Which Desk Trades WhatSome background helps

Eleven trading seats and six that sit next to them: what each one actually touches, the single number it lives by, and what a bad day looks like from that chair.

The one distinction to get right first

• Almost every seat on a markets floor is one of three things, and mixing them up is the fastest way to answer a question about the wrong job.
• Trading holds risk. The desk owns a position, is paid for warehousing it, and is measured on what that position does. Its day is bounded by limits.
• Sales holds relationships. It does not own the risk; it finds the client whose problem the desk's product solves, and it is measured on what that flow is worth to the desk.
• Structuring holds the design. It assembles an instrument out of simpler ones so that a client's exact requirement is met and the desk can still hedge what it has sold. Every autocallable and every credit-linked note came off a structuring desk before it reached anybody.
• Then there is everything that makes those three possible without holding the position: risk, quant and technology, operations, compliance and audit, treasury. The section at the bottom is about them, because those seats hire more people than the trading ones do and are far less written about.

Rates

• What it trades: government bonds, bond futures, interest rate swaps, OIS, short-term interest rate futures, FRAs, basis swaps, swaptions, linkers and inflation swaps.
• The number it lives by: DV01 — what the book gains or loses if the relevant rate moves one basis point. Positions are described in DV01 rather than notional, because notional means nothing across maturities.
• The mental model: a curve, not a rate. Everything is a view on the shape — this maturity against that one, this curve against another currency's. The yield curve and curve construction are the two pages to have read.
• What a bad day looks like: a central-bank surprise that moves the front end and the long end in opposite directions, so a position that was flat in duration terms is not flat at all. Reading a central-bank statement is the desk-relevant playbook.
• Deep dive: rates derivatives and fixed income.

Credit

• What it trades: corporate bonds, high yield, leveraged loans, single-name CDS, CDS indices, asset swaps, convertibles, CoCos.
• The number it lives by: spread — the compensation over the risk-free curve — and CS01, what a one-basis-point spread move is worth. A credit trader thinks in spread the way a rates trader thinks in yield.
• The mental model: two risks in one instrument. A corporate bond is a government bond plus a short position in the issuer's solvency, which is exactly why the asset swap and the CDS exist — to separate them. Credit spreads.
• What a bad day looks like: liquidity leaving before the fundamentals do. The bonds have not defaulted, but nobody will bid them in size, and the hedge in index CDS moves more than the cash position it is supposed to protect. That gap has its own name and its own page section.
• Case studies that belong to this desk: 2008, the 2023 AT1 write-down.
• Deep dive: credit.

FX

• What it trades: spot, forwards, FX swaps, NDFs, options and the exotics built from them — barriers, digitals, accumulators — plus cross-currency swaps where FX meets funding.
• The number it lives by: the position in each currency pair, and for the forward book, the points — the interest-rate differential expressed as a price. Almost nothing in FX forwards is a forecast; it is arithmetic on two deposit rates. Hedging covers why.
• The mental model: every price is a ratio, so there is no such thing as an absolute move. “The euro fell” is always “against something”, and the cross rates have to be consistent or somebody arbitrages them. The arithmetic is on the mental-math page.
• What a bad day looks like: a currency that has been quiet for years because a policy holds it there, and then the policy stops. The Swiss franc in 2015 is the case every FX desk knows.
• Deep dive: FX and FX derivatives.

Cash equities

• What it trades: shares, preferred stock, depositary receipts, ETFs, rights issues, and the stock borrow that makes shorting possible.
• The number it lives by: execution quality against a benchmark price. On an agency desk the position is meant to be flat by the close; what is measured is the difference between the price achieved and the price that was available.
• The mental model: microstructure. Where the order goes, who sees it, what the spread and the queue cost, and how a large order moves the price it is trying to get. Market microstructure and what happens when you press buy.
• What a bad day looks like: an index rebalance, an unexpected halt, or a crowded short with a shrinking borrow. 2021 is that last one in full.
• Deep dive: cash equities.

Equity derivatives

• What it trades: listed and OTC options, futures, equity swaps, total return swaps, variance swaps, dividend swaps, and the structured shelf built on top — autocallables, reverse convertibles, bonus certificates.
• The number it lives by: the Greeks, and above all gamma — how fast the hedge goes wrong. A delta-hedged book is not a flat book; it is a position in volatility and in time. Volatility.
• The mental model: the desk is usually short what it sold to clients and long the hedge, so it earns the spread and pays for the rebalancing. Whether that is profitable depends on whether realised volatility comes in under the implied volatility it charged.
• What a bad day looks like: a volatility spike that makes the hedge expensive exactly when it is largest. February 2018 is the canonical version; Archegos is what happens when the swap counterparty, not the market, is the risk.
• Bench tool: rebuild any of it in the strategy builder.
• Deep dive: equity derivatives.

Commodities

• What it trades: futures, options, swaps, metals spot, power, freight, weather, carbon allowances, exchange-traded commodity products.
• The number it lives by: the curve's shape, and the roll. In a market where storage costs money and delivery is physical, the difference between two contract months is the trade far more often than the level is.
• The mental model: this is the one asset class where the underlying can be a nuisance. Somebody has to take delivery, store it, insure it and ship it, and every one of those constraints shows up in the price.
• What a bad day looks like: a contract that reaches expiry with nowhere to put the physical, or a squeeze in a market too small to absorb one position. Negative oil in 2020 and nickel in 2022 are the two ends of that.
• Deep dive: commodities.

Securitised products and structured credit

• What it trades: MBS, CMBS, ABS, CLOs, CDOs and tranches, covered bonds.
• The number it lives by: the assumption set — default rate, recovery rate, prepayment speed — because the price is a function of them and there is no market price for the assumptions themselves.
• The mental model: a waterfall. The same pool of loans produces a safe claim and a dangerous one depending only on where you stand in the queue, and correlation decides how much of that safety is real. Securitisation.
• What a bad day looks like: the correlation assumption turning out to have been estimated on a sample containing no crisis. 2008, in one sentence.

Repo, money markets and the funding desk

• What it trades: repo and reverse repo, bills, commercial paper, CDs, money-market fund flows, and the FX swap when the funding need is in another currency.
• The number it lives by: the overnight rate it can actually borrow at, against the rate everybody quotes. The gap between those two is the whole job.
• The mental model: this is the plumbing every other desk depends on and nobody thinks about until it blocks. Collateral, haircuts, and who is obliged to post what by when. Margin and collateral and clearing and settlement.
• What a bad day looks like: everyone wanting the same collateral at the same time. Also the day a bank discovers that its assets were long and its funding was short — March 2023.
• Deep dive: money markets.

Alternatives, private markets and digital assets

• What sits here: private equity, private credit, venture capital, hedge funds, infrastructure, catastrophe bonds, and on the digital side crypto spot, perpetuals, stablecoins, tokenised bills.
• The number it lives by: for private markets, a valuation nobody trades against — which is why the analysis is on fundamentals rather than marks. For digital assets, funding rates and the exchange you are exposed to.
• The mental model: whatever cannot be sold has to be understood before it is bought, because there is no exit to correct the mistake with. That is the same sentence for a private loan and for a token.
• What a bad day looks like: an exchange that was also the custodian; a fund that gates redemptions exactly when redemption is what everyone wants.
• Deep dives: alternatives, digital assets.

Risk

• What it does: sets and polices the limits the desks trade inside, and answers the question “how much could we lose, and under what” before anybody has to find out.
• What you actually touch: the whole product range, but through exposures rather than trades — sensitivities, scenarios, stress tests, concentration, counterparty exposure.
• The pages that matter: risk measures, the arithmetic of drawdowns, how products fail, and every case study on the site — a risk interview is largely a conversation about them.
• The question this seat is really asked: not “what is the number” but “what does the number assume, and when does the assumption stop holding”.

Quant, strats and technology

• What it does: builds the models the desks price with, the systems they trade through, and the tooling that keeps both honest.
• What you actually touch: curve construction, volatility surfaces, pricing libraries, calibration, and the far less glamorous business of making sure yesterday's number can be reproduced today.
• The pages that matter: curve construction, volatility, valuation, and the calculators on this site, which are deliberately small enough to read end to end.
• The distinction worth knowing: a model that is wrong in a known direction is usable; a model that is wrong in an unknown direction is not. Most of the work is turning the second into the first.

Operations, settlement and collateral

• What it does: makes sure the trade that was agreed is the trade that settles — confirmation, matching, payment, delivery, margin calls, fails.
• Why it is underrated: this is where you learn what a product is in the only sense that is legally binding, and it is the fastest route to knowing more about instruments than people who trade them. Several of the failures on this site were operational before they were anything else.
• The pages that matter: clearing and settlement, corporate actions, margin and collateral, reading a broker statement.

Compliance, audit and control

• What it does: checks that what the floor does matches what it is permitted and required to do, and that the record of it would survive being read by somebody hostile.
• What you actually touch: client categorisation, suitability and appropriateness, disclosure documents, conflicts, market-conduct rules, and the paper trail behind each.
• The pages that matter: investor protection, reading a key information document, reading a term sheet, what the wrapper changes.
• The seat's own question: not “is this legal” but “could a reasonable client have understood what they were buying from the document they were given”.

Corporate treasury — the client side of all of it

• What it does: manages a non-financial company's cash, debt and exposures. It is the counterparty on the other side of a great deal of what the desks above sell.
• What you actually touch: FX forwards for receivables, swaps to fix a floating loan, commercial paper and money-market funds for the cash, factoring and trade finance for the working-capital gap.
• The mental model: you are not trying to make money on the instrument. You are trying to make a known cash flow out of an unknown one — which is why hedging and the cash-flow timing of a hedge matter more here than pricing does. Metallgesellschaft is the lesson in one case.

Choosing what to read for the seat you are actually going for

• Read the desk's asset-class page end to end, including the quiz at the bottom. That is one evening and it covers what the desk trades, who is on the other side, and what breaks.
• Then take the three instruments that desk touches most and get them to four-sentence fluency — the method is here.
• Then one case study from that market, told in three sentences: what the position was, what moved, why the exit was not available.
• If you do not yet know which seat you want, the role map works the other way round — from what you like doing to what the site holds for it.