The Carry File: How $4M+ Loans Get Done When the Income Is K-1, Lumpy, and Pegged to a Fund Vintage

A composite scenario. The PE partner calls Tuesday morning. Forty-four, eight years past promotion at a $5B AUM mid-market firm, two funds in carry across four active vintages, $850K W-2, four trailing years of carried-interest distributions on K-1: $1.8M, $4.2M, $1.1M, $3.6M. Under contract on a $5M Aspen second home, forty-five-day close. The firm's relationship private bank has quoted an asset-pledged structure at SOFR plus 2.5 percent against his GP interests. He does not want to pledge them. His tax counsel sent him to you.

The next half hour decides whether you handle this or hand it back. You have qualified borrowers off complicated returns and you have closed jumbo files for finance-sector borrowers. What you do not have, until you do, is the 2026 lender-category map for a $4M-plus loan against an income statement where carry is the dominant line and the desk has to underwrite four years of vintage-driven volatility into a single qualifying number.

You already know the partnership return

You have read a K-1. You know guaranteed payments and the difference between ordinary income and long-term capital gain on the partner's return, and you have closed jumbo files for finance-sector borrowers off W-2 plus a partnership interest. You do not need the 101.

What you need is the 2026 reality of where the carry file lives. The veteran broker who treats it as a standard high-income jumbo gets out-priced by the firm's relationship private bank on the asset side and a polite no from a non-QM aggregator that does not read the K-1 line items that matter. Treated as a different category, with a different bench and a defensible qualifying narrative, the file routes cleanly.

The carry file behaves differently from DSCR, asset-depletion, pre-IPO, and pledged-asset. The W-2 is real but rarely the qualifying number. The K-1 is real but volatile across vintages, lumpy across years, and tied to fund-level realization events the partner does not control. A partner three years past promotion has a different file than the same partner eight years in. A partner whose flagship fund is in harvest year four has a different file than the same partner whose flagship sits in deployment year one. The bench has to underwrite around all of that.

The 2026 lender-category map at $3M-plus carry files

The bench has three real entries and a fourth desk to know so the broker can position against it.

Specialty non-QM desks with finance-sector underwriting. Selective. A focused cohort of non-QM lenders engages with K-1 carried-interest distributions on a multi-year lookback, reads the partnership agreement and the fund LPA at a level that matters, and treats W-2 base and bonus as a separate qualifying floor. Most do not lend against the GP interest as collateral. They will qualify on a defensible carry-income narrative. Twenty-eight to forty-five days from clean intake to close. The workhorse desk for the partner whose primary qualifying complexity is K-1 lumpiness rather than asset structure.

Private banks. The structural workhorse on the asset and relationship side, and the desk the borrower probably already has a custody relationship with. The firm itself often banks at one or two specific institutions, and the partners get routed there by default through the firm's preferred-lender channel or the natural overlap between subscription banking and personal banking. Private banks engage with carry in two flavors: a relationship jumbo qualified off W-2 and averaged carry, or an asset-pledged structure secured against future distributions from defined fund vintages. Relationship jumbo pricing is the sharpest in the market when the AUM relationship is deep. The GP-interest pledge runs above standard SBLOC because the collateral is unrealized and tied to fund performance.

Independent specialty desks running the partnership-agreement read. Under the radar. A small set of independent mortgage banks and non-bank lenders staff underwriters who can read a partnership agreement, a fund LPA, and a clawback provision without flinching. Pricing runs above the bank's relationship rate and at or modestly below the spec non-QM rate. The structural advantage is documentation flexibility. They will engage with a partner three years past promotion who does not yet have five years of carry, with a partner whose firm changed sponsors in the trailing window, or with a file that requires a side letter from the GP confirming distribution rights. The bank rarely does that work for a non-AUM borrower.

The firm's preferred-lender channel. The default the borrower hears about first. Most mid-market and upper-mid-market PE, hedge fund, and venture firms run a preferred-lender relationship with one specific private bank. Sharp rate, familiar onboarding because the bank already knows the firm, and an asset-side pitch that often pivots toward pledging the GP interest. Know the channel exists, know what it will and will not flex on, and position the broker bench against it on structure rather than rate.

The 2026 move: build a three-name bench across the first three categories, and know the preferred-lender quote well enough to articulate the trade the partner is being asked to make.

How carry actually gets underwritten

The trailing-distribution lookback. The 2026 standard at the spec non-QM and independent specialty desks runs three to five years of K-1 carried-interest distributions averaged into a single qualifying number, with a haircut when year-over-year volatility exceeds a threshold. A four-year stream of $1.8M, $4.2M, $1.1M, $3.6M averages to $2.675M. The desk typically applies a discount and qualifies off something closer to $2.0M to $2.3M, depending on the underwriter's read of the fund cycle behind the numbers. W-2 base and bonus stack on top as a separate qualifying floor.

Cash-distributed carry versus accrued-but-not-distributed. The desk wants cash. Allocated K-1 income that has not been distributed because the fund is still inside its waterfall or subject to a clawback reserve does not qualify the same way. Pull the distribution history from the fund administrator at intake and reconcile against the K-1 trail before submission.

Recently-promoted partners. A partner three years past promotion does not have a five-year history. Most desks work with a shorter lookback when the partner has a clean W-2 trail through the promotion year, the prior funds show realized distributions to comparable partners, and the current vintage has begun returning capital. Some will not. Read the timing in the first call and route accordingly.

The fund-vintage read. A partner whose carry is concentrated in one flagship vintage that just exited two strong portfolio companies has a different forward picture than the same partner whose flagship is in deployment year one with no realizations on the horizon. Pull the fund-by-fund attribution at intake. The firm's controller will produce it on request.

Clawback provisions. Most fund agreements include a GP clawback requiring the partner to return prior distributions if the fund's overall performance falls below a hurdle by wind-down. Often partially offset by escrow holdbacks, but the underwriter will ask. A clean track record and a strong escrow position underwrites at a smaller haircut than a flagship vintage that has under-performed peer median to date.

The 2017 holding-period rule. A reminder. The 2017 tax law change extended the holding period for long-term capital gain treatment of carry to three years. The desk does not underwrite tax treatment, only cash flow. Where the rule surfaces is on the recently-promoted partner whose first realization sits inside the window and faces ordinary-income treatment on an early distribution. The cash arrives. The desk underwrites the cash.

The lumpy-K-1 problem and the recent-deposit question

The most common deal-killer at the underwriting layer on a carry file is the recent-deposit question on a year that looks like an income surge.

A partner shows $1.8M, $4.2M, $1.1M, $3.6M across four years. The $4.2M year is the year the flagship returned a meaningful chunk of its waterfall. The $1.1M year is the gap between realizations. To the underwriter who has not seen carry before, the $4.2M looks like a non-recurring spike to discount. To the underwriter who has, it looks like a normal harvest-year distribution that the next vintage will, in some form, repeat.

Reconciliation against the fund-level distribution calendar. Pull the firm's distribution history by fund and tie each K-1 year to the underlying realization events. A $4.2M year that maps to two announced exits reads cleanly. A $4.2M year that does not reconcile to anything visible reads as a question.

The forward picture from the GP firm. A side letter or attestation from the firm's CFO or controller confirming the partner's allocation across active funds, the funds' current investment status, and a directional view of remaining anticipated distributions. The desk does not require it. Some desks weight the file more aggressively when it is in hand.

The recent-deposit question is particular. A $4.2M wire hitting the partner's primary checking account inside the trailing six-week window triggers the standard sourcing inquiry. The answer is the K-1, the partnership agreement, and the distribution-calendar reconciliation. Treated as a partnership distribution from a fund whose mechanics the underwriter can verify, it routes cleanly.

The loan-against-anticipated-carry structure

The structure the borrower's relationship private bank often pitches first.

The bank lends against the GP interest in defined fund vintages, secured by the partnership interest itself and an assignment of the right to receive future distributions. Pricing in 2026 has typically run SOFR plus 200 to 350 basis points depending on fund quality, vintage maturity, and the partner's tenure. Advance rates are conservative: most banks lend at 30 to 50 percent of a discounted projection of remaining anticipated distributions, with deeper haircuts on funds in early deployment and lighter haircuts on late harvest.

The collateral is contingent. A clawback event, fund underperformance, a change in sponsor, or a partner-level event at the firm can each impair the position. The bank prices the risk through spread and haircut. The partner prices it through the constraint on flexibility, since a pledged GP interest cannot be easily transferred, restructured, or posted as collateral elsewhere without consent.

The covenants are firm-specific. The bank will want a side letter from the GP confirming distribution mechanics, often a consent to assignment, and sometimes a direct-pay arrangement where distributions route through the bank's custody account. Some firms accommodate the structure as a matter of course. Others view it as administrative load and discourage it.

The exit is the realization curve, not a refinance. The structure amortizes against actual distributions or includes a balloon tied to the fund's expected wind-down. For the partner comfortable pledging the GP interest, the structure works. For the partner who is not, the bank's quote is not the answer.

Strategic position against the firm's relationship bank

The firm-level banking relationship is the structural fact that decides most carry files by default.

A mid-market PE firm with $5B AUM banks somewhere. The fund administrator sits there, capital-call mechanics route through there, the partners bank there because it is convenient and because the firm pre-negotiated terms. The relationship bank knows the funds, the partners by name, and has a pitch for any partner-level financing the partners need.

The bank wins this file by default in 2026. Not by superiority on every dimension. By proximity.

The veteran broker's edge is rarely rate. When the firm-level relationship is deep, the bank wins the rate fight. The edge is everywhere else.

The structural unbundling. The bank's instinct on a carry-heavy partner is to pitch the asset-pledged structure that ties the personal financing to the GP interest. That produces sharp rate and bundles relationship economics on the bank's side. It also constrains the partner's flexibility in ways some partners do not want once they understand the side-letter mechanics. The broker who can quote a clean carry-qualified jumbo against W-2 and averaged K-1 distributions, no GP pledge, no consent-to-assignment, no firm-level side letter, has a structurally different product on the table.

The qualifying-narrative work. A defensible number on a lumpy K-1 stream requires the distribution-calendar reconciliation, the fund-vintage attribution, the clawback read, the W-2 stack, and a coherent story the underwriter can put in the file. The independent specialty desk often runs this more competently than a generalist bank underwriter because the desk specializes in finance-sector borrowers. The broker who builds the comparison on a one-page brief inside the first call has a position the bank's standardized pitch cannot match.

State the position cleanly. The relationship bank is a real option and the firm-level proximity is genuine. The broker bench produces a different product structure, a different set of covenants on the GP interest, and a faster timeline when the fund-level mechanics are mid-cycle. Let the partner do the math.

A composite transaction

Tuesday, 11:42am. The PE partner. Forty-four, eight years past promotion at his $5B AUM firm, two funds in active carry across four vintages, $850K W-2 (roughly $475K base and $375K bonus), four-year K-1 carried-interest stream of $1.8M, $4.2M, $1.1M, $3.6M. $3.5M liquid taxable at the firm's relationship private bank. $1.2M in a 401(k). No outside business income.

The bank has put a $3.8M asset-pledged structure on the table at SOFR plus 2.5 percent against a partial pledge of his GP interest in two active funds. He has a forty-five-day close on a $5M Aspen contract and one backup offer behind him.

You run the intake. Three facts route the file. The partner will not pledge the GP interest because the firm's general counsel has flagged the consent-to-assignment as a complication for an upcoming fund-level restructuring on the flagship. The trailing distributions reconcile cleanly against the firm's announced realizations across the four-year window. The W-2 is steady. The 401(k) is not in the qualifying math.

You run the averaging. Trailing four-year K-1 carry averages $2.675M. Apply a volatility discount and qualify off $2.1M of carry plus the $850K W-2 for a $2.95M qualifying-income line. Reserves are real, balance sheet clean outside the GP interests. The numbers support a $3.5M to $3.8M loan against the $5M property at 70 to 76 percent LTV.

Intake routes the file. Independent specialty desk first; spec non-QM as a secondary track on rate. The firm's relationship bank stays in the conversation as the comparator.

You submit Wednesday morning to an independent specialty desk you have used twice on similar files. The desk engages on a $3.6M jumbo at 7.10 percent thirty-year fixed, no prepay, qualified off the trailing four-year K-1 average plus the W-2, no GP pledge, no consent-to-assignment, no firm-level side letter. Documentation: fund-by-fund attribution from the controller, the four-year K-1 trail, the partnership agreements, a recent statement from the relationship bank.

Conditional approval Friday. The controller produces the attribution inside the week. Documentation runs cleanly through day twenty-eight.

In parallel you run the relationship-bank track for the comparator brief. The bank holds the asset-pledged quote at SOFR plus 2.5 percent, currently pricing all-in around 7.45 percent, against the partial GP-interest pledge. The bank also offers a relationship jumbo at 6.85 percent with no GP pledge but conditioned on moving an additional $4M of AUM to the bank within ninety days of close.

You walk the partner through the three options on a one-page brief. The bank's asset-pledged structure: sharper rate, GP pledge, consent-to-assignment complications during the flagship restructuring. The bank's relationship jumbo: sharpest rate, no GP pledge, AUM commitment that pulls $4M from existing custody. The independent specialty jumbo: twenty-five basis points above the bank's relationship rate, no GP pledge, no AUM commitment, no firm-level paperwork, closes inside the contract clock.

The partner picks the independent specialty track. Day thirty-six, clear to close. Day forty-three, funded.

The tax counsel calls the next month. Two partners at a peer firm have similar profiles and similar reluctance to pledge GP interests during a restructuring. He wants to know if you handle that file the same way.

That is the lane.

What this changes about your business

The veteran who has read partnership returns has the skill. What is missing is the K-1 averaging fluency, the bench that engages with carry as primary qualifying income, the reconciliation discipline against fund distribution calendars, and the strategic position against the preferred-lender channel. None of this is hard. It has not gotten done because carry files have felt like a sub-segment of finance-sector jumbo rather than a primary product line.

In 2026 that shifted. The PE, hedge fund, and venture-partner cohort is structurally large, geographically concentrated, and routed through tax counsel and trust-and-estate attorneys who already trust the broker on adjacent files. The firm's preferred-lender channel serves the easy half, the relationship bank the AUM-bundled half. The middle is the broker's. Treat the carry file as a primary lane for the partner three to twelve years past promotion who has aged into a balance sheet where carry is the dominant income line and the GP interest the dominant asset.

The play to run this week

1. Build the three-name carry-file bench. One spec non-QM desk engaging with K-1 carry on a multi-year lookback, one independent specialty desk that staffs underwriters who read a partnership agreement and a clawback provision, one private bank contact for the asset-pledged or relationship-jumbo path. Confirm each desk's lookback, volatility discount, recently-promoted-partner policy, GP-pledge stance, and side-letter requirements.

2. Pull the carry taxonomy onto a one-page intake brief: trailing K-1 carry by year and by fund, W-2 base and bonus, fund-vintage attribution, distribution status by fund, clawback exposure, GP-pledge willingness, firm preferred-lender relationship. Make it the second page of every finance-sector partner intake call.

3. Read one publicly-available LPA from a representative mid-market PE fund. Internalize the distribution waterfall, the clawback provisions, and the consent-to-assignment language. The 2026 partner you are about to qualify operates inside a document with structurally similar mechanics.

4. Identify the highest-value finance-sector partner files you have closed in the last twenty-four months. Ask each borrower what the K-1 stream looks like today and what the next vintage is doing.

The first $4M-plus carry file you close on the broker bench resets the month. The second resets the year.

The work: learn the K-1 averaging cold, build the three-name bench, run the partnership-agreement read inside the close clock, and put a fundable answer on the table inside seventy-two hours. The preferred-lender channel cannot match that without rebuilding underwriting it has chosen not to build. The bank cannot match it without bundling AUM the partner does not want to move.

---

The next Cornerstone in The HNW Lending Atlas covers the estate-tax bridge file: how 2026 lenders fund the nine-month liquidity gap when the patriarch dies and most of the estate is locked inside an operating business, the underwriting realities the executor confronts, and the strategic position with the trust-and-estate counsel that produces the file.

---

Compliance note. Authority Graph is not a lender, mortgage broker, financial advisor, attorney, accountant, or licensed financial professional. The content above is educational and reflects the author's interpretation of publicly reported market dynamics and industry conventions as of May 2026. Specific rate, LTV, advance-rate, lookback-period, volatility-discount, clawback, side-letter, and underwriting figures vary by lender, fund, jurisdiction, partner profile, partnership agreement, and the contemporary regulatory environment. References to the 2017 carried-interest holding-period rule and to fund-level mechanics are general and illustrative; partnership tax treatment, distribution mechanics, and GP-interest pledge structures must be evaluated by qualified tax counsel and qualified fund counsel for the specific partner. Composite scenarios are illustrative and do not represent specific real persons, firms, or transactions. Carried-interest, partnership-distribution, and GP-pledge financing structures involve material tax, legal, and fund-level risks; nothing in this article constitutes financial, legal, or tax advice. Consult licensed mortgage, legal, tax, and fund-counsel professionals for guidance specific to your situation.

---

Field Notes is published weekly by Authority Graph, a service for top-producing mortgage brokers serving high-net-worth and ultra-high-net-worth clients. The Cornerstone Authority Piece included in The Authority Rise is built in the broker's voice for exactly the kind of compounding the rest of this piece is about.