What To Know For An Investment Banking Interview

The Actual Breakdown Of What Goes Into These Interviews

Most people walk into an investment banking interview thinking they need to prove they're smart. They're not wrong, but that's not how it actually works. They need to prove you won't implode under a 18-hour day and that you can build a DCF without needing a hint every five minutes. The gap between what candidates study and what interviewers actually care about is huge.

I remember sitting across from a vice president once who spent the first ten minutes of a 45-minute interview just watching me struggle with a simple valuation premise. He didn't correct me. He just waited. I'd built a LBO model in my head backwards, starting with the exit multiple before figuring out the purchase price. His exact words when I finally caught myself: "So which way around does it go?" That was the whole interview right there. It wasn't about the math. It was about whether I'd catch the structural error before presenting garbage as analysis. Let's get through the technicals first because that's the gate you have to pass through. You need to be genuinely comfortable with the core skills: accounting, valuation, and financial modeling. Not "I can recite the formulas" comfortable. I mean "someone could tear my model apart and I'd rebuild it on a whiteboard in 20 minutes" comfortable. Here's the thing nobody tells you about the accounting section. Most candidates know the three statements link up. What they don't know is that the interviewer is specifically testing whether you understand the causal chain, not just the mechanical link. Ask them to trace a $10 depreciation increase through all three statements and half of them will say net income drops by $10. They forgot the tax shield. Cash flow from operations goes down by $6, not $10, because taxes decrease by $4. The balance sheet doesn't balance at that point. You've now got to explain where the other $4 goes and why.

I had a candidate once who got this wrong three separate ways in one interview. Each time I corrected her, she'd pivot to a new wrong answer instead of re-examining the original assumption. That's the red flag. The technical answer matters less than whether you can think through a correction gracefully. In the actual job, you will make mistakes. The question is whether you panic or you recalibrate.

Valuation Is Where People Actually Get Tested

DCF, comparables, precedent transactions. You need to know all three, know when to use each, and know the limitations of each. The limitation part is what separates people who read a blog post from people who've actually done this work. For DCFs, the single most important thing to understand intuitively is that terminal value usually represents 60 to 80 percent of total enterprise value in a standard model. If your terminal value is 50 percent or less, you're probably making an error somewhere. If it's over 90 percent, you picked a terrible growth rate or WACC assumption. I once reviewed a model from a candidate who had a terminal value of 94 percent because he'd used a 4 percent terminal growth rate against a 7 percent WACC. The spread was too thin. The output looked impressive but it was built on fantasy assumptions. Pointing that out during an interview shows you've actually built these things, not just studied them. With comparables, the pitfall is selection bias. Picking comps isn't just about finding companies in the same industry. It's about finding companies with similar growth profiles, margin structures, and capital intensity. Two companies in the same sector can trade at completely different multiples if one is growing at 30 percent and the other is flat. I've seen candidates use a median EV/EBITDA of 12x for a high-growth tech company and then wonder why their implied valuation was way off. The answer is that the median was dragged down by mature peers. You need to segment your comps, not just cluster them.

Precedent transactions add a control premium on top of everything. That's usually 20 to 40 percent above trading multiples. The practical issue is that precedent transaction multiples are backward-looking and every deal is structurally different. You'll never find a truly comparable acquisition. The skill is in adjusting for differences, not just copying a multiple.

Modeling Tests Your Patience More Than Your Intelligence

The break-even case interview is brutal because it sounds simple and is almost impossible to nail cleanly under pressure. You'll be asked to adjust a single assumption and recompute everything. If you're working from memory instead of having a model in front of you, you need to understand the full sensitivities. A 50 basis point change in WACC can shift enterprise value by 3 to 5 percent depending on the profile. A 1 percent change in terminal growth rate can move it by 8 to 12 percent. These aren't trivial. They're the difference between a buy and a pass. Here's a practical edge case I ran into repeatedly. Candidates often forget that working capital changes affect free cash flow but not EBITDA. If someone asks you how an increase in accounts receivable impacts your DCF, the answer is it reduces free cash flow in that period but has no effect on EBITDA. Getting this wrong makes you look like you've never built a model from scratch. I learned this the hard way during my first summer internship when I accidentally double-counted a working capital adjustment in a pitchbook and the MDCT caught it mid-meeting. Never forget it again.

Behavioral Questions Are Not Afterthoughts

This is where most firms actually filter people out. Technical competency is table stakes. Everyone in the final round can do the math. What they're evaluating is whether you're someone they'd want trapped in an elevator for three hours on a Sunday night, whether you'll represent the firm decently to clients, and whether you'll stay longer than six months. Your story needs to be coherent. I've seen candidates talk about wanting to go into investment banking because they love markets, then pivot to saying they want to advise CEOs on strategic decisions, then finish with "and I just really like solving complex problems." Three different motivations that don't connect. Pick one thread and follow it. "I started in equity research, got fascinated by how valuations drive deal decisions, and realized I wanted to be on the deal side where those valuations actually get tested" is a complete narrative. It's specific, it's credible, and it shows you understand what the job is. For the "why our firm" question, don't say "you're a top-tier firm with great reputation." They get that answer 200 times a day. Mention a specific deal they did in the last two years and what about it attracted you. Mention a sector they're strong in that aligns with your background. It takes maybe 30 minutes of research and it separates you from everyone who didn't bother.

What Most Candidates Miss Completely

The current macro environment matters more than it used to. Interest rates, deal flow volume, sector rotation. If you're interviewing in 2026 and you can't discuss how higher rates have compressed LBO returns or changed M&A timing, you're behind. I had a candidate who gave a textbook DCF answer without acknowledging that in a 5 percent rate environment, the discount rate assumption changes the story entirely. He wasn't wrong, he was just incomplete. In practice, the right answer includes both the theoretical framework and the practical adjustment for current conditions. Another thing: know your resume cold. Every line on it will be fair game. If you listed "analyzed comparable company valuations" on your resume, expect a grilling on how you selected comps, what multiples you used, and what conclusions you drew. I once asked a candidate what the most interesting finding was from a project he described in two bullet points. He couldn't remember. He'd written the resume line but never revisited the actual work. That's a red flag for a reason. There's also the unspoken test of whether you're going to be difficult to manage. Senior bankers can spot a candidate who's argumentative disguised as confident within the first five minutes. There's a difference between pushing back thoughtfully and being combative. "I see it differently because..." followed by a structured reason is fine. "Actually, that's wrong" without any framing is not. I've passed candidates with weaker technical answers over candidates who couldn't accept a gentle correction without getting defensive. The job is client-facing. How you handle pushback in an interview is a preview of how you'll handle pushback from a managing director at 2 AM.

Practical Preparation Steps That Actually Work

Build a DCF from scratch without looking at a template. Time yourself. Then build an LBO model. Then do a comps analysis. Do these on paper first, not in Excel, because in the interview you'll likely be working from a whiteboard or a napkin. If you can only do them cleanly in Excel, you'll freeze under pressure. Practice explaining your answers out loud, not just in your head. There's a gap between knowing something and articulating it clearly when you're nervous. Record yourself answering "walk me through a DCF" and listen to it. You'll catch filler words, rambling, and logical gaps you didn't notice while speaking. Read recent deal announcements in your target sector. Not the press releases, the actual transaction details. What was the purchase price? What multiple did they pay? How was it financed? This comes up naturally in conversation and shows genuine interest rather than rehearsed enthusiasm.

The Honest Limitations

No amount of preparation guarantees an offer. Fit matters. Random chance matters. The hiring team's mood that day matters. I've seen genuinely strong candidates get rejected because the team already had someone with a similar profile and was looking for someone with a different skill set. I've also seen average candidates get offers because they fit a specific gap the team had. Don't internalize rejection as a reflection of your capability. Treat it as data points and keep moving. Over-preparing on technicals at the expense of coherence is a real trap. I'd rather interview someone who gives a slightly imperfect but clearly explained answer than someone who recites a perfect textbook answer that sounds like they memorized it. Understanding beats memorization every time. If you genuinely understand why a DCF works the way it does, you can reconstruct it even if you forget the exact formula. If you memorized the formula, you're stuck the moment anything changes. The interview process itself is a simulation of the job. They're testing whether you can think clearly under pressure, communicate complex ideas simply, handle correction without crumbling, and maintain composure when you don't know the answer. Master those four things and the technical details become easier to manage. Not the other way around.