floor. Bitcoin has a fixed supply schedule and a much shorter track record, and its volatility is five to eight times gold's, which makes it unusable as a reserve asset regardless of the thesis.
   5. The honest possibility that it changes: as the holder base institutionalises, correlation could fall and behaviour could converge towards gold. There is some evidence of that in the post-ETF period. I would want several full cycles before believing it.
   6. So the way I would frame it for a client: gold is a hedge you hold and forget, crypto is a risk position with an option on monetary regime change. Sizing them the same way is the error, and calling them both hedges is how that error gets made.

   Where candidates lose it

   Accepting the premise that both are hedges and looking for a reason one is underperforming. Reject the premise: the correlation data says bitcoin is a risk asset. And name the central bank gold bid post-2022, because that is the specific flow story behind the divergence.

   Expect next

   • Could crypto's correlation profile change as the holder base institutionalises?
   • Why has central bank gold demand been so strong since 2022?
   • How would you size the two differently in a portfolio?

   Reported by candidates at Nomura (Global Markets, New York, 2026). Source: Wall Street Oasis.

4. 080How does AI affect equities and rates?Trading and marketsHardsuperdayNomuraGlobal Markets · New York · 2026

   Say this

   In equities it has concentrated the index and shifted the story from software margins to capital expenditure, which changes the quality of the earnings. In rates the channel is more interesting and less discussed: a genuine productivity shock raises the neutral real rate, and the capital spending itself is a large new demand for financing. So AI is arguably a steeper-curve, higher-real-yield story as much as an equity story.

   Then walk it

   1. Equities first, and the honest structural fact: index concentration is at multi-decade highs, with a handful of names driving most of the return. That makes the index itself a different instrument than it was — higher single-name risk inside a supposedly diversified product, which shows up as index volatility being low while dispersion is high.
   2. The earnings-quality shift matters for valuation. The hyperscalers moved from asset-light software economics to spending a large share of cash flow on data centres and chips. Depreciation follows with a lag, so reported margins face a headwind two to three years after the spending, and the return on that capital is the open question.
   3. The derivatives expression of that: correlation is low and dispersion high, so index volatility understates single-name risk. Being long single-name volatility and short index volatility — long dispersion — is the natural way to express scepticism without taking a directional view.
   4. Rates channel one: if AI genuinely raises productivity growth, the neutral real rate rises, which means the whole curve settles higher than pre-2020 assumptions and long-duration assets are structurally repriced.
   5. Rates channel two, which is nearer term: the capital expenditure is enormous and increasingly debt-financed, including a fast-growing data-centre securitisation and private credit market. That is a new, large supply of credit issuance, and it concentrates exposure to a single technology thesis inside the credit market.
   6. Where I would be honest: nobody knows if the productivity effect is real, and previous technology capital cycles — railways, fibre in 1999 — delivered the technology and destroyed the capital. So I would hold the equity view loosely, and note that the trade with the clearest logic is the dispersion trade, because it profits from the concentration being mispriced regardless of which way the thesis resolves.

   Where candidates lose it

   Giving a generic technology-optimism answer. On a Global Markets desk the differentiator is the rates channel — neutral rate plus financing supply — and the derivatives expression, which is the dispersion trade. And having the humility to name the fibre 1999 comparison keeps it from sounding promotional.

   Expect next

   • What is a dispersion trade and how would you put it on?
   • Why would AI raise the neutral rate?
   • What does the 1999 telecom build-out tell you about this one?

   Reported by candidates at Nomura (Global Markets, New York, 2026). Source: Wall Street Oasis.

5. 081What is crude trading at right now? And walk me through what has moved it this month.Trading and marketsIntermediatetechnicalRCRBC Capital MarketsSales and Trading · London · 2025

   Say this

   Give the level, name which contract you are quoting, then the two or three drivers with a direction and rough magnitude. And say the curve shape, because on a commodities or macro desk that is half the information.

   Then walk it

   1. Be specific about the instrument. Brent front month versus WTI front month are different numbers with a spread between them, and quoting one when asked about the other is the first mistake. Say 'Brent front month is around X' and give the WTI spread if you know it.
   2. Then the curve: is the front in contango or backwardation and by how much. That tells the interviewer whether inventories are tight, and it is the piece a generalist candidate never has.
   3. Then two or three drivers with direction and size. On the supply side: OPEC+ quota decisions and actual compliance, US shale production, and any outage or sanctions development. On the demand side: Chinese import data, refinery margins, and the growth outlook.
   4. Distinguish flow from fundamentals. A move driven by managed-money positioning unwinding is different from one driven by an inventory draw, and a trader should be able to say which they think it was.
   5. Then the technical level if asked, without pretending it is more than it is: 'the market has failed twice around X, and positioning is long, so a break below Y probably accelerates.' Say it as a description of where the stops are, not as a forecast.
   6. The honest framing if you genuinely do not know the level: say so, give your best estimate with a range, and say when you last checked. 'Brent was around the mid-60s when I looked yesterday' is a perfectly good answer. Inventing a precise number you cannot defend is the one thing that ends the conversation.

   Where candidates lose it

   Quoting a number without saying which contract, or being unable to name the curve shape. And never invent a level — an S&T interviewer knows the screen and will catch a fabricated number instantly. Have three or four markets you genuinely follow daily rather than a shallow view on everything.

   Expect next

   • Is the curve in contango or backwardation?
   • Where is the Brent-WTI spread and what drives it?
   • Was that last move flow or fundamentals?

   Reported by candidates at RBC Capital Markets (Sales and Trading, London, 2025). Source: Wall Street Oasis.

6. 083What do you think this index closes at by the end of the year?Trading and marketsIntermediatetechnicalMSMorgan StanleySales and Trading · Tokyo · 2025

   Say this

   Give a number and build it, do not dodge it. Decompose into earnings growth and multiple: start from current index earnings, apply a growth rate you can defend, apply a multiple with a reason, and you have a level. Then give a range and say what the options market is currently implying, because that is the market's own answer.

   Then walk it

   1. The build: index level equals earnings times multiple. If earnings are growing 8 percent and the multiple is unchanged, you get 8 percent, and then you argue about the multiple — rates, risk premium and the growth outlook.
   2. Give one number and one range. 'My central case is up 6 to 8 percent from here, so roughly X, with a plausible band of minus 10 to plus 15.' Point forecasts are dishonest and no forecast is evasive; the number plus band is the professional answer.
   3. Then the market-implied cross-check, and this is where a derivatives candidate distinguishes themselves: the options market gives you a distribution for free. The at-the-money implied volatility annualised over the remaining period tells you the one standard deviation range the market is pricing, and the skew tells you the market's asymmetry.
   4. So you can say: the market is pricing about a plus or minus 12 percent one standard deviation range with a fat left tail, and my view is inside that range but with less downside than the skew implies — which is a trade, not just a forecast.
   5. Then the trade expression. If your view is modest upside with low volatility, sell a put spread or buy a call spread rather than buying outright calls. Matching the structure to the shape of the view is the point of being on a derivatives desk.
   6. And the risk to name: the level is driven by the multiple far more than by earnings over a one-year horizon, and the multiple is driven by rates and risk appetite, neither of which I can forecast. So the honest version is that my earnings number is a view and my multiple number is an assumption, and I would sensitise it.

   Where candidates lose it

   Refusing to give a number, or giving one with no construction. Both fail. Build it from earnings and multiple, then use the options market to give the range — that second step is free evidence and almost nobody does it.

   Expect next

   • What is the options market implying for the range?
   • Which part of your build are you least confident in?
   • How would you express that view in options rather than futures?

   Reported by candidates at Morgan Stanley (Sales and Trading, Tokyo, 2025). Source: Wall Street Oasis.

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.