The Pre-IPO File at $3M+: How Tech-Wealth Borrowers Get Funded When Most of the Money Is Still Locked Up

A composite scenario. The engineering director calls Tuesday afternoon. Thirty-eight, runs a platform team at a public tech company that IPO'd eighteen months ago, holds $4.2M of vested RSUs still inside the company-imposed quarterly trading windows, $800K cash, and $1.5M of pre-IPO common stock at a side venture where he sits as a technical co-founder. He is under contract on a $2.8M Palo Alto primary residence with a thirty-day non-extendable close. His employer's in-house mortgage program has already quoted: conforming rate, 70 percent LTV, but the LTV calculation excludes everything except vested-and-already-sold liquid. That gives him a $560K loan ceiling against an $800K cash position. The wealth advisor at his RIA suggested SBLOC against the diversified sleeve she manages, which sits at $600K. His tax counsel sent him to you.

The next half hour decides whether you handle this or hand it back. You have seen RSU income on a tax return for a decade. You know what restricted stock is, you have closed jumbo files for tech borrowers before, and you have watched plenty of these route to the company-partnered program by default. What you do not have, until you do, is the 2026 lender-category map for a $2.2M loan against a balance sheet where most of the wealth is vested-but-illiquid, restricted, or pre-IPO.

You already know this product

You have done jumbo files for tech borrowers. You have seen RSU vesting schedules, you have read 10b5-1 elections on a tax return, and you have qualified borrowers off the cash they have sold and the W-2 the company reports. You do not need the 101.

What you need is the 2026 reality of where files live when the borrower is asset-rich but most of the wealth has not yet legally vested, has vested but cannot be sold inside a window, or sits in pre-IPO common stock with no public market. The veteran broker who treats this file as a normal jumbo gets out-leveraged by the company program on rate and out-structured by the private bank on the asset side. Treated as a different category, with a different bench, the file routes cleanly.

The tech-wealth segment behaves differently from DSCR, asset-depletion, foreign national, and pledged-asset. The borrower's net-worth statement is real. The liquidity profile underneath it is not. Vested RSUs inside a blackout window are not cash. Pre-IPO common is not collateral most lenders will price. A 10b5-1 plan with quarterly windows is a forecasting document, not a payment guarantee. The bench has to underwrite around all of that.

The 2026 lender-category map at $3M-plus tech-wealth

The bench has four entries. Three are real plays. The fourth is the desk to know about so you can position against it.

Specialty non-QM desks. Selective. A focused cohort of non-QM lenders has built underwriting that engages with vested RSUs on a lookback basis, treats restricted stock awards with proper documentation, and accepts the structural realities of the post-IPO trading window. They typically will not price pre-IPO common as collateral and will discount single-stock concentration heavily. Twenty-one to thirty-five days from clean intake to close. This is the workhorse desk for the tech borrower whose primary qualifying complexity is RSU income concentration and recent-IPO timing.

Private banks. The structural workhorse on the asset side. The product is rarely a stand-alone tech-wealth mortgage. It is a pledged-asset structure against the vested liquid sleeve, sometimes blended with a smaller jumbo against the property. The bank can credit a wider asset base than the spec non-QM desk because it controls a custody account and marks daily. For the borrower with $4.2M of vested RSUs held at a custodian the bank can integrate with, the math opens up. Available only to existing relationship borrowers or borrowers willing to move the AUM.

Correspondent paths via specialty wholesale. Under the radar. A growing set of independent mortgage banks operate correspondent channels with non-bank investors funding tech-wealth files at the upper jumbo tier without putting the program on a public rate sheet. Not on aggregator portals. Accessed by direct relationship. For the broker who has built the bench, this category quietly wins a meaningful share of files where the company-partnered program is the headline competitor.

Tech-employer mortgage partnerships. The default the borrower will hear about first. Most large public tech employers run a partnered program with one specific lender at conforming or near-conforming rate. The trade is usually structural: low rate, restrictive LTV against liquid only, no credit given to RSUs that have not been sold, no engagement with pre-IPO equity. The program serves the dual-W-2 engineer buying a $1.4M starter home cleanly. It does not serve the $2.8M file where the wealth profile is restricted-and-illiquid. Know the program exists, know what it will and will not do, 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 company-partnered program well enough to articulate the trade the borrower is being asked to make.

The equity-comp taxonomy lenders actually use

The asset side is where the file lives. The borrower will use the terms loosely. The lender will not.

Vested RSUs held in a brokerage account. The cleanest piece of the file. Most specialty non-QM desks count vested-and-deliverable RSU shares at 60 to 75 percent of trailing thirty-day average market value, subject to single-stock concentration discounts. Underwriting treats these shares as liquid securities with a sale restriction limited to the company's blackout calendar. Verification: a current brokerage statement, the active 10b5-1 plan if any, and the company's trading-window schedule.

RSUs vested but inside an active blackout window. Counted at the same haircut for asset purposes by most desks if the window will release before close. If the window releases after close, the file routes to a desk that engages with that timing or to a private bank that can price against the position via pledged-asset.

Vesting schedules and forward income. RSU income is most of the W-2 line. The desk wants the vesting schedule on company letterhead, the most recent two pay stubs showing in-period vesting, and ideally a 10b5-1 plan. The lookback convention varies. A common 2026 pattern is twenty-four months of trailing RSU income averaged, with a haircut on forward years that are unvested or contingent. The mistake is treating the entire grant value as forward income. The lender wants vested-and-paid history plus a credible forward picture, not a hypothetical.

ISOs and NSOs. Different beasts. ISOs that are vested but unexercised do not count as income or assets. Once exercised and held, the underlying shares may count as restricted stock subject to AMT and concentration rules. NSOs hit the W-2 at exercise and can spike a single year, creating a noisy two-year average. The desk will want the option-grant history, the exercise pattern, and a clean reconciliation against the W-2 and 1099 trail.

Pre-IPO common stock. Almost universally not counted at face value. The pre-IPO position is the borrower's asset. It is not the lender's collateral. A small set of private banks will price a pledged structure against pre-IPO common where the issuer has a defined secondary market or a recent funding round at a documented per-share price, with advance rates usually 25 to 40 percent of the round price. For the file in front of you, default to assuming the $1.5M of pre-IPO common is balance-sheet color, not lendable collateral.

Founder vesting schedules with reverse-vesting cliffs. The pre-IPO co-founder position is often subject to a reverse-vesting agreement: shares issued but contractually clawed back if the founder leaves before a milestone. Pre-cliff, the position underwrites at zero. Post-cliff and inside the vesting curve, sometimes a fractional credit is available. The borrower's instinct to count the full $1.5M is wrong.

Restricted stock awards under Rule 144 holding periods. A separate question from RSUs. RSAs granted to an early employee or founder may sit inside a Rule 144 affiliate holding period after IPO. Typical six-month minimum holding plus volume limitations on subsequent sales. The desk will want the affiliate letter, the 144 trail, and a clean read on whether the position can be sold inside the file's relevant window.

The recent-IPO and lock-up choreography

The lock-up clock is the most underappreciated underwriting variable on this file.

Standard post-IPO lock-up runs 180 days from the offering. Some companies stagger it, some run early-release on a fraction of insider shares, some shorten the window via an SEC-cleared modification. Inside the lock-up, vested RSUs and any RSAs are technically owned but legally unsellable. Most non-bank lenders will not lend against locked-up shares at any haircut. A small set of private banks will price a pledged structure inside the lock-up at a deep discount, but the file becomes a relationship transaction, not a rate transaction.

The window the file usually wants is the ninety days after lock-up expiration. The borrower has shares that have legally cleared the lock-up and is sitting on a position the desk can engage with on either an asset basis or a pledged basis. Trading-window blackouts still apply, and the 10b5-1 plan still governs scheduled sales, but the structural barrier is gone.

For the file in front of you, the IPO eighteen months prior means the lock-up cleared a year ago. You are inside the workable window. For a borrower four months post-IPO, the file either waits, the borrower brings outside cash, or it routes to a relationship private bank that can structure inside the lock-up.

The 10b5-1 plan is the underwriter's friend. A pre-cleared scheduled-sale plan creates predictable liquidity events. A borrower with an active 10b5-1 selling roughly $200K per quarter has a forecastable cash trajectory the underwriter can plug into reserves and forward-income math. A borrower dependent on discretionary trading-window sales presents the same wealth picture and a noisier underwriting profile.

The pitfall is timing alignment. A 10b5-1 selling in March against a close date in early February does not produce cash in time. A blackout window that opens four days after the close-by date does not close the file. Pull the company's trading calendar and the borrower's 10b5-1 schedule at intake. Match them against the close-by clock before submission.

The asset-pledged path against vested-but-restricted equity

The private bank category exists on this file because of one structural fact: the bank can price an over-collateralized pledged structure against a single-name concentrated position the specialty non-QM desk will not engage with on an asset basis.

Typical 2026 advance rates against a single-stock vested RSU position pledged to a private bank: 50 to 70 percent of marked-to-market value, with an additional concentration discount of roughly 25 percent on positions that exceed a quarter of the pledged base. A borrower pledging a $4M single-stock position into a structure where the bank also holds $1M of diversified securities sees the single-stock haircut stack on top of the concentration penalty. The number that comes out the other end is meaningfully smaller than the borrower's mental model. Set the expectation in the first conversation.

Margin-call mechanics are the second piece. A single-stock pledged structure marks daily, and a bad week for the underlying ticker can trigger a margin call inside a structure that worked at origination. Walk the borrower through cure-window mechanics, margin-call protocol, and the realistic worst-case sequence before signing.

For the engineering director, the math directionally: $4.2M of single-stock pledged at 60 percent advance, less a 25 percent concentration discount on the portion above the quarter-of-base trigger, produces a working collateral base in the low $2M range. With $800K of additional cash, the structure supports a $2.2M mortgage against the $2.8M property. The pre-IPO common is not in the math. The 10b5-1 plan governs the forward sale schedule and supports the reserve picture.

The strategic position against the company-partnered program

The borrower's first mortgage conversation usually happens through HR or a benefits portal. The company-partnered program is a real benefit. It is also calibrated for a different file than the one in front of you.

The program's pitch is rate. The trade is structure: conforming or near-conforming pricing in exchange for a tight LTV cap against liquid-only assets, no engagement with restricted equity, no flexibility on the close timeline beyond the partner lender's standard pipeline. For the dual-W-2 engineer with a clean cash position buying a $1.4M starter home, the program works. For the file you are looking at, it fails on the asset side and gets out-leveraged by a private bank pledged structure on the rate side.

Position cleanly in the first call. The company program is a real option. It will quote a sharp rate and a $560K loan against $800K cash. The broker bench can run a $2.2M loan against the same balance sheet by engaging with the asset profile the program ignores. The trade is a modest rate premium for a structurally fundable file. Let the borrower do the math.

The wealth advisor's SBLOC suggestion is also a real option. Against the $600K diversified sleeve, the SBLOC produces a $400K to $480K line at SOFR plus a spread. Liquidity, not a mortgage. Useful as a reserve sleeve or as a piece of a blend, not the primary structure for a $2.2M acquisition.

The broker's lane is unbundling. Engage with the asset profile the program ignores, structure across the equity-comp taxonomy the program excludes, and put a fundable answer on the table inside seventy-two hours.

A composite transaction

Tuesday, 4:14pm. The engineering director. Thirty-eight, eighteen months post-IPO at his public-tech employer, $4.2M of vested RSUs at one custodian, $800K cash at the same custodian, $1.5M of pre-IPO common at his side venture under a reverse-vesting agreement that cleared its cliff fourteen months ago and now sits at 78 percent vested. Active 10b5-1 plan selling roughly $180K per quarter, scheduled cleanly through the next four windows. No outside W-2 income beyond the employer.

The company-partnered program has put $560K on the table against the cash. The wealth advisor has mentioned SBLOC. He has thirty days, a $2.8M Palo Alto contract, and two competing offers behind him on the property.

You run the intake. Equity-comp picture clean: RSUs vested and trading-window-deliverable inside the close clock, lock-up cleared a year ago, 10b5-1 supports forward reserves, pre-IPO common is balance-sheet color, no AMT trail that complicates the picture. Single-stock concentration is the obvious file-killer if mishandled.

The intake routes the file. Specialty non-QM desk first, private bank desk as secondary on the asset side, correspondent partner as tertiary.

You submit Wednesday morning to a specialty non-QM desk you have used twice on similar profiles. The desk engages on a hybrid: $1.6M against the property, qualifier built on the trailing twenty-four months of RSU income, 10b5-1 supporting forward reserves at thirty-six months. Conditional approval Friday. Documentation is the heavy lift: vesting schedule on company letterhead, trading-window calendar, two years of pay stubs, the 10b5-1 plan, recent brokerage statements, the side-venture cap table, and the reverse-vesting agreement showing the post-cliff position.

In parallel you run the private bank track. The bank prices a pledged-asset blend at SOFR plus 1.45 percent against a $3.5M pledge of vested RSUs and cash, advancing roughly $2.1M, with a $200K standalone reserve sleeve untouched. Rate is sharper than the spec non-QM track. Structure encumbers the entire vested liquid position, including the cash needed for the borrower's next two quarters of household and tax flow.

You walk the borrower through both. The spec non-QM track funds at 6.85 percent, no asset pledge, no AUM commitment, full retention of the cash sleeve, and supports the diversification plan the wealth advisor is already running. The bank track is sharper on rate, encumbers the position, and complicates the diversification timing. The borrower picks the spec non-QM track.

Day twenty-six. Clear to close. $1.6M funded, the borrower bringing $1.2M of cash from trailing 10b5-1 sales plus an interim window-release sale. Rate: 6.85 percent thirty-year fixed, no prepay, no asset pledge.

The wealth advisor calls the next week. Two colleagues at the RIA have similar profiles in motion. She wants to know if you handle that file the same way.

What this changes about your business

The veteran who has closed jumbo files for tech borrowers has the skill. What is missing is the equity-comp taxonomy fluency, the bench that engages with vested-but-illiquid positions, and the strategic position against the company-partnered program. None of these are hard. The work has not gotten done because tech-wealth files have felt like a regional curiosity rather than a primary product line.

In 2026 that shifted. The tech-wealth segment is structurally large, distributed across a half-dozen geographic concentrations, and underserved by the broker channel. The company-partnered program serves the easy half. The private banks serve the relationship half. The middle is where the broker plays.

Stop treating the pre-IPO and post-IPO file as the loan you run when nothing else fits. Treat it as a primary lane for the engineer, the founder, the executive who has aged into wealth that is mostly equity-comp and mostly restricted. The cohort is routed by tax counsel, RIAs, and estate attorneys who already trust you on adjacent files.

The play to run this week

1. Build the three-name tech-wealth bench. One specialty non-QM desk that engages with RSU income on a trailing lookback, one private bank desk where you have a direct contact for pledged-asset structures against single-stock positions, one correspondent partner for the upper-tier file. Confirm each desk's RSU haircut, lookback period, single-stock concentration treatment, lock-up policy, and 10b5-1 documentation requirements.

2. Pull the equity-comp taxonomy onto a one-page intake brief: vested RSUs, RSUs in blackout, ISOs, NSOs, pre-IPO common, founder vesting schedules, RSAs, lock-up status, 10b5-1 plan status. Make it the second page of every tech-borrower intake call.

3. Read the most recent 10-Q section on insider trading windows for any public tech employer in your top three referral zip codes. Internalize the trading-window cadence so you can read a borrower's 10b5-1 against the company's calendar at intake.

4. Identify the two highest-value tech-borrower files you closed in the last twenty-four months at the company-partnered program rate. Ask each borrower what the balance sheet looks like today. The wealth has compounded since you saw it. The next acquisition is closer than the file thinks.

The first $2M-plus tech-wealth file you close on the broker bench resets the math of your month. The second resets the trajectory of the year.

The work is to learn the equity-comp taxonomy cold, build the three-name bench, run the lock-up and 10b5-1 timing against the close clock, and put a fundable answer on the table inside seventy-two hours. The company-partnered program cannot match that without rebuilding underwriting it has chosen not to build. The bank cannot match it without bundling the AUM.

---

The next Cornerstone in The HNW Lending Atlas covers the carried-interest file: how 2026 lenders treat partnership interests in private equity and hedge fund GPs, the K-1 patterns that route to one bench versus another, and the deal-killers that surface at the partnership-agreement review rather than at the borrower file.

---

Compliance note. Authority Graph is not a lender, mortgage broker, financial advisor, attorney, 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, single-stock concentration, lock-up, 10b5-1, RSU haircut, and underwriting figures vary by lender, jurisdiction, employer policy, borrower profile, and the contemporary regulatory environment. Composite scenarios are illustrative and do not represent specific real persons or transactions. Equity compensation, pledged-asset, and securities-based lending structures involve material tax, legal, and market risks; nothing in this article constitutes financial, legal, or tax advice. Consult licensed mortgage, legal, tax, and securities 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.