How Far the Qualifying Envelope Actually Moves: Escalators, Bonus Seasoning, and Pledged Assets
The question
Verbatim: "What expands a jumbo construction-to-perm qualifying envelope beyond the ~$1.0-1.2M DTI cap at $315k gross — specifically, how do Florida portfolio lenders underwrite bonus and cert-escalator income at under 12 months of W-2 tenure, and how is a pledged-asset / securities-backed structure treated in DTI versus a cash down payment?"
Context: this merges two of the five follow-ups spawned by [[2026-08-14-construction-to-permanent-loans-florida]]. Both ask the same underlying thing, which is how far the income-side cap can move, and both resolve against the same underwriting-guide layer. The parent brief's closing line was "the loan is not the lever that makes an $1.8M build work; income is." This brief tests that claim quantitatively.
⚠️ THE $315k PREMISE IS STALE BY ONE ESCALATOR. Combined gross is $325,000 as of 2026-08-17, not $315,000. Escalator #1 paid on the Anthropic CCA-F pass and moved base from $195k to $200,000 ([[cert-progress]]). Escalator #2 takes it to $205,000 around late Nov 2026, putting combined at $330,000. Every figure below is recomputed on the current numbers, not the [[readme]] baseline the parent brief inherited.
What we already know (from the vault)
- The cap the parent brief carried forward was never derived, it was inherited. [[readme]] states "Practical mortgage ceiling on income alone: ~$1.0-1.2M" and the parent brief affirmed it. Reverse-engineering the README's own table ($500k loan → $3,900 PITI, $1.2M → $9,400) shows it modeled escrow as roughly 1.4%/yr of the LOAN amount. Escrow scales with property value, not loan size, and the finished value here is ~$2.5M in FEMA Zone AE ([[2026-08-04-flood-zone-ae-construction-loan-requirements]], [[2026-07-22-florida-ae-zone-insurance-repricing-tampa]]). The inherited ceiling is therefore optimistic by construction.
- Tenure is the binding constraint, and it is short. phData W-2 started 2026-05-26. At a Q1 2027 application that is ~9 months; at a mid-2027 application it clears 12. Prior income was 1099 through Ray Data LLC (Mammoth Growth, ~$225-230k), which phData replaced rather than added to.
- The escalators are base salary, not bonus. [[cert-progress]] records the founder verbatim on 2026-08-17: "officially locked in my first tranche of salary boost. I'm at $200k base now." Two order-agnostic +$5,000 legs, confirmed against the primary vault doc rather than assumed. The parent brief lumped "bonus and cert-escalator income" together as variable comp. That framing is wrong and it costs us the answer to question 2. (
01-projects/phdata/milestones.mdstill shows the superseded fixed-sequence framing; do not cite it for escalator terms.) - The balance sheet the pledge would draw on: taxable brokerage $487,381 across four accounts at two custodians (Wealthfront ~$447,506 including a $109,640 Nasdaq-100 direct-indexing sleeve; Schwab $39,875), plus $290,150 cash, per [[2026-08-01-capital-deployment-brainstorm]]. Deployable liquid ex-retirement is $777,531. The founder is "not capital-constrained… cash-flow constrained and demand-constrained."
- Vault white space confirmed. A full-vault grep for pledged asset / securities-backed / SBLOC / asset depletion / portfolio line of credit returns exactly one hit: the parent brief's own follow-up paragraph. Nothing here duplicates existing vault coverage. Non-housing debt is ~$1,516/mo (auto $362 + an unitemized $1,155 "Loan Repayment" open since [[2026-06-04-home-affordability-build-vs-buy]]).
What the web says
Bonus and variable income seasoning (question 1). Fannie Mae B3-3.3-02, effective 03/04/2026, is unambiguous: "A minimum two-year history is recommended; however, income received for a shorter period, but no less than 12 months, may be considered as acceptable if there are positive factors to reasonably offset the shorter income history." Critically, the calculation itself is hard-floored: "The calculation must include a minimum of 12 months' income." The "positive offsetting factors" carve-out moves you from 24 months to 12, not below 12. Freddie Mac's parallel rule at 5303.4 sets the same 12-month floor. There is no agency path to counting bonus income at 9 months of tenure.
Base income has no history requirement at all (question 2, the crux). Fannie B3-3.3-01, same 03/04/2026 revision, states that for fixed base income "A minimum history is not required for inclusion as qualifying income." Fixed base is "a set salary or fixed hourly rate with guaranteed minimum hours." Only variable base income (fluctuating hours) carries the 12-month floor. A contractual certification escalator that permanently raises annual salary is fixed base income the moment it is effective, and it seasons in zero months.
Forward-dated raises are also usable before they hit a paystub. Fannie B3-3.2-02 permits future increases from the current employer when: the increase is "fully verified with the employer using Form 1005… or other documentation directly from the employer that provides the terms of the future pay increase"; "The increase must take effect no later than 60 days after the note date"; "The income is fixed base only"; the transaction is a purchase or limited cash-out; and the borrower is not employed by a family member or interested party. Freddie is more generous. Section 5303.2 Option 1 covers a salary increase from the current employer commencing "no later than 90 days after the Note Date", requiring documentation that "the increase is fully approved and is explicitly granted", a 10-day pre-closing verification, and extra reserves equal to housing plus liabilities times the months to the start date, plus one. (Note: Freddie's live guide is JavaScript-rendered and would not fetch; 5303.2 was read through the homebuyer.com guideline mirror and should be re-verified against the primary guide before anyone relies on the 90-day figure.)
Offer letters are for new jobs, not raises. Fannie B3-3.3-03 permits qualifying on a "fully executed and non-contingent offer or contract for future employment" with a start date "no earlier than 30 days prior to the note date OR no later than 90 days after the note date" and either six months' PITIA in reserves or a bridge-to-start-plus-one reserve. But the section is scoped to borrowers "scheduled to begin new employment." It does not reach a raise at an existing employer. That path runs through B3-3.2-02 instead.
Borrowed funds secured by your own securities are DTI-neutral under agency rules (question 3). Fannie B3-4.3-15 is the load-bearing citation and it is long-standing (effective 10/30/2009, unrevised): "Borrowed funds secured by an asset are an acceptable source of funds for the down payment, closing costs, and reserves, since borrowed funds secured by an asset represent a return of equity." Eligible collateral explicitly includes "stocks, bonds, and 401(k) accounts." The DTI rule: the general case is "the lender must consider monthly payments for secured loans as a debt," but there is an express carve-out, "When loans are secured by the borrower's financial assets, monthly payments for the loan do not have to be considered as long-term debt." The offsetting cost is reserves: "If the borrower uses the same financial asset as part of their financial reserves, the lender must reduce the value of the asset by the amount of proceeds and related fees for the secured loan."
SBLOCs are demand loans, and that is the whole risk. Per the FINRA/SEC OIEA investor alert on securities-backed lines of credit, a decline in collateral value triggers a "maintenance call" requiring additional collateral or repayment "within a specified period (typically two or three days)," and if the borrower cannot meet it "the firm can sell your securities and keep the cash." SBLOCs are callable by the lender at any time irrespective of payment history. Advance rates on diversified equity portfolios generally run 50-70% at full-service custodians and materially lower at robo-advisors; I could not fetch Schwab's Pledged Asset Line disclosure (403) or verify Wealthfront's Portfolio Line of Credit advance rate in this run, so treat those percentages as unverified.
Pledged-asset mortgages are a different product and they point the wrong way. A PAM substitutes pledged securities for the down payment, eliminating PMI and avoiding a capital-gains realization while keeping the portfolio invested. A targeted search for a 2026 PAM offered on jumbo construction-to-perm found no lender program documentation, only general product explainers and unrelated SEC filings. The arithmetic matters more than the availability: a PAM finances a larger principal, which raises the monthly payment, which shrinks the DTI-constrained envelope.
Convergences and contradictions
- Contradiction with the parent brief, and it is the useful one. [[2026-08-14-construction-to-permanent-loans-florida]] warns that "the vault warns most lenders want a 2-year history for variable comp (bonus, cert escalators)." The agency guides split that sentence in half. Bonus, yes: 12-month hard floor, 24 preferred. Cert escalators, no: they are fixed base income under B3-3.3-01 with no minimum history, and are additionally coverable under B3-3.2-02 / Freddie 5303.2 even before they take effect. The founder has been carrying a pessimism on the escalators that the guides do not support.
- Convergence on where the ceiling actually binds. The vault says tenure is the risk; the guides say tenure binds on the bonus line only. Both point at the same conclusion from opposite directions: the ~$25,625 bonus is the single largest recoverable income item, and it is recoverable by calendar rather than by negotiation.
- Contradiction between agency rules and portfolio practice, unresolved. B3-4.3-15's exclusion of securities-secured loan payments from DTI is an agency rule. This loan is jumbo and portfolio by construction (Hillsborough's 2026 conforming limit is ~$832,750). Portfolio lenders may adopt the agency treatment, tighten it, or refuse borrowed funds for the equity contribution on a construction loan entirely. Construction lenders in particular tend to want unencumbered borrower equity. Calibration: treat "SBLOC payment excluded from DTI" as the agency baseline and a plausible ask, not as the expected outcome.
Synthesis for RDCO
The escalators are good news and small news. Question 2 resolves cleanly in the founder's favor: a contractual base-salary increase tied to a completed certification is fixed base income the moment it is effective, seasons in zero months, and is usable at a 9-month-tenure application without any exception request. Escalator #2 is expected around late Nov 2026 and would already be on paystubs by a Q1 2027 application; even if it slipped, Fannie's 60-day and Freddie's 90-day forward-raise windows cover it with a written phData confirmation of the approved increase. The one thing that would break this is a documentation shape mismatch. If phData papers the escalator as a recurring bonus rather than an amended base salary in the comp letter, it flips from B3-3.3-01 to B3-3.3-02 and dies on the 12-month floor. That is the single highest-leverage document to get right, and it costs nothing. The value, though, is modest: $10,000 of annual base is $833/mo of gross, which at 43% back-end supports roughly $54,000 of additional loan at 7%. Against a gap measured in hundreds of thousands, that is about 13%.
The bonus is the real income lever, and it is a calendar problem, not an underwriting problem. At a 12.5% target on a $205k base, the bonus is ~$25,625/yr, worth roughly $140,000 of qualifying loan once it can be counted. Fannie B3-3.3-02's requirement that "the calculation must include a minimum of 12 months' income" is not a soft preference a portfolio lender routinely waives at high loan amounts with thin tenure; it is the line that separates a saleable file from an exception file, and exceptions at $1.3M+ get priced or declined. With a 2026-05-26 W-2 start, the 12-month mark is late May 2027, and a full bonus cycle needs to have actually paid and shown on paystubs or a W-2. This produces the cleanest recommendation in the brief: applying in mid-2027 rather than Q1 2027 is worth more than both escalators combined and costs nothing but calendar. It also compounds with the parent brief's single-close recommendation, since a single-close means one underwrite and you want that underwrite to happen on the strong side of the seasoning line.
Pledged assets solve the tax problem and expand the qualifying envelope by exactly zero. This is the correction the parent brief's follow-up implicitly invited. Ranked by effect on DTI, a cash down payment is best (smallest loan), an SBLOC-funded down payment is roughly neutral under B3-4.3-15's financial-asset carve-out and strictly better than a HELOC or personal loan whose payments do count, and a pledged-asset mortgage is worst because it finances more principal. None of the three adds a dollar of qualifying income. What an SBLOC genuinely buys is avoidance of a capital-gains realization against $487,381 of taxable brokerage, and it buys that at three real costs: B3-4.3-15 requires reducing the reserve credit for that same asset dollar-for-dollar by the proceeds, so the pledge cannibalizes the reserve the construction lender is separately demanding (TD holds 10% of construction cost, roughly $160-180k on this build); the founder has $290,150 in cash already, so the marginal gains-avoidance benefit applies only to the slice beyond what cash covers; and the risk pairing is bad on its own terms. An SBLOC is a demand loan with a two-to-three-day maintenance call, secured against a portfolio that includes a $109,640 direct-indexed sleeve, funding a 12-18 month construction project. A drawdown mid-build is the one moment you cannot sell the collateral asset (a half-finished house) and cannot pause the draw schedule. Recommendation: pledged-asset structures are a tax-efficiency tool worth pricing for the marginal slice above cash, not a qualifying-envelope tool, and the margin-call exposure should be sized so a 30% portfolio drawdown does not force liquidation mid-construction.
The arithmetic, and it does not close. Holding rate at 7.00%/30-yr (the parent brief's stress rate; MIDFLORIDA's posted 7.000% was on a $300,000 example and a jumbo prices above it), non-housing debt at $1,516/mo, and escrow at a central estimate of $4,200/mo on a ~$2.5M Zone AE Hillsborough property:
| # | Income counted | Monthly gross | DTI | P&I budget | Qualifying loan |
|---|---|---|---|---|---|
| S1 | Base $200k + spouse $125k (provable today) | $27,083 | 43% | $5,930 | ~$891k |
| S2 | + escalator #2 (base $205k) | $27,500 | 43% | $6,109 | ~$918k |
| S3 | same, 45% jumbo max | $27,500 | 45% | $6,659 | ~$1,001k |
| S4 | + bonus $25,625 counted (needs ~June 2027) | $29,635 | 45% | $7,620 | ~$1,145k |
| S5 | same, 50% portfolio overlay (rarely granted) | $29,635 | 50% | $9,102 | ~$1,368k |
Loan required, reconciled from the parent's sources-and-uses on current balances ($415,622 payoff + build + $108k temp housing, against $777,531 deployable plus $1.37M at $1.8M**, both with zero reserve left standing. Three things fall out. First, at today's provable income the $1.5M build is short by roughly $180k, which is a harder result than the parent brief's "it closes with zero margin" and comes entirely from scaling escrow to property value instead of loan size. Second, the $1.8M build needs $1.37M and the absolute maximum stack produces $1.368M, meaning the best case lands exactly on the zero-reserve line; any reserve preservation makes it infeasible, and S5's 50% overlay is an exception-priced ask at sub-12-month tenure that should not be planned on. Third, the whole table is hostage to one number I did not resolve: ±$1,000/mo of escrow swings every row by ~$150k of loan. That makes the parent brief's already-queued escrow follow-up the highest-value open item in this project, ahead of anything in this brief. For scale on the other levers at 43%: each 2 points of DTI is worth ~$83k, each 25bp of rate is worth ~$29k at this loan size, and a ~75bp rate improvement is worth roughly what the bonus is.$175k new savings): **$1.07M at a $1.5M build** and **
Why this is in the vault
This resolves two of the five follow-ups from [[2026-08-14-construction-to-permanent-loans-florida]] and changes the home-rebuild-2027 decision in three specific ways: it retires the "cert escalators need 2-year history" belief carried in [[readme]] Open Question #3, it converts the $1.8M build from "financeable if levers stack" to "not financeable with a reserve intact at 7%," and it identifies application timing (mid-2027 over Q1 2027) as a larger lever than any product or negotiation, which is a milestone-schedule input, not a financing input.
Open follow-ups
- Can the founder's prior 1099 income through Ray Data LLC (Mammoth Growth, same line of work) be bridged to satisfy variable-income history at phData? Fannie B3-3.1-01 contains line-of-work continuity provisions, and a 1099-to-W-2 transition within the same field is a documented pattern. If a lender will count the prior self-employed history toward the 12-month variable-income floor, the ~$140k bonus lever unlocks at a Q1 2027 application instead of mid-2027, which is the difference between the $1.5M build penciling and not. Nothing in the vault or in this brief answers it.
- Does 12-18 months of temporary-housing rent count as a monthly obligation in DTI on a single-close construction-to-perm where the subject property is the borrower's future primary residence? At the repriced ~$8,000/mo temp-housing figure this is not a rounding error; if it counts, every row of the table above collapses. Agency construction-to-perm guidance (Fannie B5-3.1) addresses the treatment of the borrower's current housing expense during the construction period, so this is researchable from the guides rather than lender-dependent.
Related
- [[2026-08-14-construction-to-permanent-loans-florida]] — parent brief; mechanics, Florida lender list, sources-and-uses
- [[2026-08-04-flood-zone-ae-construction-loan-requirements]] — Zone AE, BFE 11.0 ft, flood-insurance timing
- [[2026-07-22-florida-ae-zone-insurance-repricing-tampa]] — the insurance half of the escrow line
- [[cert-progress]] — primary source for the two-leg $5k escalator structure and the 2026-08-17 CCA-F pass
- [[2026-08-01-capital-deployment-brainstorm]] — verified balances; the $487,381 taxable brokerage a pledge would draw on
- [[2026-06-04-home-affordability-build-vs-buy]] — the $1,516/mo non-housing debt line
- [[readme]] — home-rebuild-2027 project baseline; source of the stale $315k and the inherited $1.0-1.2M ceiling
A sibling brief on asset-depletion qualification for the founder's parents' purchase (README Critical Open Question #1, a $450k mortgage on retirement income) was queued for the same night as this one. It covers a different borrower and a different qualification method; asset depletion is deliberately not treated here. Cross-link once its slug exists.
Sources
Vault
06-reference/research/2026-08-14-construction-to-permanent-loans-florida.md06-reference/research/2026-08-04-flood-zone-ae-construction-loan-requirements.md06-reference/research/2026-07-22-florida-ae-zone-insurance-repricing-tampa.md01-projects/phdata/cert-progress.md01-projects/home-rebuild-2027/README.md04-finance/2026-08-01-capital-deployment-brainstorm.md04-finance/2026-06-04-home-affordability-build-vs-buy.md
Web — primary underwriting guides (fetched and read)
- Fannie Mae Selling Guide B3-3.3-02, Bonus, Commission, Overtime, and Tip Income (eff. 03/04/2026) — https://selling-guide.fanniemae.com/sel/b3-3.3-02/bonus-commission-overtime-and-tip-income
- Fannie Mae Selling Guide B3-3.3-01, Base Income (eff. 03/04/2026) — https://selling-guide.fanniemae.com/sel/b3-3.3-01/base-income
- Fannie Mae Selling Guide B3-3.2-02, Standards for Employment-Related Income (eff. 03/04/2026) — https://selling-guide.fanniemae.com/sel/b3-3.2-02/standards-employment-related-income
- Fannie Mae Selling Guide B3-3.3-03, Employment Offers or Contracts (eff. 03/04/2026) — https://selling-guide.fanniemae.com/sel/b3-3.3-03/employment-offers-or-contracts
- Fannie Mae Selling Guide B3-4.3-15, Borrowed Funds Secured by an Asset (eff. 10/30/2009) — https://selling-guide.fanniemae.com/sel/b3-4.3-15/borrowed-funds-secured-asset
Web — secondary, flagged
- Freddie Mac Guide 5303.2, Income commencing after the Note Date — read via the homebuyer.com guideline mirror (https://homebuyer.com/guidelines/freddie-mac/income-commencing-after-the-note-date-5303-2) because the primary guide at https://guide.freddiemac.com/app/guide/section/5303.2 is JavaScript-rendered and returned no body. The 90-day window and reserve formula should be re-verified against the primary guide. Freddie 5303.4 (https://guide.freddiemac.com/app/guide/section/5303.4) failed to render for the same reason; its 12-month bonus floor is taken from search-result summaries only.
- FINRA / SEC OIEA Investor Alert, Securities-Backed Lines of Credit — https://www.investor.gov/introduction-investing/general-resources/news-alerts/alerts-bulletins/investor-alerts/investor-28 and https://www.finra.org/investors/insights/securities-backed-lines-credit (read via search summaries; the direct FINRA alert URL 404'd)
Not resolved in this run (explicit)
- Schwab Pledged Asset Line disclosure (403 on fetch) and Wealthfront Portfolio Line of Credit terms. Advance rates of 50-70% are a general-practice band, not verified against either custodian's documentation.
- No Florida portfolio lender publishes its jumbo income or pledged-asset overlays. Every "portfolio lenders may differ" statement here is a structural inference from the agency baseline, not a documented overlay.
- The escrow line on a ~$2.5M Zone AE Hillsborough property. The $4,200/mo central estimate drives the entire table and is the parent brief's still-open follow-up #1.
Research caps: 5 WebSearch + 8 WebFetch against a template cap of 3 + 3. Over-budget by design and disclosed: this is a merged brief answering two queued follow-ups, and the dispatch required agency-guide and lender-program citations rather than blog summaries. Vault retrieval was delegated to one sub-agent to keep a 21KB parent brief and seven vault docs out of parent context.