What Investment Banking Is Actually Like

When you tell people you're in investment banking, what do they picture?

Wolf of Wall Street. Every single time. They think I'm on a trading floor yelling, or I'm doing something with stocks. Neither of those things is my job. Trading is a completely different division. I don't touch securities. What I do is closer to, like, management consulting meets financial modeling meets graphic design. I build pitch books, which are presentations that the bank uses to win advisory mandates from companies. And I build financial models, which are Excel spreadsheets that value companies based on their financial data. That's 80% of my life. PowerPoint and Excel. Not exactly what Leonardo DiCaprio was doing.

The other 20% is research. Pulling comparable company data from Capital IQ, reading 10-Ks and investor presentations, looking at precedent transactions in our sector. TMT is technology, media, and telecom. My coverage universe right now is mostly enterprise software companies and a couple of media conglomerates. When Vikram, my VP, needs a "market update" for a client meeting, that means I'm building a 15-page section showing every relevant M&A deal that's happened in our sub-sector over the last 18 months, with multiples and deal terms. That section takes me about six hours. It shows up on one slide during a one-hour meeting.

Walk through a specific day. Not a typical one, a real one.

OK, so, three weeks ago. Wednesday. I came into the office around 9:30, which is actually late for some groups but normal for TMT because we tend to work later at night. Checked my email. There were 47 unread messages. Twelve of those were from Vikram, who had been awake since 6 AM because he had a client call at 7. Most of Vikram's emails are action items disguised as casual observations. He'll write something like "interesting that their margins compressed 200bps in Q3" and what he means is "update the model to reflect margin compression and have the revised output to me by noon."

I worked on the model updates until about 12:30. Ate lunch at my desk, which is not a choice, it's just what happens. A turkey wrap from the deli downstairs. Then Charlotte pinged me on Teams. Charlotte is the associate I work under on most deals. She's second-year, very good, and she has this thing where she sends you a message that just says "hey" and then doesn't follow up for ten minutes. Which means you're sitting there for ten minutes wondering what's about to land on you.

What landed was this: the client CEO had called our managing director at 4 PM the day before. During a routine management meeting with his board, the CEO had decided he no longer wanted to pursue the acquisition target we'd been analyzing for three weeks. He wanted to explore a different company in a completely different sub-sector of enterprise software. Infrastructure monitoring instead of cybersecurity. Which meant the entire target section of our pitch book, about 40 pages, was useless. And we had a client meeting at 2 PM the next day.

So you had to rebuild 40 pages in less than 24 hours.

Eighteen hours, technically, because Charlotte told me at 1 PM and the meeting was at 7 AM the next morning for final review with the MD. The pages needed to be in his inbox by 7. Charlotte was on another live deal and couldn't help with more than reviewing my output. Vikram was traveling. It was me.

First thing I did was pull up Capital IQ and start searching for comparable companies in infrastructure monitoring. Needed at least eight good comps. Found eleven, narrowed to eight based on revenue size and business model similarity. Pulled their financials, built the comp table, calculated the trading multiples. That took about two and a half hours. Then I needed precedent transactions, which are M&A deals in the space. Found six. Pulled the deal terms, built that table. Another hour and a half.

Then the DCF. I had to build a discounted cash flow model for the new target from scratch. I had their investor presentation and two years of 10-K filings. Built the revenue model, the expense assumptions, the working capital schedule, the WACC calculation. The DCF alone took about four hours. By then it was 9 PM and I was on my second Seamless order. Pad thai, which I ate while formatting. The formatting is the part people don't understand. Every chart has to be in the firm's template. Every font is Calibri 10. Every axis label has to match. If the blue in your chart is #4472C4 instead of #4472C8, Vikram will catch it. I'm not exaggerating. He keeps a color swatch reference.

What happened the next morning?

I finished at 5:40 AM. Went home, which is a 12-minute walk to Murray Hill. Showered. Came back at 6:50. The pages were in the MD's inbox by 6:55. He reviewed them on his phone in the elevator, apparently, because by 7:10 he sent one comment: "Page 23, footnote 3, update the date." I updated the date. He said "good, send it." Charlotte packaged it into a PDF and sent it to the client's team. The client meeting happened at 2 PM. I was not in the meeting. Analysts are rarely in client meetings at this stage. Vikram told me afterward that the client liked the new target and wanted to move forward with preliminary due diligence.

Eighteen hours of my life became a PDF attachment in an email. Nobody in the meeting knew I'd built it overnight. The MD said "good, send it." That was the entire acknowledgment. And honestly, that's fine. I don't need applause. But when people ask me what investment banking is like, that's the most honest answer I have. It's building something under extreme time pressure, making it perfect, and then watching it disappear into a meeting you weren't invited to.

You stayed a third year. Most people leave after two. Why?

Money and indecision, in that order. My base salary is $120,000 and my bonus last year was $85,000, so total comp was $205,000. At 27. In New York that doesn't go as far as it sounds, but it goes far enough. I have a studio in Murray Hill that costs $3,200 a month. I save about $4,000 a month after taxes and rent and the Seamless habit I refuse to audit.

The indecision part is about private equity. My co-analyst Trevor left after year two for a PE megafund. He's making more money and working slightly fewer hours, but the work is basically the same, just on the buy side instead of the sell side. He's still in Excel. He's still formatting. He still doesn't sleep. I looked at PE recruiting and I just, I don't know. The whole recruiting cycle is insane. You start interviewing for PE jobs six months into your first year as an analyst. You've been on the job for six months and people are asking you to model LBOs in a case interview for a job you'd start eighteen months later. I wasn't ready to make that decision that fast, so I stayed a third year, and now I'm promoting to associate, which is the first time I'll be in client meetings, managing analysts, and theoretically thinking about things instead of just building things. We'll see.

You said Trevor is "still in Excel." Is that the job at every level?

Not exactly. The job completely changes as you move up, which is one of the weird things about banking. As an analyst, I'm a production machine. I build the product. As an associate, I'll be reviewing the product and managing the analysts building it. As a VP, you're running the deal process, managing the timeline, quarterbacking the client relationship day to day. As an MD, you're selling. You're a salesperson. You're taking CEOs to dinner at that omakase place and convincing them to hire your bank instead of the other five banks pitching them. The skills that make you a good analyst, speed, accuracy, endurance, have almost nothing to do with the skills that make you a good MD. It's like being promoted from carpenter to architect to real estate developer. Same industry. Completely different job.

What's yours?

The optionality trap. Everyone tells you that banking gives you "optionality." You can go to PE, you can go to hedge funds, you can go corporate, you can do an MBA. And that's true. The doors are wide open. But what nobody tells you is that having all those doors open becomes its own kind of paralysis. Trevor went to PE and now he's locked in for another two years. My friend Meredith went to business school and she's $180,000 in debt. Another guy from my class, Daniel, went to a corporate development role at a tech company and he's happier but he makes $140,000 and he wonders if he left too early.

The optionality is real but it also means you spend your entire analyst experience not living in it. You're always looking ahead. Always positioning. Always thinking about the next move instead of whether you actually like the current one. I've been here three years and I still don't know if I like this job or if I just like what this job can get me next. That's a strange way to spend your twenties. My mom calls every Sunday and asks if I'm eating enough. She doesn't ask if I'm happy. I think she stopped asking because my answers were too complicated and they worried her. The honest answer is: I don't know yet. Ask me when I figure out whether the doors were worth the hallway.

You did the bulge bracket thing and then left. What was wrong with it?

Nothing was wrong with it exactly. It was, structurally, very good for my career. Two years at a top firm, strong deal experience, the name on the resume. What was wrong was that I was a cog. A very well-compensated cog, but a cog. At the bulge bracket, as an analyst, you're on a team of maybe 15 people working a deal. You build one section of the model. You do one part of the pitch book. You never see the client. You never hear the negotiation. The deal closes and you get an email from the group head that says "great work, team" and that's it. I wanted to be in the room.

After two years I went to Kellogg for my MBA. During recruiting, I specifically targeted middle-market banks because the deal teams are smaller. Here, a deal team is me, my MD Danny, an analyst, and maybe a second analyst if the deal is big. Four people. I'm in every client meeting. I hear the negotiation. When Fatima, the CFO of a portfolio company we sold last quarter, wanted to understand why the buyer was discounting her EBITDA for customer concentration, I was the one who walked her through it on a Tuesday at 8 PM on the phone. At the bulge bracket, someone three levels above me would have had that conversation and I would have been in a conference room updating a data room index.

Tell me about a specific deal moment that stuck with you.

Last month. Thursday. We were presenting a management presentation to the board of a company we were advising on a sale. Industrial distributor in the Midwest, about $180 million in revenue. Danny and I had been working this deal for four months. The management presentation is the document the company uses to market itself to buyers. It's maybe 60 pages. Financial overview, growth story, market positioning, all of it.

Kwame is my analyst. He's sharp, first-year, came from a good school, works hard. He'd built the financial model for this deal. The model takes the company's historical financials and projects them forward five years. Revenue growth, margin expansion, capex, working capital, all flowing into an enterprise value range that we'd present to buyers. The model is the backbone of the entire deal.

I was reviewing the sensitivity table the morning of the board meeting. Sensitivity tables show how the company's value changes under different assumptions. And something felt off. The base case EBITDA for year three was $42 million, but when I changed the revenue growth assumption from 8% to 6%, the EBITDA only dropped to $41.2 million. That's not enough. A two-percentage-point change in revenue growth on a $180 million revenue base should move EBITDA by more than $800,000. The math didn't feel right.

I opened the model. Went into the revenue build tab. Followed the formulas through to the EBITDA bridge. Found it on the third tab: a circular reference in the debt schedule. The interest expense was referencing a cell that referenced the debt balance, which referenced a cell that pulled from the cash flow statement, which pulled from the income statement, which included the interest expense. A loop. Excel was iterating through it and landing on a number that was close to right but overstated EBITDA by $4.2 million in the projection years.

$4.2 million. And the board meeting was that day?

The board meeting was in 40 minutes. Danny was already in the lobby of the client's office making small talk with the general counsel. I was sitting in a conference room with my laptop open and Kwame across from me. I didn't yell. I didn't panic. I said, "Kwame, there's a circular reference in the debt schedule. The EBITDA is overstated by about four million in the out years. I need you to fix it. I need the updated sensitivity table and the summary output page in my inbox in 20 minutes." And I got up and went to find Danny.

Danny was mid-sentence with the GC about the Bears. I pulled him aside and said the meeting would start a few minutes late because I wanted to double-check one of the financial exhibits. Danny looked at me for about two seconds. He trusts me. He said OK. Then I went back to Fatima, who was the CFO and was also in the lobby waiting, and I asked her a question about her Q3 inventory levels. Not because I needed the answer. Because I needed 15 more minutes.

Kwame had the fix to me in 22 minutes. The updated EBITDA was $37.8 million in year three instead of $42 million. The enterprise value range shifted down by about $15 million on the low end. I swapped the pages in the presentation file, printed fresh copies, and walked into the board room 18 minutes late. Nobody knew why.

What happened to Kwame?

Nothing happened to Kwame. He made a mistake. Analysts make mistakes. The job exists, in part, so that mistakes get caught before they reach the client. That's what I'm here for. I sat down with him the next morning and walked through the model structure. Showed him how to check for circular references. Showed him the quick audit formulas I use. He took notes on a legal pad and filled three pages. He'll probably never make that specific mistake again.

What I didn't tell Kwame is that I made the exact same mistake in my second year as an analyst. My VP caught it. Her name was Tara and she handled it the same way I did, quietly, without drama, and she fixed it before anyone above her knew. I only found out years later when she told me the story over drinks after she'd left the industry. The chain continues. Someday Kwame will catch an analyst's circular reference and he'll handle it the same way, and the client will never know, and the deal will close, and nobody will write a case study about the 22 minutes that saved a $15 million valuation error from reaching a boardroom.

You mentioned 11 pens. Which one matters most?

The fourth one. That was the first deal where I was the lead VP, not the supporting one. A $95 million sale of a specialty chemicals manufacturer in Indiana. Danny let me run the process. I sourced the buyers, managed the data room, negotiated the purchase agreement markups with the buyer's counsel. When we signed, I was sitting in a law firm conference room in downtown Chicago at 11:40 PM and the CEO of the company, a man named Gerald who'd built the business over 30 years, looked at me and said "thank you for taking care of this." He meant the deal. But what he was really saying was: I trusted you with the thing I spent my life building, and you didn't mess it up.

That pen is the one I'd grab if the drawer caught fire. JP has asked me about the pens a few times. He doesn't fully get it. He designs buildings. When a building is done, people live in it. You can walk past it. My deals close and they become a paragraph in a press release and then they're over. The pen is my building. It's the only physical proof that the thing happened.

What's yours?

How much of the job is performing confidence you don't always feel. I'm a Black woman in a room that is, most of the time, older white men. The CEO, the GC, the board members, the buyer's team, the lawyers. Danny is great. He advocates for me. But Danny isn't in every room. And in the rooms where he isn't, I notice things. The CFO who directs his question to the junior associate sitting next to me instead of to me. The buyer's counsel who calls me "miss" while calling my male counterpart by his first name. The board member who asks Danny to "confirm" something I just said, as if hearing it from me wasn't enough.

I don't let it slow me down. I just work. I let the pens speak. Eleven deals closed. But the energy it takes to perform confidence in those moments, to answer the question calmly and completely while also processing the microaggression and also keeping the meeting on track, that's a tax. Not a financial one. A metabolic one. By Friday I'm tired in a way that has nothing to do with the hours. JP notices. He'll say "you seem heavy tonight" and I'll say "long week" because explaining the specific weight of it would take more energy than I have left.

You left a bulge bracket as a director. That's one step below MD. Why?

Because I could see what the next 15 years looked like, and I didn't want them. At a bulge bracket, the promotion from director to managing director takes two to three years if things go well. You're running deal teams, managing client relationships, and doing a tremendous amount of internal politics. There are committees. Revenue attribution arguments. The question of who "owns" a client relationship, which at a bulge bracket is genuinely one of the most political things you'll encounter in any industry. I spent about 30% of my time as a director on internal nonsense that had nothing to do with clients or deals.

Stefan and I had been talking about starting something for years. Stefan was in healthcare banking at the same firm. We covered different sectors but we had the same frustration: we were spending more time selling internally than selling to clients. The boutique lets us choose our clients, choose our deals, and spend zero time on internal politics because there's nobody to politic with. It's 22 people. Everyone knows what everyone is working on. We had breakfast together this morning and in 20 minutes we covered everything that would have taken three committee meetings and a PowerPoint deck at the old firm.

What kind of deals does a 22-person firm do?

Tech M&A. Sell-side and buy-side advisory between $50 million and $500 million. We don't do IPOs, we don't do debt capital markets, we don't do anything except advising companies on buying or selling businesses. That focus is the point. When a SaaS founder with $40 million in ARR is thinking about selling, he doesn't want a bulge bracket that will assign a 26-year-old analyst and a rotating cast of VPs. He wants someone who has done this exact deal fifteen times and will pick up the phone on a Saturday. That's what we sell. Specialization and availability.

Audrey is our most senior associate. She came from Credit Suisse, which doesn't exist anymore, and she's very good at the modeling and diligence work. But the thing that makes our firm different is that I'm in every meeting. Not an associate. Not a VP who reports to me. Me. The founder talks to the same person from the first pitch to the closing dinner. At a bulge bracket, the MD shows up for the pitch, disappears for four months, and reappears at the signing. The mid-level people do the actual work. Here, I do the pitch and the work. It's more hours for me but better outcomes for the client, and better outcomes mean repeat business and referrals, which is how a 22-person firm stays alive.

Tell me about the $280 million conversation.

Peter built a SaaS company over nine years. Infrastructure monitoring for mid-market enterprises. Good product, loyal customers, about $38 million in ARR growing at 22% year over year. Reasonable capital efficiency. We ran a sale process, solicited interest from strategic and financial buyers, and the best offer came in at $280 million.

Our model had projected a range of $310 to $340 million. The $280 million offer was below the low end. Three things drove the gap. First, the buyer used a 7x ARR multiple instead of the 8.4x we'd modeled, because the comparable transactions from Q3 and Q4 of the prior year had compressed due to rising interest rates. Second, they excluded about $3 million of revenue that Peter considered recurring but the buyer classified as professional services, which trades at a lower multiple. Third, they applied a customer concentration discount because Peter's top three customers represented 31% of total ARR. Any one of those adjustments is defensible on its own. Together they explain the full gap.

I had to call Peter and walk him through this. Peter had spent nine years building this company. He'd missed vacations, taken below-market salary, burned through his savings in the early years. The $40 million gap between what he expected and what he got was not a financial abstraction to him. It was, in his words, "three years of my life."

How did that conversation go?

I told him the truth. I said: "The buyer's analysis is methodologically sound. I don't agree with all of it, and there are two areas where I think we can push back. The revenue classification is arguable, and I think we can recover $8 to $12 million by reclassifying that $3 million as recurring with supporting data. The concentration discount, I can argue against it but I don't think the buyer will move much because it's objectively true that losing any one of those three customers would materially impact the business. The multiple compression is market-driven. We can't argue with the market."

Peter was quiet for a long time. Maybe 30 seconds, which on a phone call feels like five minutes. Then he said, "I thought we'd get more." And I said, "I know. Let me work on it." I spent the next two weeks negotiating. We got the purchase price to $296 million. The revenue reclassification worked. The concentration discount stayed. Peter signed. At the closing dinner, he shook my hand and said, "You were honest with me." He didn't say thank you. He didn't say he was happy. He said I was honest. That's the job.

Liz was a banker too. What's that like at home?

It's useful because she doesn't need me to explain why I'm checking my phone during dinner. She knows. She did the same thing for six years before she went in-house at a tech company. She's the head of corporate development there now, which means she's on the buy side. Sometimes we're on opposite sides of the same kind of deal, theoretically. We don't discuss specifics, obviously. But she'll say something like "we passed on a deal today because the seller's banker overpriced it by 20%" and I'll think, I wonder if that was one of mine. We've never confirmed it. It's become a running joke. "Was that yours?" "I can neither confirm nor deny."

Liz told me once, maybe four years into the bulge bracket, that the person I was becoming at that job wasn't someone she liked very much. That stayed with me. She didn't mean I was a bad person. She meant I was becoming someone who couldn't turn it off. The phone checking. The emails at dinner. The way I'd come home and still be running deal scenarios in my head while she was telling me about her day. The boutique is better. Not because the work is less intense but because the work feels like mine. When I'm stressed about Peter's deal, I'm stressed about something I chose. At the bulge bracket, I was stressed about something that was assigned to me by a committee I'd never met.

What's yours?

The loneliness of the advisory role. You're the person who knows everything about both sides of a deal. You know the seller's real floor price. You know which buyer is most motivated. You know the terms that will kill the deal and the terms that are just posturing. You carry all of this information and you can't share it with anyone. Not Liz. Not Stefan, unless he's on the same deal. You are a vault.

Peter trusted me with the most important financial event of his life. The buyer's CFO trusted me to present fair analysis. My job is to serve Peter, my client, but if I lie to the buyer, I lose my reputation, which is the only thing a boutique has. I live in a permanent state of informed isolation. I know more about my clients' businesses than their own families do. The founder who's selling because his marriage is falling apart and he needs liquidity. The buyer who's acquiring because their organic growth has stalled and their board is restless. I know all of it. I go home and Liz asks how my day was and I say "productive." The actual answer would take an hour and include information I'm ethically obligated not to share. The run at 5:30 AM is the only time I process any of it. The Embarcadero doesn't have an NDA.