Rolling Layoffs and Talent Upgrades: Are RIFs a Thing of the Past?

The short answer

• Banks are still shrinking headcount, but mostly through performance-based cuts and other quiet tools, not classic reductions in force (RIFs).
• The two are different decisions. A RIF eliminates positions. A performance-based cut removes a person, and the seat can be refilled.
• A record M&A market makes the shift possible. Cuts are easy to fund, and a thin deal sheet has no macro excuse.
• Banks favor performance-based tools partly because a RIF carries a stigma. It reads as distress for the firm, and the stigma shifts to the individual.
• Pressure is also landing at promotion time, below the MD level. MD classes at Goldman and Morgan Stanley grew, but the filter appears to be tightening at Associate 3 and VP 3/4.
• The new toolkit is cheaper and far less public than a RIF, and its purpose is a talent upgrade more than a smaller bank.
• True RIFs survive as large, announced structural programs. Whether the new toolkit holds in a downturn is untested.

For the last two decades, a Wall Street layoff followed a script. Markets turned, deal flow dried up, and a bank announced a RIF with a number attached, a memo, a severance package and a headline. The downturn was the explanation, and even the timing was predictable.

The 2026 cycle has broken that script. Goldman Sachs just posted the best quarter in its 157-year history. Bloomberg’s tally shows headcount at Bank of America, Wells Fargo, Citigroup, Goldman and Morgan Stanley fell by more than 10,000 in that same quarter. In the first quarter, the biggest banks trimmed more than 5,000 jobs while the six largest earned $47.3 billion. That followed 2025, when the six largest banks cut about 10,600 jobs, the most since 2016.

So are RIFs a thing of the past? The better question is whether banks still need them. The answer starts with what a RIF actually is.

1. The Old System: A Calendar and a Headline

As we wrote in 2024, large-scale reductions at major banks clustered in two windows, and bankers planned their careers around them.

Outside those windows, terminations were generally isolated cases tied to misconduct, negligence or client conflicts. Boutiques could run on different cycles, but the rhythm was well enough understood that people used it to time job searches, focus their efforts and plan their finances.

Goldman’s annual Strategic Resource Assessment (SRA) was the purest example. It was a once-a-year cull that typically removed 1% to 3% of the workforce. As recently as March 2025, the bank planned to cut 3% to 5%. On Goldman’s roughly 46,500 employees, that works out to about 1,395 to 2,325 jobs. Even the ritual we remember as a layoff targeted underperformers, so it was performance-based in substance. What made it feel like a RIF was its calendar, its scale and its visibility.

2. RIFs vs. Performance-Based Cuts

The two terms get used interchangeably in headlines, but they describe different decisions.

A reduction in force eliminates positions. The firm decides it needs fewer seats because of a downturn, restructuring, divestiture or cost program, and then works out which ones go. The question is “which roles do we no longer need?” The seat generally isn’t refilled.

A performance-based cut removes a person, not a position. The firm decides an individual isn’t meeting the standard for the role. The question is “who isn’t performing at the level we need?” The business can be growing, and the seat is often backfilled.

Three points matter for everything that follows.

The label changes the story. A RIF says the business needs less. A performance-based cut says the business is fine and this person fell short. In a record quarter, only the second story is comfortable. Goldman’s spokesperson called its approach “regular, consistent headcount management,” a phrase built to sound routine.

The line is not always clean. If a bank cuts people for performance and doesn’t refill the seats, the effect on headcount resembles a RIF even though the label differs. That is our analysis, not a bank statement. It also matters legally, because notice rules and documentation requirements can differ between the two. Anyone relying on the performance label should be sure the record supports it, and that is a question for counsel.

Both are running at once. Citi’s multi-year program is a classic structural reduction. Goldman’s rolling cuts are performance-based. The industry isn’t choosing between them. It is using each where it fits.

The RIF stigma

A RIF carries a stigma, and part of the shift toward performance-based tools is an effort to avoid it. The stigma doesn’t disappear. It moves.

For the firm, a RIF reads as distress. In a record quarter, an announced reduction invites the question of why. “AI lets us run the same business with fewer people” isn’t a message every audience wants to hear. Goldman’s own analysts have noted that clear evidence of AI-motivated layoffs remains limited, so a bank that leans on that explanation invites scrutiny. Recruiting is a second cost, since banks run campus programs and forensic lateral searches at the same time and want to look like places where careers are built. Citi shows the exception. A RIF is tolerated when it reads as strategy and the audience is investors who want to hear the savings.

For the individual, the stigma flips. After a RIF, “my group was cut” is a story a banker can tell a recruiter. After a performance exit or a “repeat the year” outcome, the message is about the person, and in a record market there is no macro cover. The firm avoids the stigma of a RIF by moving it onto the employee.

If you’re on the receiving end, ask how the exit will be characterized. What will be said internally, and what will the firm say if a recruiter calls? Does any severance agreement state a reason or include non-disparagement terms? How will a termination be reported if you hold registrations? Those answers depend on your agreement, so this is a question for an employment lawyer.

If you’re a recruiter or hiring manager, a 2026 departure can’t be read from the headline. Someone who left in a rolling cut, a promotion hold or a structural program can look identical on a résumé, so ask before assuming.

The stigma analysis is our inference, not something the banks have discussed.

3. Why Now: A Record M&A Market

A great deal market is the enabling condition for the new model, not a contradiction of it.

The numbers are hard to overstate:

• Goldman’s investment banking fees rose 55%, with an advisory backlog at its highest in five years and large-cap M&A volume up 90% through the first half.
• JPMorgan’s IB fees rose 30%, Morgan Stanley’s investment banking revenue jumped 58%, Bank of America’s IB fees rose 50% and Citi’s rose 44%.
• Across the five big U.S. banks, investment banking fees rose 46% to $12.9 billion in the second quarter, according to S&P Global Market Intelligence.
• Equities revenue was up 69% to 86% at the three banks that broke it out.
• Johnson Associates raised its 2026 bonus forecast by about 3 points on average from its first-quarter estimate and calls it the “Year of the Bank.”

That environment matters for cuts in four ways:

1. Banks can afford it. Severance and technology spend are easiest to absorb at record profits. Wells Fargo’s $612 million in severance pushed up its Q4 2025 expenses. JPMorgan is spending $19.8 billion on technology in 2026, up 10%.
2. There is no macro excuse. A banker with a thin deal sheet has nowhere to point, and a firm can see cleanly who is converting deal flow into revenue. That also makes a performance rationale credible.
3. Operating leverage is visible. Goldman’s CFO described “material operating leverage” in the second quarter, with revenue up roughly 40% and compensation expense up roughly 30%. When comp grows more slowly than revenue, the firm is rewarding top performers, not lifting every seat equally.
4. The front office is protected on purpose. Banks are cutting the back office in the best revenue environment in years so they don’t have to cut the front office when the cycle turns. Johnson Associates projects M&A advisory bonuses up 15% to 20% and equity sales and trading up 20% to 30% or more.

The boutiques show the market sorting platforms, though for different reasons:

• Evercore posted record second-quarter revenue of about $1.0 billion, up 19%, with advisory fees of $776 million and underwriting up 201%. It added 19 senior managing directors year-to-date. PJT also reported records.
• Houlihan Lokey reported $511 million, down 15.5% from $605 million. The cause was delayed large-fee Corporate Finance deals, Middle East instability and software-sector disruption. Management calls it temporary. This is a deal-timing story, not a talent story.
• Lazard is the case to read carefully. Total revenue was $808 million, up 1%, and Financial Advisory revenue fell 9% to $450 million while Evercore’s rose. GAAP net income fell to $5 million from $55 million, a 91% drop. The company attributes that mainly to an elevated tax rate of 63.5%, and adjusted net income was $13 million. The CFO said the firm is exiting the most substantial period of repositioning its advisory talent.

A strong tape doesn’t lift every platform or banker equally. Lazard’s advisory revenue fell in the same quarter that peers set records, and that is where “no macro excuse” applies.

4. The New Toolkit: Five Ways Banks Reshape Headcount

Rolling performance cuts. Instead of the annual SRA, Goldman moved to smaller cuts beginning in April and running through the summer. The April start sits inside the old spring window, but the summer continuation falls outside it, and that is the break with the calendar. Divisional leaders gained control over when to move on poor performers, without waiting months for a centralized review, and the cuts are expected to reach every corner of the bank. A Goldman spokesperson said the firm is “constantly assessing” performance and talent across divisions. Reuters reported that the bank did not disclose how many roles would be affected. Morgan Stanley had recently taken a more traditional route, letting go about 2,500 people, around 3% of its workforce, across all divisions.

Promotion-year holds and exits. This is the quietest performance cut. Based on what we are seeing in the market, banks are increasingly telling Associate 3s to repeat the year or leave, and applying similar discipline to VP 3s and 4s facing the step up. Titles and timelines vary by bank. None of it is characterized as a RIF. No position is eliminated and no headcount target is announced. Someone is told they aren’t ready for the next title. Promotion years are the ideal place for this because:

• The decision point already exists. Promotion decisions were part of why the fall sweep happened when it did, so performance and promotion conversations have always been linked.
• The record market removes the excuse. A banker who isn’t promoted in the strongest market in years can’t blame a soft tape.
• It controls the pyramid. Headcount at the six largest banks has barely moved in five years while bonus pools rise. If senior seat counts are effectively capped, every promotion adds cost and uses a slot. That is our inference about the economics, not a stated bank policy.
• It is invisible to the press. A person who repeats a year or quietly leaves never appears in a quarterly headcount tally.

This section reflects our market observations, not disclosed policies. See the next section for what the MD data does and doesn’t show.

Attrition and backfill review. Wells Fargo has posted 24 consecutive quarters of headcount reduction, and CEO Charlie Scharf has said he prefers natural attrition where possible. The bank reportedly cut about 3,500 in the second quarter, to roughly 197,000. Bank of America held headcount roughly flat at 213,000 through 2025 and hired about 17,000 people, largely to backfill departures. Its CFO described reviewing whether a role needs to be refilled at all. Headcount shrinks with no announcement, just seats that never reopen. The old spring window treated attrition as something to wait out before deciding whom to cut. The new model treats it as a headcount tool in its own right, running all year.

Redeployment. JPMorgan held headcount near 318,000 while Jamie Dimon said on the July 14 call that AI has already cut headcount 30% to 40% in certain specific areas. He said the vast majority of affected employees were offered other roles. Net headcount looks flat while the mix changes underneath, and JPMorgan’s own results cite front-office hiring growth.

Structural programs. These are the true RIFs of the cycle, but note when they were planned. Citi announced its 20,000-role reduction in January 2024, after a $1.8 billion quarterly loss, cutting from a 239,000 base. With about 40,000 more jobs leaving via the Banamex spinoff, the target is 180,000. The initial reorganization eliminated about 5,000 mostly managerial roles. AI became part of the explanation later. Citi ended 2025 at about 226,000 employees, and its CFO expects headcount to keep falling in 2026 and beyond. The reported second-quarter figure of about 219,000 suggests the 20,000 target is already close by headcount, though Banamex has not yet left. HSBC has floated up to 20,000 cuts, about 10% of its workforce, and Commerzbank has announced 3,000. These programs were planned before the boom and are being executed inside it, when profits can absorb the cost.

5. What the MD Numbers Show

If promotion discipline were broad, MD classes would be shrinking. They aren’t, except at Citi.

Three takeaways:

• MD promotions are not being squeezed at the largest banks. Goldman and Morgan Stanley grew their classes. Citi is the exception.
• The tightening we describe sits below MD. Associate 3 and VP 3/4 outcomes aren’t captured in class-size data, so treat that as a market observation.
• Lazard shows the other side of MD management. The firm has been cutting and hiring MDs at the same time. The CEO says its MD investments are shifting from a headwind to a tailwind. That is repositioning, not a retreat.

6. Where AI Fits

AI is the obvious explanation for all of this, and executives are happy to point that way. The evidence is more nuanced.

Where it is clearly landing. Nearly every bank is aiming AI at the same functions: operations, middle-office reconciliation, client onboarding and KYC, regulatory reporting, compliance and audit.

• Goldman’s “OneGS 3.0” explicitly targets sales support, client onboarding, lending operations, regulatory reporting and vendor management.
• Citi CEO Jane Fraser told staff in January that the bank is “not graded on effort” and that AI would let it run middle-office and operational functions with fewer people.
• Wells Fargo’s CFO said the bank “should be able to run this company with less headcount,” and Scharf has called AI’s long-term staffing impact “extremely significant.”
• Bank of America’s Moynihan cited audit as a function AI could eventually support with a smaller team.

Where the evidence is thinner.

• Goldman’s own analysts have cautioned that clear evidence of layoffs directly motivated by AI remains limited, and that the more likely driver for many firms is cutting costs the old-fashioned way.
• Bloomberg noted in April that none of the executives that week linked their job reductions to AI.
• Revelio Labs data shows Goldman’s total headcount up about 6.4% to roughly 49,384 while active job postings fell 44.7%. That conflicts with Bloomberg’s finding that Goldman contributed to the Q2 decline, so its net trend is the least settled of any bank. AI may show up first as slower hiring, not as layoffs.
• Alan Johnson of Johnson Associates told Axios that hiring normally rises when business is booming and isn’t doing so this year, which he called a seismic change.

The savings aren’t free. Dimon warned that AI doesn’t uniquely benefit any one bank, because competitors adopt it too and the gains flow to customers. JPMorgan’s CFO said generative AI spending will climb sharply in the second half. Labor savings get partly recycled into compute.

The junior pipeline is the wrinkle. Goldman research has floated automating roughly a quarter of banking work hours, and reporting via McKinsey’s QuantumBlack described banks cutting incoming analyst classes by as much as two-thirds. Some banks are converting fewer summer interns, with candidates hearing “we don’t have the seats.” But a 2026 benchmark found even the best AI agent failed nearly half its evaluation criteria on real banking workflows, and human bankers rated none of the output client-ready. As we reported on Rogo, associates are spending real hours fixing wrong enterprise values and off footnote dates, and our searches skew toward seasoned associates and VPs.

AI’s effect on headcount looks less like mass replacement and more like three things:

1. It lowers the bar for cutting back-office seats.
2. It slows backfills and junior hiring.
3. It raises the premium on senior judgment.

The third is a talent-upgrade story, and it helps explain why promotion decisions are getting harder. If judgment is the scarce asset, banks want more scrutiny on who has it.

7. Is This Cheaper for Firms?

Broadly yes, but not because any individual departure costs less. The savings come from avoiding the hidden costs of the big-bang approach.

• Attrition is the cheapest tool. There is no severance, no announcement and no hard conversation. The cost is time, since reduction only happens as fast as people leave.
• Redeployment avoids severance on the way out and recruiting and training costs for a replacement hire.
• Promotion holds defer the cost of a promotion, involve no separation and keep an experienced person working. If the person leaves voluntarily, the firm has avoided a cut as well.
• Rolling performance cuts spread the expense across quarters, unlike a single large charge like Wells Fargo’s $612 million, and let a firm time it against strong results. The larger saving is precision. A broad RIF sweeps up producers along with underperformers, and replacing a good banker later means recruiting fees, guarantees and a ramp period. This is analysis, not a bank disclosure.
• Structural programs carry large one-time costs. Citi booked a $780 million severance charge when it launched its restructuring. Firms accept those costs in exchange for permanent savings and a stated headcount target.

Timing changes the math. The fall sweep was timed so severance didn’t collide with year-end compensation, and the spring window came after bonuses were paid. A cut in the middle of the year, months before pools are set, changes what a departing banker is owed and what the firm avoids paying. Whether that saves money depends on plan documents and severance terms, including deferred compensation, which is a legal question. It matters because 67% to 70% of MDs in our 2025 survey defer more than 25% of their bonus.

The costs on the other side:

• Uncertainty and regretted attrition. Removing the annual season replaces it with continuous doubt. The people most exposed are in the gray area, solid but not standouts, and they are also the most likely to take recruiter calls. Someone told to repeat a year may simply leave.
• Consistency and manager time. Handing timing to divisional leaders puts a sensitive decision in many more hands, and different desks may apply different standards.
• Legal exposure. When a cut rests on individual performance, the paper trail matters. A broad, market-driven RIF largely avoided this risk. Treat this as a question for counsel.
• The cost of AI itself. Savings have to be netted against spending like JPMorgan’s $19.8 billion technology budget and rising generative AI costs.
• The hours may only move. The cleanup work AI was supposed to remove has partly landed on associates and VPs, who cost several times as much per hour as the analysts whose work was automated.

8. Is This Less Public?

Yes, and that may matter as much to firms as the cost.

The old SRA and the fall and spring sweeps were known events on a known calendar. The rolling model has no calendar, no announced figure and no single moment for a reporter to write about. When Goldman began the April cuts, it didn’t disclose the number of affected employees, and a source told Reuters only that a small number of underperforming staff would go. Business Insider reported on the change through people familiar with it, so it doesn’t stay hidden entirely, but it is far harder to turn into a headline than a firmwide number.

The quietest tools never generate a story at all. An unfilled seat, a redeployment or a promotion that doesn’t happen has no press cycle. Much of what we know about 2026 came from Bloomberg’s tally of quarterly workforce disclosures, and promotion outcomes never appear in it.

Low visibility and low stigma are linked. The same features that keep a cut out of the press keep it from reading as distress.

There is also a narrative incentive. Goldman hasn’t officially linked AI to its April cuts, and “we are managing performance” is an easier message for employees, clients and regulators than “we are replacing people with software.” That is our read of the incentive, not a bank statement.

Visibility isn’t uniformly unwelcome. Citi’s structural program is announced, quantified and repeated by its CFO because investors need to hear the savings. Banks say the strategic story loudly and handle individual decisions quietly.

One legal caution: notice requirements such as the federal WARN Act and New York’s state version generally depend on how many employees are affected at a site, so smaller, staggered actions may fall below those thresholds. The rules are technical, and any firm relying on this should consult counsel.

9. The Talent Upgrade Is the Point

Put the tools together and the logic is upgrading more than cutting. A bank that trims a percent or two continuously, holds back some promotions, declines to backfill certain seats and moves other people into new roles isn’t shrinking its business. It is reshaping it toward producers, AI-fluent operators and experienced reviewers.

On the way out: rolling cuts, promotion gates and attrition remove people who aren’t clearly differentiated.

On the way in: as we wrote in “The Great Poach,” banks stopped casting wide nets and started running forensic, deal-by-deal interrogations of who did what on which transactions. They want proven A2s, A3s and junior VPs with public-to-public M&A experience and show almost no interest in uptiering candidates from lower-tier platforms. Evercore’s 19 senior MD additions and PJT’s citing of senior hires in its rising comp expense show the same pattern at the boutiques. Dimon has said JPMorgan will likely hire more AI specialists and fewer traditional bankers, so the mix is shifting even where headcount is flat.

The apparent tension. Banks are tightening promotion for associates and VPs while hunting for proven associates and VPs. These fit together once you see that banks want proven, seasoned people, not everyone who reaches a given tenure. A forensic lateral market and a stricter internal gate are two halves of the same sort.

Lazard is the clearest MD-level example. It has cut more than 80 MD roles while hiring and promoting, with a focus on healthcare, technology and defense. Its CEO said the firm’s strongest announced half-year league-table position since 2014 supports the view that its MD investments are turning from headwind to tailwind. Goldman’s rolling process applies the same logic at bulge-bracket scale, and the promotion gate applies it in the middle of the pyramid.

10. So, Are RIFs Dead?

Not dead, but demoted, and largely replaced in the day-to-day by something with a different name.

• The cyclical RIF is fading. Waiting for a downturn, cutting broadly and blaming the market is no longer the default.
• Performance-based tools are doing the work. Rolling cuts, promotion holds and managed exits accomplish much of what a RIF would, without eliminating positions on paper and without the headline.
• The calendar is blurring. Year-round review is replacing the two-window pattern, at least at banks using the rolling model.
• The structural RIF is alive. Citi, HSBC and Commerzbank show that large announced programs still happen when a bank wants to change its shape, and AI gives them a strategic reason.
• The old ritual may return. A source told Business Insider that a more traditional SRA could still occur later in the year, so rolling cuts may supplement the annual cull, not replace it. October and November are the window to watch, since that is when the fall sweep has traditionally landed and promotion decisions get made.
• The test hasn’t come yet. Every data point here comes from a record revenue environment. A bank running rolling cuts and tight backfills in a boom may still reach for a classic RIF in a real downturn, though from a leaner base. Cutting the back office now so the front office isn’t cut later is a hypothesis, not proof.

11. What This Means for You

Associate 3s. The promotion year is now a real gate, and “repeat the year” is a performance outcome, not a neutral delay. Know your deal contribution and what makes the case for promotion on your team. If the case is weak at your bank, compare the cost of waiting a year against a lateral move. If you are separated, ask how the exit will be characterized.

VP 3s and 4s. The same discipline applies at the step toward Director or MD, and banks are selective about both who moves up and who they bring in from outside. If you are told to wait, ask specifically what has to change and by when.

Producers. The environment rewards you. Bonus projections are up for revenue-generating groups and the lateral market wants proven deal experience. The rising tide no longer covers for anyone, so document what you originated and executed.

The gray zone. This model is hardest on solid-but-undifferentiated performers. Under the old calendar you could time a job search around the review. Now there is no season to brace for, so being ready to move at any point in the year matters more.

Senior bankers. In our 2025 data, MD (4+ Yrs) was the one level where the bonus recovery stalled, with only 47% seeing a higher bonus and 26% seeing a decline. Lazard’s advisory revenue fell in a record market, so a thin sheet has no macro excuse.

Operations, onboarding, reporting and compliance. This is where cuts concentrate at nearly every bank. Redeployment, the JPMorgan model, is possible but not guaranteed everywhere.

Junior bankers. The entry-level pipeline is tighter at some banks, driven by headcount allocation more than intern performance. The durable skill is knowing the mechanics well enough to catch an AI tool when it is wrong.

Everyone, on bonus timing. Because rolling cuts can happen well before year-end pools are set, check how your deferred compensation and bonus eligibility work if you are separated mid-year. The answer depends on your plan documents and any severance agreement.

Hiring managers and recruiters. The market looks flat in aggregate while churning underneath. Openings are created by cuts and promotion holds and filled with very specific profiles, so the value is in knowing exactly which profile a client is trying to buy. Ask how a candidate’s last role ended before assuming.

The Bottom Line

RIFs aren’t a thing of the past, but they are no longer the main instrument. A historic M&A market gives firms the profits to fund a change, and AI gives them a reason to rethink which seats they need. The result is a continuous toolkit of performance-based cuts, promotion gatekeeping, attrition discipline, redeployment and a small number of large structural programs. Only the last resembles the layoffs of the past. The rest are cheaper than a blunt RIF and far quieter, and they never carry the label. The shift also has a human cost: firms avoid the stigma of the RIF label, and the person who is cut inherits it. Whether the toolkit holds when the cycle turns is the open question, and this is the first cycle where we get to find out. With the traditional fall window and promotion decisions close, the next few weeks will show how much force the old calendar still has.

Sources

• Bloomberg, “Wall Street Banks Cut Over 10,000 Jobs Despite Record Trading Quarter” (July 16, 2026)
• Bloomberg, “Wall Street Banks Cut 5,000 Jobs Even as They Notched Record Profits” (April 15, 2026)
• Bloomberg, “Wall Street Eliminated 10,600 Jobs Last Year, Most Since 2016” (January 15, 2026)
• Bloomberg, Johnson Associates bonus forecast coverage (August 5, 2026), and Axios, “Wall Street bonuses are poised to rise” (August 5, 2026)
• Lazard, Second Quarter and First Half 2026 Results (July 23, 2026), and Reuters, “Lazard has cut over 80 managing director roles, CEO says” (July 23, 2026)
• Evercore, Second Quarter 2026 Results and earnings call (July 29, 2026)
• Houlihan Lokey, First Quarter Fiscal 2027 Results (July 29, 2026)
• JPMorgan Chase, Q2 2026 earnings call (July 14, 2026), and 2026 technology spending guidance
• Business Insider and Reuters reporting on Goldman Sachs’ rolling cuts and the Strategic Resource Assessment (March 2026)
• Goldman Sachs MD class of 2025 announcement (November 2025); Morgan Stanley MD classes (Bloomberg, January 2025; Business Insider, January 2026); Citi MD class of 2025 (Business Insider)
• Reuters, “Citi to cut 20,000 jobs through 2026” (January 2024), and Banking Dive on Citi’s fourth-quarter 2025 call (January 14, 2026)
• Global Finance, “Big Banks Signal Strong 2nd Half After Q2 Earnings Soar” (August 19, 2026)
• Q2 2026 and Q4 2025 earnings calls and releases for Goldman Sachs, Morgan Stanley, Bank of America, Wells Fargo, Citigroup, PJT Partners and Moelis
• Revelio Labs workforce data on Goldman Sachs, as cited in our August 1 article
• Prospect Rock Partners, “Record Profits, Deeper Cuts” (August 1, 2026); “What Investment Bankers Can Actually Expect to Earn in 2026” (August 2026); “Goldman Sachs Layoffs 2026” (March 21, 2026); “Rogo Was Supposed to Shrink the Junior Class” (August 21, 2026); “When Do Layoffs Typically Occur in Investment Banking?” (August 27, 2024); “The Great Poach” (November 2025); 2025 Investment Banking Compensation Report (March 2026, N=866)