Getting the Interview

There is a version of the job search that almost every student runs. You polish the CV, you write the cover letter, you find the graduate portal, you upload, you wait. Then you do it thirty more times and you wait some more.

That process has a response rate close to zero, and the reason is not your CV.

This piece is about why the process fails, where the seats actually are, and the three or four things that change your odds by an order of magnitude. None of them are hard. Most of them take under a minute each.

Part 1 — Why it is hard

There are three reasons. Two of them are structural: you cannot change them, you can only route around them. The first one below is different. It is entirely within your control, almost nobody fixes it, and it is the one that actually costs you the offer.

1. You cannot explain how a desk makes money

This is the single most important thing in this document. Everything else here improves your odds at the margin. This one decides whether you convert an interview once you have it, and it is the reason most candidates who look good on paper do not get an offer.

It is not a trick question. It is the first thing anyone on a desk wants to know about you, and it is the thing no academic course teaches.

Most candidates can price a straddle. Almost none of them can say how the desk turns that price into P&L. So they fall back on the only story they have, which is a market view: I think this name is going to fall, so I would short it. The moment you say that in a market-making interview, you have told the interviewer you do not know what the job is.

A market maker does not make money by being right about direction. There are three real sources of revenue, and none of them are taught in an academic course:

• Bid-offer spread. Quote below fair value, quote above it, capture the difference. Volume times spread.
• Hedging efficiency. When a client buys the call, you hedge with whatever instrument is cheapest and most liquid. No individual hedge is perfect; across thousands of trades the imperfections wash out, and being systematically better at this than the next bank is the edge.
• Funding and balance sheet. You hold inventory overnight. It consumes capital, it earns or pays repo and rates. On a large book this is often the biggest single line.

If you can talk about any one of the following with a concrete number attached, you are immediately in a different pile: the implied-versus-realised spread, skew, correlation sold through worst-of structures, dispersion, or a term-structure dislocation. One of them. With numbers.

If you fix nothing else after reading this, fix this. It takes a weekend of reading and it is the difference between being one of five thousand CVs and being someone the desk can imagine sitting next to.

2. You are all applying to the same twelve doors

Every student in every business school in France, the UK and the US is sending the same CV to the same desks. Goldman, JPMorgan, BNP, SocGen, BofA, Citi. Maybe a hedge fund or two if they are ambitious.

There are somewhere around forty to fifty trading, structuring and sales intern seats in CIB each year. There are thousands of applications for them. Even if you are top of your class, so is the person next to you in the queue, and so is the person at the top of every other class in Europe.

Apply to thirty banks and you are competing with five thousand people. That is the whole problem, and it is arithmetic, not merit.

3. The bottom of the pyramid is shrinking

The business that built sell-side careers between 2000 and 2015 — selling structured products to retail and private banking clients — is contracting. Three forces at once:

• Margins have collapsed. Every bank now offers the same bonus certificates, the same discount certificates, the same autocalls. When the product is identical everywhere, the only differentiator is price. A structure that carried a 3% margin now carries fifty basis points.
• Regulation removed the hiding places. MiFID II and PRIIPS force full cost transparency to retail clients. A client can compare issuers in seconds. There is nowhere left to put an invisible margin.
• Software ate the easy work. Pricing a vanilla bonus certificate used to take a junior an afternoon. It now takes an algorithm a few milliseconds. Headcount followed.

The consequence for you is specific: the value of a junior who can only do standardised work has gone to roughly zero. The remaining value sits at the top of the business, in bespoke institutional flow and in aggregate risk management, and neither of those is an entry-level job description.

Part 2 — Where the seats actually are

The interesting work is frequently outside the bulge bracket, and the competition there is a hundred times thinner.

• Asset managers. Amundi, AXA IM, BlackRock, Carmignac. AUM is at record highs, hiring runs every year in cycles far less brutal than a bank's, and they do not freeze in a downturn the way banks do. You also learn the buy side first, which is the harder direction to travel later. Banks like hiring people who already understand their clients.
• Market makers you have not heard of. Not Optiver, IMC and Flow Traders, which every student already knows about. The tier below them, and the regional ones.
• Insurance hedging desks. Every European life insurer runs a derivatives book to hedge variable annuities. Enormous volumes. Almost no applicants.
• Corporate hedging desks. Total, EDF, Airbus. They hedge billions in commodities and FX with desks that look exactly like a bank's, often pay well, and keep saner hours.

The career path nobody mentions: start in derivatives operations or risk at an asset manager, move to a derivatives overlay desk, then move to the sell side with two years of experience. You skip the analyst grind entirely and you arrive knowing what the client wants.

Add ten buy-side firms, five prop shops, three corporate treasuries and five insurance derivatives teams to your list. That single change does more for your odds than any rewrite of your CV.

Part 3 — The channel nobody uses: brokers

There is one group of people that students overlook entirely, and it is the group with the best information and the strongest incentive to help.

Inter-dealer brokers sit between the banks, executing and matching liquidity. They talk to every desk, every day. They know which desks are growing, which are cutting, who just resigned, and who is about to launch a new strategy. They know all of this before HR posts anything.

Names you will see: TP ICAP, BGC Partners, Tradition, Marex, Sucden, Aurel BGC, Tullett Prebon. There are also specialist equity-derivatives brokers such as Octo, Louis Capital and Makor, plus smaller boutiques across rates, FX, commodities and structured products.

Why they are worth more to you than a headhunter:

• They have a direct interest in placing you. A broker's job is to maintain the relationship with the desks they cover. Introducing a strong candidate to a head trader earns a favour, and that favour turns into flow later. Helping you costs them nothing and pays them something.
• They will be honest with you. A broker will tell you plainly that a desk is about to cut headcount, or that a bank is launching an exotic team in March and you should send your CV this week. You will not get that from anyone inside a bank.
• They work for free. A broker who likes you will forward your CV to five desks in an afternoon with a personal endorsement attached. No recruiter does that for nothing.
• They are easy to reach. Brokers are paid to be sociable. A polite, specific message from a student almost always gets a reply.

The approach is close to the one you use with a trader, with one deliberate difference: you can ask a broker for time, because giving it away is part of how they work.

Bonjour [Last name], I am a Master in Finance student specialised in equity derivatives. I noticed you cover [structured equity / vol / autocalls] at [firm]. I would like to start my career in this area. Would you have ten minutes after the close to talk about it? Best, [First name Last name]

Two things are doing the work here.

“After the close.” This is the line that separates you from every other student who writes to a broker. It says you know their day has a shape, that the session is not the moment, and that you are not going to interrupt them while they are working. It also removes the only real objection they have. Nobody has ten minutes at eleven in the morning. Almost everybody has ten minutes at half past five.

“I would like to start my career in this area.” Not I am looking for an internship, which asks them for something they cannot give you directly, and not could you tell me who is hiring, which asks them to do research on your behalf. You are telling them where you want to go and inviting them to react. A broker who covers that product will react, because that is the conversation they have all day with everyone else.

Note the difference from the trader message below. A trader gets one question with a yes or no answer, because ten minutes is a large ask for someone whose day you are interrupting. A broker gets the ten minutes, because relationships are the product they sell.

Add five to ten brokers to your list today. It is the most underused asset in any student's search.

Part 4 — The three-line rule

Picture the trader's Bloomberg on a normal morning. Brokers pinging prices every minute. Data vendors selling a platform. Headhunters fishing. Salespeople from other banks pitching. Dozens of internal chats. The unread count is in the hundreds before the open.

Now drop this into it:

Dear Sir, I hope this message finds you well. My name is […], I am currently a student in a Master of Finance programme, where I am specialising in Capital Markets. I have always been passionate about financial markets, especially derivatives and structured products. After conducting an in-depth analysis of your team and your bank's positioning in the equity derivatives space, I am convinced that I would be an excellent fit for…

By line three the chat is closed. Not because the message is rude or badly written. Because it takes too long to work out what you want.

A trader's time is measured in seconds. They work in an environment where a two-second hesitation on a quote costs money. Their brain filters noise instantly and without malice.

Here is the message that works. Three lines, and that is the entire thing.

Bonjour [Last name], I am a Master in Finance student specialised in equity derivatives. Are you currently looking for a Summer 2026 intern on the desk? Best, [First name Last name]

Read what it communicates without stating any of it:

• You know who they are, because you used their last name.
• You know what they do, because you matched their specialty.
• You know what you want, and it fits in six words.
• You respect their time, because you did not write twelve lines.
• You are easy to deal with, because you asked one question with a yes or no answer.

Their reaction is: this one I can answer in five seconds. Either “send me your CV” or “not right now, try again in November.” Both take them under ten seconds, and both are a reply. You have just done what two hundred other students failed to do.

Why brevity is itself the signal

On a desk, being precise and fast is the job. The trader who quotes quickly and accurately wins the trade. The salesperson who compresses a complex client request into two sentences wins the meeting. The junior who answers “what is the delta on that book” in three seconds without scrolling earns trust.

A three-line message says, without claiming it, that you already think in that rhythm. That signal is worth more than any paragraph about how passionate you are.

The twelve-line message says the opposite: I have never spoken to anyone on a desk, I do not know what their day looks like, and I am probably going to be exhausting to manage.

Cut anything that does not change the answer

Once you have three lines, look at them again and ask of every clause: does this change what the reader does next? If it does not, delete it.

Apply that test to the school name.

If you are writing to an alumnus of your own school, the school is the whole hook and it goes first. It is the reason they will read past the greeting, and you should say it before you say anything else.

Everywhere else, it is doing less work than you think. A broker in London, a market maker in Amsterdam, a trader who did not go to your business school: to them, your school is a name that means nothing in particular. It sits in the message taking up one of your three lines and buying you almost nothing.

The specialisation is what the reader is filtering on. The institution is not. Every word in the message is being read by someone who is deciding, in about a second, whether to reply. Anything that does not push them toward yes is pushing them toward no. Say less. You will be read more.

Part 5 — Never lead with the code

There is one sentence that shows up in a large share of student messages, and it should not be there.

I have also built a Streamlit option pricer in Python, you can find it on my GitHub.

Delete it. Not because it is untrue, and not because building it was a waste of your time. Building it was probably the most useful thing you did all year. Delete it because in a first message it buys you nothing and it can cost you.

Coding stopped being a differentiator. Two years ago, saying you could write Python moved you into a smaller pile. It does not any more. Anyone can now produce serviceable code by asking for it. “I know Python” carries roughly as much information as “I know Excel”: everyone says it, most of them are telling the truth, and none of it separates you from the person below you in the inbox.

The evidence is all over LinkedIn. Scroll for ten minutes and you will find dashboards that would collapse at a thousand rows, repositories with CSV files committed straight into them, notebooks that only run on the author's laptop. The people posting them believe they are demonstrating capability. To anyone who has put something into production, they are demonstrating the opposite, and doing it publicly.

You will be judged on the ten percent that is hard to explain. Nobody on a desk is going to read your repository. If it comes up at all, they will find the one part that is genuinely difficult — the calibration, the barrier handling, the reason you structured the data that way — and ask you to explain it in plain language. That single exchange is the whole test. The other ninety percent is scaffolding and everyone in the room knows it.

When it does come up, unroll gently

If a trader, a broker or a salesperson asks what you can do technically, then you talk about it, in this order:

• The problem you were trying to solve.
• What you built.
• What broke, and what you changed as a result.

Never the reverse. Leading with the tool and arriving at the problem is how somebody describes a hobby. Leading with the problem and arriving at the tool is how somebody describes work. The third point is the one that convinces, because it is the only one an interviewer cannot get from your CV.

The rule underneath all of this

Our job is to make the complicated simple. You are not on the desk to prove that you diagonalise matrices in your head. You are there so that a salesperson, a client, or a risk manager without your background walks away actually understanding the risk. If you can explain the gamma profile of an autocall to somebody who has never traded, you are doing the job. If you can only explain it to somebody who already knows, you are not.

A smart person makes the complicated simple. Someone unsure of themselves does the opposite, because complexity is where they feel protected. Everyone on a floor can tell the difference within two minutes, and they are not being subtle about which of the two they want sitting next to them.

Part 6 — Match the desk to the request

The second mistake is sending one message to fifty people regardless of what they do.

If the person is a commodities trader, do not ask them for an equity derivatives internship. They will ignore you, and correctly, because there is nothing they can do for you. They do not hire on the equity desk and they have no visibility on its headcount. You have wasted their time and yours.

Finding out who someone is takes ninety seconds. LinkedIn is usually enough. The bank's structured products platform or investor relations pages sometimes list desk heads. Bloomberg's people search often shows the function. And if you know one student already inside the bank, ask them who covers what.

Then change one word:

Bonjour [Last name], I am a Master in Finance student with a focus on rates derivatives and swap pricing. Are you currently looking for a Summer 2026 intern on the rates desk? Best, [First name Last name]

“Rates derivatives” instead of “equity derivatives” is the entire personalisation, and it is enough. It tells the reader you know who they are, which puts you ahead of nearly all inbound.

Part 7 — Tone, hierarchy, timing

In French banking, hierarchy is real. The person you are messaging is five to fifteen years your senior and has a job you want. The relationship is asymmetric and your register should reflect that.

Vouvoyer by default. Bonjour, not Salut. Vous, not tu. Last name, not first name. This is not old-fashioned, it is professional. A trader who opens “Salut Pierre, tu cherches un stagiaire ?” from a twenty-two-year-old they have never met is annoyed before they finish the sentence.

Switch only when invited. They tell you to tutoyer, or they start using tu, or they sign off with their first name alone. Then you switch immediately, because continuing to vouvoyer after that reads as cold. Match their register, never lead it.

Send mid-morning or just after lunch. Not at the open, not into the close, not during an expiry. Fewer pings competing with you.

No follow-up before seven days. If they did not reply, they did not reply. One polite reminder after a week is acceptable. After that, move on and spend the energy on the next name.

If they say no, ask one more question. Thank them, then ask whether they know anyone else who might be hiring. A large share of the internships in this industry come through a referral from someone who first said no.

Part 8 — The hidden test

Every message is a small, unconscious interview. The reader is not deliberately scoring you, but their brain runs four checks in about a second and a half:

• Does this person waste my time?
• Do they know what my desk does?
• Can they communicate clearly under pressure?
• Are they going to be exhausting to manage?

Pass all four and you get a reply. Fail any one and your message joins the other fifty in the void. The three-line message is not a shortcut around the work. It is a demonstration that you already understand the job.

Part 9 — Volume, and what rejection actually is

Two things are true at once and you need to hold both.

The first: students from a school like yours land roles at the top of the industry every year. Goldman, Citi, Barclays, JPMorgan, BNP, SocGen, Natixis, Morgan Stanley, and the major asset managers. Even the cohorts who graduated into the post-Covid hiring freeze got through. Your school's alumni network is your single largest asset and you should use it without hesitation. Many of the people you message were in your seat two, five or ten years ago, and a fifteen-minute call with one of them can change the direction of your search.

The second: some of them will never reply, and your message will have been perfect. Not because they are unkind and not because you did something wrong, but because they are on a floor in the middle of a hundred things and yours slipped through.

Do not read that as rejection. It is noise, not signal, and you cannot tell in advance which message is going to be the one that lands.

A personal data point. I failed an interview at a bulge-bracket bank in April 2023. That desk closed at the end of 2025. The rejection I took personally at the time turned out to be information about a business, not about me.

The students who get in are not the ones with the best CV. They are the ones who sent fifty messages while everyone else sent five, and who treated every silence as data rather than as a verdict.

The checklist

Where you apply. Add ten buy-side firms, five prop shops, three corporate hedging desks and five insurance derivatives teams to your list. Add five to ten brokers as contacts, not as applications.

What you know before you message anyone. Be able to explain in sixty seconds how a desk turns a quoted price into P&L — this is the one that matters most. Be able to discuss one of implied-versus-realised, skew, correlation, dispersion or term structure, with a number. Know which desk your contact sits on before you type a word. Be able to explain the hardest part of anything you built, in plain language, to someone with no background.

How you write. Three lines: greeting, who you are, what you want. One question, with a yes or no answer. One word of personalisation matching their desk. Name the school only when it earns its place, which mostly means when you are writing to an alumnus. Never lead with Python, a pricer or a GitHub link — wait to be asked, then start with the problem. Vouvoyer, last name, until told otherwise. No life story, no passion paragraph.

Timing and persistence. Mid-morning or early afternoon, never the open or the close. With brokers, ask for the ten minutes after the close, never during the session. One follow-up after seven days, then stop. If they say no, ask who else might be hiring. Fifty messages, not five.

Do not copy this

Everything above is what worked for me and for the people I have watched get in. It is not scripture, and you should not treat it as such.

Some of you will do the opposite of what is written here and it will work. Somebody will send a long, funny, completely unstructured message to a head trader and get a reply within the hour, because it was the right message to the right person on the right day. That is good. That is how it should be.

Here is why the rules still matter. They are a floor, not a ceiling. What they buy you is that you stop making the mistakes that guarantee silence: the twelve-line introduction, the message to the commodities trader asking about equity, the “Salut Pierre”. Once those are gone, what remains is you, and you is the part that actually gets the reply.

Which brings the real risk of a document like this one. A message that has been copy-pasted reads exactly like a message that has been copy-pasted. The people you are writing to read hundreds of these; they detect a template in about a second. And the moment fifty students from the same school send the identical three lines to the same desk, those three lines stop working for all fifty of them. The structure is worth borrowing. The wording is not.

So take what is useful and rewrite the rest in your own voice. If a phrase in here feels wrong in your mouth, it will feel wrong in the reader's eye, and you should change it. Spontaneity is not a risk you are taking, it is information you are giving: somebody who writes naturally under constraint is somebody who will be able to talk to a client under pressure.

Learn the rules well enough that you stop thinking about them. Then write like yourself, and trust that. The person on the other side of that chat gets fifteen messages a day from people trying to sell them something. Be the one who is selling them back their own time. Three lines, one question, one answer, in your own words. That is how you get noticed in a place where attention is the scarcest thing on the desk.