The number where the paths stop separating
Ask how much money you actually need to more or less do what you want. In New York it is high, somewhere around $200,000 to $300,000. It is also lower than most students assume, and once you are past it the lifestyle gap between these jobs goes marginal.
Think about what the extra dollars buy at that level. Four-star hotel or five-star hotel. First class, which most of these paths already cover if you want it. You can eat at any restaurant you want, take a vacation anywhere you want, and stop checking prices. That is the ceiling of the experience.
A first-year analyst at a large bank starts under that ceiling and does not stay under it long. Out of undergrad at a bulge bracket you are at $160,000 or better all in, with a $40,000 to $50,000 bonus and a sign-on inside that number, and it grows quickly from there. Which is the whole point. You can always make more, and the lifestyle ceiling arrives well before the pay stops rising.
Your savings are not marginal. Compounding is real and the paths diverge sharply there. But if you are choosing based on how your daily life will feel, your life will probably be very good on any of these paths. The ranking you built is measuring a variable that has already maxed out.
What the money is actually buying
Here is the part people skip. Banking buys your time in bulk. It is a wholesale purchase, and the discount it gets is enormous.
If you sign up, you are on board for two years and there is not much getting out of it. You are never far from your phone or your laptop. You are on call in the way you are on call during finals week, except it runs for two years. You have dinner plans, the VP calls, the deal went live, and the dinner is gone. That happens often enough that it stops being a story.
So the dollar power is weaker than the number on the offer letter suggests. Money you cannot spend on a Tuesday night is money at a discount. Once you accept that, the variable that still moves is the one worth optimizing: the ratio of dollar to life value. How much life does each dollar cost you here.
Where the ratio actually differs
Run that ratio across the paths and they reorder in ways the prestige list never shows.
The quiet seat. Somewhere in the syndicate structure of a large debt deal, there is a small desk at a smaller bank. The lead banks with the huge balance sheets have to bring in other banks as selling group members, partly to show the client that everyone was included. That desk gets business by structure. It might be two or three people. Much of the job is joining the working group call and supporting the lead bank's position. Rise through it and you are still making a few hundred thousand a year. On the dollar-to-life ratio, that beats almost everything on the prestige list, and the economy has a lot more of those seats than most students realize. They are attached to interesting work and good lives.
Private equity. The power dynamic flips. You become the client, which means you stop being in the service industry and start being the one who asks. Hours are still long, seventy a week is unremarkable, but they are predictable, because a principal is unlikely to appear at 5pm on a Friday needing a sixty-page deck by Monday. The stress moves rather than disappears. Teams are leaner, fewer people check your numbers before a partner sees them, and if the investment goes badly it is your investment.
Corporate finance. The median is below banking. The tail is above it. Go to a company that might actually go public, say on day one that corporate finance is the career you want, and you are competing against far fewer people than you would be for the banking seat. I have had conversations with the corporate finance team at SpaceX. If a company like that lists, the people inside it get an outcome no Goldman analyst gets, because analysts do not receive options.
Big Four valuation and audit. A lower beta asset. It is a different kind of purchase from banking. Someone has to sign the audit and someone has to look at the taxes, and legally a human has to do both, so the work is durable. The door to deals is not shut either. Inside a firm you can transfer onto a due diligence or a valuation team and inch closer to deals from there. I coached a candidate who was doing deal diligence at a consulting firm rather than a Big Four, and he lateraled to a boutique. He knew the technical material cold, and it showed the moment he talked to bankers.
Startups. The widest range of anything here, with no floor at all, since anything can call itself a startup. A classmate of mine left the MBA program after the first year for an AI startup. It raised, and a few of us sat in a pub running the math on his equity against the new valuation. On paper he had made a couple of million dollars that year. Venture capital is the adjacent bet and a worse one at entry level: pay is low, returns are concentrated in a handful of winners, and most of the daily work is sifting through spam from people with bad ideas.
Two things that break the math
Your starting point sets the return. Think of a career move the way you would think about an LBO. The student debt is the acquisition financing and the career is the asset you are buying. The IRR comes out strong, but it depends entirely on where you started. If you are already making $200,000 with good exits ahead of you, an MBA into banking probably does not produce an outsized return. My own case ran the other way. I was at Deloitte making about $125,000, living well in Chicago and saving. I got a partial scholarship for the MBA, then an offer at Centerview: a sign-on somewhere in the mid-teens to mid-twenties, another bonus when I started, and $80,000 over the ten weeks I worked that summer. That is the shape of the upside. Then it plateaus. Associate pay grows roughly linearly, VP is a step up, and the enormous outcome is not in the path itself without an exit.
The heuristic goes stale. Most students choose by looking at what the class ahead of them did and assuming there was a reason. Usually there was. Sometimes it expired. An alum came to Booth and told us that when she graduated ten years earlier, nobody wanted private equity, everyone wanted hedge funds. I have also heard that private equity take-home has come down as carry rules changed, and I want to be precise about that one: I have no data on it and it is a thing I heard, not a thing I can document. Treat it as a rumor. The lesson is the shape of the problem rather than the specific claim. Before you chase the outcome the previous class chased, check that the party is still going.
The move that raises the ratio on any path
Specialize earlier than feels comfortable.
At Deloitte I was more of a specialist than a generalist. I helped CTOs make investment decisions about technology infrastructure, which meant learning data centers, networks, server racks, fiber, software, and the operational side of all of it. At the time none of that was cool. Now people ask me how I know about data centers. It made recruiting into tech banking easy, because I already understood how technology companies work, which is not something you pick up in a three-month prep sprint.
The banker I work with puts the same point more bluntly: the money is made when you become an expert. You become the go-to tech banker, you know your clients inside out, and that specific knowledge is what makes you hard to replace. He also concedes the obvious, that it is very hard to know what to specialize in at 22, which is exactly why most people default to generalist and stay there.
The trade is honest in both directions. Making a distinct choice costs you optionality and buys you a shot at the upside. Keeping your options open costs you the upside and buys you time. Just make the trade on purpose, and notice that it applies inside banking too. Pick a sector, learn how those companies actually make money, and become the person in the group who knows it.
If the ratio still points at banking, the part you control is arriving with the technicals already handled. MBA recruiting moves almost immediately once you set foot on campus, and nobody learns three-statement mechanics while running a networking campaign. The learning pathway is the order I would take it in.
