SBA Administrator Loeffler Signals $10M+ Cap Expansion + Underwriting Rollback — What The FY2026 SBA Policy Shift Means For Your Funding Strategy

Introduction

If you've spent any time in a merchant cash advance funnel, you already know the pitch: fast money, no real underwriting, sign here. That world sells urgency because urgency is profitable for the lender, not because it's what's actually good for your business. SBA financing has never been that world, and the policy signals coming out of Washington in the back half of July 2026 make clear it's about to be even less that world — a good thing, not a warning sign, for the business owner who actually wants a durable capital stack rather than a payday.

On July 27, 2026, SBA Administrator Kelly Loeffler told Forbes something worth sitting with: she "absolutely" supports raising the $5 million cap on 7(a) loans, a ceiling that hasn't moved since 2010 (Forbes). In the same conversation, she explained — without apology — why FY2026 loan approvals are running roughly a third below last year's pace: SBA is deliberately unwinding underwriting latitude that let "loans get approved that shouldn't have," in her words, and is now managing "with an eye toward risk management, not headline numbers" (Forbes).

Read those two data points side by side and the shape of the moment becomes obvious: the ceiling is going up and the floor is going up at the same time. That is not a contradiction. It is, if anything, the most coherent version of what a well-run federal lending program is supposed to look like — expand capacity for the borrowers who've earned it, while tightening the standards that determine who gets in the door. For years, capital access and underwriting discipline have been treated as a zero-sum tradeoff in the public conversation around SBA lending. Loeffler's July signals argue they aren't. You can raise the maximum loan size available to a qualified manufacturer and simultaneously require a real cash equity injection from an acquisition buyer. Both moves come from the same underlying philosophy: match the size of the loan, and the ease of getting it, to the actual creditworthiness of the borrower.

This matters enormously for how you should think about your own funding strategy right now, and it's the reason this two-part analysis exists. Too many business owners approach SBA financing as a single event — a loan they either get or don't, on a timeline dictated by whenever they happen to need the money. That's backwards, and it's precisely the mindset that gets punished by a tightening underwriting environment. We think about it differently at Stacking Capital, and it's worth stating plainly here because it frames everything that follows: funding is for today. Becoming bankable is a repetitive process. The loan you close is a single transaction. The bankability that made that loan possible — the compliance data, the trade lines, the banking relationships, the financials — is not a one-time achievement. It's a discipline you maintain, quarter after quarter, whether or not you're actively raising capital in that moment.

We organize that discipline around what we call the Four Legs of Bankability: compliance (the boring, correct, consistent data across your Secretary of State filing, your EIN registration, and your business bureau listings), business credit (the trade lines, the scores, the payment history that builds a file independent of your personal credit), banking relationships (real depository history with the institutions that will eventually underwrite you), and financials (a full, clean, defensible set of books that can survive the scrutiny SOP 50 10 8 now demands). A business that's strong on three of the four legs but weak on the fourth is not a bankable business — it's a business with a gap that an underwriter will find, because underwriters are now, by explicit SBA policy, looking harder than they were 18 months ago.

Part 1 of this analysis — what you're reading now — walks through exactly what Loeffler said, on the record, and reconciles it against the data trail SBA itself has published. We cover the FY2026 approval numbers with full transparency about where different sources diverge, the shutdown's cost to the lending pipeline (also reconciled across sources that don't fully agree with each other), and the mechanics of the underwriting rollback that's driving the current approval slowdown. Part 2 will pick up with the cap-raise legislative mechanics, the manufacturer segment specifics, the Preferred Lender Program landscape, and the concrete tactical moves that follow from all of it. Nothing here is designed to make you nervous about SBA financing. It's designed to make you accurate about it — because an accurate read on where the program is headed is worth more to your capital stack than any amount of urgency-driven marketing copy from a lender who profits when you don't read the fine print.

One more framing note before we get into the sourcing: this article deliberately treats every figure Loeffler has cited publicly with the same standard of scrutiny we'd apply to any other primary source, which means naming the places where her numbers don't perfectly reconcile with SBA's own published data rather than smoothing over the gap for narrative convenience. That is not an attempt to undermine the substance of what she's signaling. If anything, it's the opposite — an administrator who is directionally correct about a genuine, verifiable tightening cycle doesn't need every supporting statistic to be presented as flawless in order for the underlying policy shift to be real and worth planning around. We'd rather hand you the reconciled numbers and let the strength of the actual trend speak for itself than hand you a single clean statistic that quietly ignores the parts that don't line up. That's the standard this article follows section by section.

Section 1 — What Loeffler Actually Said

Policy analysis is only as good as its sourcing, so before we interpret anything, it's worth establishing precisely what SBA's Administrator said, when, and to whom. The core interview is Forbes contributor John Schroyer's July 27, 2026 piece, headlined "SBA Chief Loeffler: It's Time To Raise The $5 Million SBA Lending Cap," conducted via direct Q&A with follow-up email exchanges (Forbes). It is the single most important primary source underpinning this entire analysis, and it did not arrive in a vacuum — it followed two weeks of Congressional testimony that laid the groundwork for nearly everything she told Schroyer.

The cap-raise question, verbatim

Asked directly whether she supports raising the $5 million 7(a) cap, Loeffler didn't hedge. Her answer, in full: "I absolutely would. And we've been very vocal about returning to the mission of the small business administration. Our loan limits stopped going up in 2010. So 16 years ago was the last increase" (Forbes). That 16-year framing recurs across her public statements almost verbatim — the Coleman Report's recap of her Congressional testimony from the week prior quotes an essentially identical line: "Our loan limits in the SBA have not been raised since 2010. So 16 years ago, it was set at $5 million" (Coleman Report). When an Administrator repeats the same specific number, in the same framing, across multiple venues over the same two-week window, that's a strong signal the figure is a deliberate talking point rather than an offhand remark — which tells you the administration intends to keep leaning on the "16-year gap" argument as its primary public justification going forward.

She grounded the cap-raise case in a concrete, sector-specific example rather than an abstract inflation argument. Discussing why $5 million no longer covers what it used to for capital-intensive borrowers, she pointed to manufacturing: "98% of all of America's factories are small businesses," and outfitting one with modern equipment — "robotics and automation and machining... CNC machines and software and precision optics tooling" — now costs "much more than $5 million" (Forbes). That 98% figure appears with a slightly different rounding in her Senate testimony — 99% of U.S. manufacturers qualify as "small manufacturers" under the relevant statutory definition — a trivial discrepancy that's worth noting only because it establishes a pattern we'll return to throughout this article: Loeffler's numbers are directionally consistent across venues but not always precisely identical, and a rigorous analysis owes readers that distinction rather than pretending every figure she cites is a single, unambiguous number.

The legislative vehicle: MAMFA

Loeffler named a specific mechanism for the cap increase rather than speaking only in aspirational terms: the Made in America Manufacturing Finance Act (MAMFA). Her description of its status: it "passed on a bipartisan basis out of committee in both the House and the Senate, and it's passed from the House floor," with the administration "hoping for passage as part of defense funding" (Forbes). That description checks out against the legislative record with one important caveat we'll flag transparently rather than gloss over: H.R.3174, the House companion, did pass the House floor unanimously and on a bipartisan basis (SBA.gov). Its Senate companion, S.1555, reached committee hearings but this desk could not independently confirm a full Senate floor vote as of publication — meaning the bill is further along than a typical piece of SBA-adjacent legislation, but not yet law. We cover the full legislative mechanics, including the defense-authorization pathway Loeffler referenced, in Part 2 of this analysis.

It's worth being precise about what MAMFA actually is, because the distinction matters for how you should read the "cap raise" conversation generally. MAMFA is a targeted carve-out for small manufacturers — businesses in NAICS sectors 31 through 33 with all production facilities located in the United States — not a general-industry increase to the 7(a) cap for every type of business. Loeffler's personal support for raising the cap more broadly ("I absolutely would") is a separate, larger-scope policy position from the specific bill currently moving through Congress. Conflating the two would overstate how close a universal $10 million 7(a) cap actually is; this article treats them as related but distinct developments throughout, and we'll return to that distinction with the full legislative mechanics in Part 2.

The July 4 decoupling, referenced directly

Loeffler's cap-raise comments came against the backdrop of a change that had already taken effect three weeks before the interview: on July 4, 2026, SBA decoupled the 7(a) and 504 loan programs so that a borrower's balance in one no longer counts against their eligibility in the other, enabling qualified borrowers to access up to $10 million in combined capital (SBA.gov). This is a meaningfully different mechanism from what MAMFA would do. The individual per-program caps didn't move — 7(a) is still capped at $5 million per loan, and 504 is still capped at roughly $5 to $5.5 million per project. What changed is that the two no longer compete against each other for the same borrower's aggregate exposure. A qualified manufacturer, for instance, can now hold a full $5 million 7(a) balance and a full $5 million 504 balance simultaneously — a real, immediately usable $10 million combined ceiling that required no act of Congress, because SBA executed it under its own regulatory authority via Policy Notice 5000-879058. We'll cover the full stacking mechanics, including the separate $3.75 million cap on SBA's actual guaranteed exposure per borrower, in more detail in Part 2 — but it's worth flagging here because Loeffler referenced it directly as evidence the administration is already moving on capital access even before Congress acts on MAMFA.

The zero-subsidy caveat she attached to every expansion claim

What's notable — and what separates this from a simple "we want to lend more money" pitch — is how consistently Loeffler paired every expansion signal with a risk-management qualifier. On the manufacturer segment specifically, she said: "We've seen manufacturers by the hundreds hit the cap at that $5 million. And certainly we're testing the default rates at every size. We're modeling what that looks like to go forward. So we would just want to make sure that it comports with our commitment to be cost-neutral to taxpayers. That means the loan programs should operate at zero subsidy by taxpayers" (Forbes). That "zero subsidy" phrase is not marketing language — it's a legal standard. The 7(a) program is statutorily required to cover its own expected losses through the fees it collects, with no ongoing federal appropriation subsidizing it (House Small Business Committee). Every cap-raise conversation inside SBA is therefore constrained by an actuarial question first and a political question second: can the fee revenue collected on larger loans plausibly cover the expected losses on larger loans? We unpack the mechanics of that math, and why it's a live open question at the $10 million tier specifically, later in this analysis.

The Coleman Report's recap of her Congressional testimony from the week before the Forbes interview adds detail that sharpens this picture considerably (Coleman Report). Per that account, roughly 200 loans have hit the existing $5 million ceiling "in the last couple of years" — a more precise figure than the Forbes interview's looser "by the hundreds" phrasing, and one worth citing specifically when discussing the manufacturer segment. The same testimony reportedly included a claim that default ratios in SBA's smallest loan programs have risen sharply, and that "nearly 500 lenders" have exited SBA lending programs entirely over concerns about underwriting discipline. We flag that default-rate figure as one this research could not fully reconcile against SBA's own independently published loan-level data — a distinction we address plainly in Section 4 rather than repeating an unreconciled number as settled fact.

Why this matters beyond the headline

Put the pieces together and a consistent policy philosophy emerges across every venue Loeffler has spoken in over the past several weeks — the Forbes interview, the House Appropriations Subcommittee hearing on July 14, and the Senate Small Business Committee's MAMFA hearing the same week. She is on record telling the House Appropriations Subcommittee, in her opening testimony: "Program integrity was absent from our lending programs where underwriting standards were dismantled, contributing to an estimated $2.2 billion in losses that taxpayers will bear" — a direct reference to the pre-2025 underwriting regime (Forbes Breaking News/YouTube). In her formal Senate testimony on MAMFA specifically, she stated: "We also took urgent action to restore the financial integrity of our core loan programs by reinstating lender fees and restoring underwriting standards — eliminating the 'Do What You Do' criteria that put taxpayers on the hook for billions of dollars" (Senate Small Business and Entrepreneurship Committee testimony, submitted the week of July 14–21, 2026). That same testimony credited the current administration's first hundred days with an 80% increase in SBA loan approvals compared to the equivalent period a year earlier, with loans to businesses of five or fewer employees "nearly doubling" — figures that, taken at face value, sit in some tension with the FY2026 approval decline discussed later in this article, and which we address directly in that reconciliation rather than letting the apparent contradiction stand unaddressed.

The throughline across all of it: expand access at the top of the ceiling, tighten discipline at the point of underwriting, and justify every move against a cost-neutrality standard that Congress itself wrote into the statute. That is not the posture of an agency retreating from small business lending. It's the posture of an agency trying to make the program durable enough to expand responsibly — which is a meaningfully different story than either "SBA is cutting off small businesses" or "SBA is loosening the reins," and it's the story this analysis is built to tell accurately.

Section 2 — FY2026 Approval Data, Reconciled

The single most-quoted statistic to come out of Loeffler's Forbes interview is the FY2026 approval decline. Asked directly about the drop, she cited a decline from "just over 63,000" 7(a) loans approved in FY2025 to "just over 43,000" so far in FY2026 — a decline of roughly 32% (Forbes). That number has circulated widely since the interview ran, and it's directionally correct — but it does not cleanly match SBA's own record-setting full-year FY2025 total, and a rigorous analysis owes you that discrepancy rather than repeating the headline figure as though it were the complete picture.

The reconciliation, laid out plainly

Here is what the data actually shows across every source this desk could independently verify:

Sources: Forbes; Lumos SBA lending report; LenderHawk; Manufacturing Dive.

Notice the gap: Loeffler's "63,000 in FY2025" does not match SBA's own record-year full-fiscal-year total of roughly 77,600 to 78,000 loans. If you compared FY2026's 43,000 against the full 78,000-loan FY2025 record, you'd calculate something closer to a 45% decline — a materially scarier number than the 32% Loeffler herself cited. So which is right, and why the gap?

The most defensible explanation, based on cross-referencing Lumos Data's independently compiled figures, is that Loeffler is comparing a same-period, year-to-date window rather than full fiscal years. Lumos's own nine-month comparison (October through June, spanning both fiscal years on an equivalent calendar basis) shows 61,270 loans in the FY2025 nine-month window against 40,824 in the equivalent FY2026 window — a decline of 33.4% by count and 20.9% by dollar volume (Lumos SBA lending report). That 61,270 figure sits very close to Loeffler's "just over 63,000," and the resulting decline percentage — right around a third — matches her "roughly 32%" framing almost exactly. Independently, Senator Ed Markey, ranking member of the Senate Small Business Committee, cited a separate figure of 7(a) lending having declined 32% under the administration's lending restrictions, a number that corroborates the same-period comparison rather than the full-year comparison (Legis1).

Why FY2025 was an outlier, not a normal baseline

There's a second layer to this reconciliation that matters just as much as the raw numbers: FY2025 was not a typical year to measure against. Per Lumos's analysis, nine-month dollar volume in FY2026 — $21.8 billion — actually falls right on the trajectory the 7(a) program has followed for the better part of a decade, up roughly 110% since 2010 on a nine-month basis. The genuine outlier is FY2025's $27.6 billion, a spike year (Lumos SBA lending report). Why the spike? Loeffler's own explanation, from the Forbes interview, is instructive here: much of FY2025's volume reflected borrowers "rushing to get loans in under that lax underwriting standard" before SOP 50 10 8 took effect. In other words, part of what looks like a FY2026 "decline" is really FY2025 borrowing activity that got pulled forward, ahead of a known tightening. Comparing FY2026 against that specific pulled-forward peak year overstates how much of a contraction is actually happening relative to the program's longer-run trend line.

Lumos's own attribution analysis, developed independently of Loeffler's public statements, arrives at essentially the same three-factor explanation she gave Forbes: "Three factors: a 43-day government shutdown that froze new approvals in October and November 2025, tighter underwriting under SOP 50 10 8 (effective June 2025), and a comparison against an inflated FY2025 base that had pulled volume forward" (Lumos SBA lending report). When an administration official's public explanation and an independent data firm's analytical conclusion converge this closely without any apparent coordination, that's a meaningful corroboration signal worth naming explicitly.

Quarterly granularity: where the decline actually concentrates

Breaking the fiscal year into quarters clarifies the story further. Q1 FY2026 (October through December 2025) was distorted heavily by the shutdown — zero approvals in all of October, followed by a partial snapback: November logged 5,227 loans worth $2.46 billion, down only 12.8% year-over-year by count, and December logged 5,125 loans worth $2.78 billion, down 27.6% by count. Q2 FY2026 (January through March 2026) approvals fell to 18,400 loans — the lowest quarterly count since 2021 — with sub-$150,000 approvals down 22% year-over-year, while loans above $1 million were down only 6% year-over-year.

Sources: Lumos SBA lending report analysis; SBA current-month activity data via SBA.gov.

That pattern — a much sharper pullback at the small-dollar end of the portfolio than at the large-dollar end — tells you something important about where the tightening is actually concentrated, and we'll return to this same size-band pattern with more granularity in Section 4, because it directly informs which borrowers should expect the most friction and which should expect comparatively little change in their approval experience. It also connects directly to the average loan size trend: as smaller loans retreat by a larger percentage than bigger ones, the portfolio's average loan size mechanically rises even in a year where total loan count is falling — a nuance that matters if you're benchmarking your own deal size against "typical" SBA loan data without accounting for this shift.

Section 3 — Government Shutdown Impact: Two Numbers, Reconciled

The second major data-reconciliation question this article needs to address head-on involves the government shutdown that ran through much of October and November 2025. Loeffler told Forbes that one shutdown — 43 days long — "probably cost us $2 billion to $2.5 billion dollars in lending opportunity" (Forbes). That's a real, quotable figure from the person running the agency. But it is not the number SBA itself published as the final, official reconciliation of that same shutdown's impact — and the gap between the two is large enough that it deserves direct treatment rather than a passing mention.

The full trail of figures, in chronological order

Shutdown-impact estimates evolved considerably as the 43-day closure dragged on, which is itself useful context — these were live, evolving estimates made in real time, not a single static number that someone simply misquoted later. Here's the complete trail:

Sources: ExecutiveGov; Washington Times; Curinos; SBA.gov.

The authoritative, final number is SBA's own November 13, 2025 press release, published once the shutdown ended and the agency could tally the actual backlog: the agency estimated it was unable to deliver $5.3 billion to 10,000 small businesses during the 43-day closure, a pace of roughly 320 small businesses per day and about $170 million per day (SBA.gov). That figure is roughly double Loeffler's own Forbes estimate of $2 billion to $2.5 billion. We're not going to paper over that gap or quietly pick whichever number sounds better — both numbers are on the record from the same agency, and the honest read is that Loeffler was likely speaking from memory in a Q&A format months after the fact, or possibly referencing the 7(a) program specifically rather than the combined 7(a)-plus-504 impact that SBA's official release covers. Either way, if you're citing shutdown impact figures elsewhere, the SBA's own $5.3 billion/10,000-business figure is the more authoritative, final, and precisely documented number — corroborated independently by Manufacturing Dive's coverage of the same reopening period, which also confirmed the shutdown froze the launch of the new MARC manufacturer working-capital program (Manufacturing Dive).

Two shutdowns, one 43-day closure that matters most

It's worth being precise about scope here too: the 43-day figure refers to a single continuous shutdown period, not a cumulative total across multiple separate closures within the fiscal year. That 43-day window is the one both Loeffler and SBA's official release reference when citing their respective dollar figures, so the two numbers are at least measuring the same event — they simply arrive at different totals for it. In broader economic terms, Loeffler separately cited a projected 40,000 jobs lost and up to $100 billion in total economic impact nationally from the shutdown overall (not SBA-lending-specific), consistent with separate U.S. Chamber of Commerce commentary citing a Congressional Budget Office estimate of a 1.5-point hit to Q4 GDP and $18 billion in delayed federal contract payments to 6,500 small businesses on top of the SBA lending freeze specifically.

Impact by product line: 7(a) hit hardest, 504 less, Express least

The shutdown's impact was not evenly distributed across SBA's loan products, and understanding that unevenness matters if you're deciding which SBA product to pursue in an environment where future shutdown risk is a real, if intermittent, variable. The 7(a) program — SBA's largest and most general-purpose lending vehicle — bore the brunt of the freeze, since new 7(a) loan numbers simply could not be issued while the agency's core loan-processing functions were shuttered. The 504 program, run in partnership with Certified Development Companies rather than directly through SBA loan officers, experienced disruption but to a comparatively lesser degree, since CDCs could continue portions of their own underwriting and packaging work even while SBA-side final approvals stalled. SBA Express, with its much smaller loan ceiling and higher degree of lender delegation, was the least affected of the three — delegated lenders retained more independent approval authority throughout, meaning Express-eligible borrowers experienced comparatively less disruption than those pursuing standard 7(a) or 504 financing during the closure window.

This desk's July 27 coverage of the SBA Office of Advocacy's business-formation report included a practical 30-60-90 day checklist for businesses navigating funding timing decisions in this environment — worth cross-referencing if you're trying to sequence your own application around policy and macro uncertainty rather than reading this shutdown-impact data in isolation. The throughline between both pieces is the same: SBA lending timing risk is real and worth planning around, but it is a planning variable, not a reason to abandon the program as a funding source.

Post-reopening, SBA reported it was processing around $500 million in loan value per day to work down the accumulated backlog — a recovery pace that, if sustained, would clear a $5.3 billion backlog in roughly ten to eleven business days, though real-world recovery curves rarely move in a perfectly straight line once new, non-backlog applications resume competing for the same processing capacity.

Section 4 — The Underwriting Rollback: "Do What You Do" and What Replaced It

Everything discussed so far — the approval decline, the shutdown's compounding effect, the cap-raise conversation itself — sits downstream of a single underwriting policy shift that is, in our view, the most consequential and least understood part of this entire story for the average business owner. Understanding it precisely is the difference between reading the FY2026 numbers as a warning sign and reading them as exactly what they are: the predictable, intended result of a deliberate standards reset.

What "Do What You Do" actually was

"Do What You Do" was introduced via SOP 50 10 7, effective August 1, 2023, and refined in SOP 50 10 7.1, effective November 15, 2023, under the prior administration. Its core principle was permissive rather than prescriptive: SBA gave Preferred Lenders latitude to apply their own internal commercial underwriting standards wherever SBA's own guidance was ambiguous, rather than holding every lender to SBA-mandated minimums (MMCG Invest). In practice, that meant no mandatory minimum equity injection for most transactions — genuinely zero-down acquisitions were routinely approved — seller standby notes could satisfy up to 100% of whatever equity a given lender chose to require, collateral was only required above $500,000, and the streamlined "7(a) Small Loan" processing track extended up to a $500,000 ceiling with a comparatively low minimum SBSS credit score of 155. Loeffler's own characterization of the result, delivered to the House Appropriations Subcommittee on July 14, 2026: underwriting standards "were dismantled, contributing to an estimated $2.2 billion in losses that taxpayers will bear."

What replaced it: SOP 50 10 8

Effective June 1, 2025, applying to every loan that received an SBA loan number on or after that date, SOP 50 10 8 is described in SBA's own Information Notice as a policy that is "largely re-implementing requirements that were in place before January 2021... eliminating the 'do what you do' philosophy from the SOP" (NAGGL). That framing matters: this is explicitly a restoration of prior standards, not an entirely novel or punitive tightening invented from scratch. A second wave of the same reset took effect March 1, 2026, ending mandatory reliance on the SBSS credit-scoring model for the 7(a) Small Loan category and replacing it with an explicit debt-service-coverage-ratio floor for every file regardless of size.

Sources: MMCG Invest; NAGGL; SBA.gov.

One line in that table deserves its own callout, because it's the single most misunderstood requirement in SBA lending: the personal guarantee for any owner holding 20% or more equity is required by federal regulation, specifically 13 CFR §120.160(a). This is not part of the rollback, and it never went away during the "Do What You Do" era either — it's a constant that predates both regimes and will outlast whatever underwriting philosophy SBA runs next. If you've encountered marketing claiming "no personal guarantee" SBA financing exists for a 20%+ owner, that claim is not accurate under current or prior policy.

The financial case SBA cites for the reset

SBA's own numbers make the case for why this reset happened, whatever you think of the politics around it. The 7(a) program posted approximately $397 to $400 million in negative cash flow in FY2024 — its first deficit in thirteen years. The default rate reached 3.7% in FY2024, the highest since 2012, per SBA's own FY2024 Annual 7(a) Risk Analysis Report, and rose further to a reported 4.8% by March 2026 on an independent loan-level analysis — the highest since 2013 (Lumos SBA lending report). Defaulted-loan purchases by SBA reached $1.6 billion in FY2024 alone, up from $1.1 billion in FY2023 and $733 million in FY2022. Separately, roughly $460 million in guarantee fees went uncollected from FY2022 through FY2024 due to fee waivers that were later reversed, contributing directly to the FY2024 cash-flow deficit that catalyzed the entire reset (House Small Business Committee).

Independent loan-level analysis from Lumos Data offers useful context on how to read that default-rate trajectory honestly. The 12-month conditional default rate hit 4.8% by March 2026 — roughly three times the 2021 stimulus-suppressed trough of 1.6%. But Lumos's own analysis flags that the 2021 trough was artificially deflated by pandemic-era relief programs under which the federal government was literally making loan payments on borrowers' behalf, meaning that trough is not a fair baseline for comparison. The more honest historical comparison is to the pre-pandemic dot-com-bust-era peak of roughly 5.1%, which puts the current 4.8% figure "within about a quarter point" of the worst level the program has seen in three decades — a real deterioration, but one that should be measured against the right historical benchmark rather than an artificially low pandemic-era floor (Lumos SBA lending report). It's also worth noting that broad bank commercial-and-industrial delinquency rates, outside the SBA-specific population, remain benign — in the 1.3% to 1.9% range as of Q1 2026 — meaning this deterioration is concentrated specifically in the "credit elsewhere" statutory population that 7(a) exists to serve, not a signal of broader economic distress.

A nuance worth surfacing: volume decline isn't purely a risk-reduction story

Here's a complicating detail that a purely celebratory reading of the tightening would miss, and that we think is important to include for full accuracy. Lumos's own size-bucket data shows that the loan-size band that fell hardest in volume — $350,000 to $500,000, down 64% — is not actually the riskiest band in the portfolio. It ranks fourth of seven on default rate, at 3.9%. The genuinely riskiest bands — sub-$150,000 loans at 4.8%, and $150,000-to-$350,000 loans at 4.7% — fell by considerably less. Lumos's own conclusion is worth quoting directly: "If credit were driving the retreat, the worst-performing bands should have been cut hardest. They were not... Lenders did not exit the riskiest loans. They exited one specific band, and that band is defined by a processing rule, not by credit" (Lumos SBA lending report). The explanation: loans above $350,000 lost their streamlined "Small Loan" processing track under SOP 50 10 8 and now require full manual underwriting, which raises the lender's cost to originate relative to the loan's size — a processing-cost effect layered on top of, and somewhat distinct from, a pure credit-risk effect. Some borrowers in that specific band may be getting squeezed out or resized downward for reasons that are about lender processing economics as much as their own creditworthiness — a distinction worth knowing if your deal happens to land in that exact range.

A related and genuinely striking finding involves the largest loans in the portfolio. In 2016, bigger SBA loans were uniformly the safest — a clean risk ladder where size and safety moved together. By 2025, the $3 million-plus bucket had become the third-riskiest band in the entire portfolio, with a 4.4% default rate, up 4.1 times since 2016 — worse than every size band between $350,000 and $3 million. Lumos attributes this to $3 million-plus loans being disproportionately structured as leveraged acquisitions and partner buyouts on floating, prime-linked rate terms with thin debt-service coverage margins, which got repriced unfavorably as the underlying benchmark rate climbed from roughly 6% to 10.5% over the relevant period (Lumos SBA lending report). This finding is directly relevant to the cap-raise conversation from Section 1: if the ceiling genuinely does rise toward $10 million, underwriting quality on the largest loans specifically — not simply loan size as a proxy for safety — will need much closer scrutiny, since the data no longer supports treating bigger loans as automatically safer loans.

Where the tightening lands by segment

It's worth being specific about who actually feels this reset most, because "underwriting got tighter" is not a uniform experience across every type of borrower. The smallest loans — under $150,000 — have seen approvals fall roughly 22% year-over-year as of Q2 FY2026, and that segment also carries the portfolio's highest default rate at 4.8%. The $150,000-to-$350,000 band retained its streamlined "Small Loan" processing track on paper, but with the SBSS mandate now retired in favor of a universal 1.10:1 DSCR calculation, every file in that range effectively requires the same documentation rigor as a manually underwritten loan — and partial-year 2026 data shows default rates in that band annualizing sharply higher, a trend worth watching rather than dismissing. Loans between $500,000 and $1 million currently carry the lowest standing annual default rate in the portfolio at 3.3%, though the early-default rate in that band has nearly quadrupled since 2019, a possible leading indicator tied to loans originated during the 2023-to-2025 fee-waiver window that examiners are still working through. By contrast, approvals above $1 million fell only about 6% year-over-year in the same quarter — evidence that the pullback is concentrated at the small-dollar end of the market rather than applied evenly across the whole portfolio.

By industry, the current data shows real dispersion worth knowing if you're benchmarking your own sector. Manufacturing runs a 70% to 76% approval rate — comparatively strong, helped by the fact that equipment itself frequently serves as usable collateral — while Accommodation and Food Services, historically the largest volume category by dollar share, runs a lower 62% to 68% approval rate alongside a higher historical default rate specific to restaurant concepts. On pure default-rate performance, Transportation and Warehousing currently runs the highest default rate in the portfolio at 7.6%, while Health Care and Finance and Insurance sit at the low end around 3.6%. None of this is a reason to avoid SBA financing if you're in a higher-default-rate sector — it's a reason to expect your file to be underwritten with the sector's track record already priced into the underwriter's expectations, and to prepare your documentation accordingly rather than being caught off guard by additional scrutiny.

What this means for the lenders you're actually working with

It's worth being clear about how this filters down to the Preferred Lender Program banks doing the actual underwriting on the ground — names like Live Oak Banking, Newtek, Byline Bank, and Regions, among the most active dedicated SBA lenders in the country, alongside the Tier 1 institutions many business owners already bank with day to day. Chase, U.S. Bank, Wells Fargo, and Bank of America all maintain active SBA lending desks, and American Express has continued expanding its small-business lending footprint beyond its traditional card business — but none of them get to apply a looser standard than SOP 50 10 8 requires, regardless of the size or reputation of the institution. These are not banks setting their own looser standards anymore; SOP 50 10 8 requires them to follow SBA's restored standards more strictly than they were required to under "Do What You Do," which means the latitude that individual PLP lenders previously had to approve a marginal file based on their own internal risk appetite has narrowed considerably. That's a structural shift in the relationship between SBA and its lending partners, not just a change in paperwork — and it means the specific PLP lender you choose to work with matters less for whether unconventional structuring will get approved (that latitude has shrunk across the board) and more for how efficiently and knowledgeably they can walk your specific file through the now-more-rigid standards. We'll cover the current PLP lender landscape in detail, including recent earnings commentary from several of the largest SBA lenders, in Part 2 of this analysis.

Section 5 — MAMFA: The Legislative Vehicle For The Cap Raise

Part 1 introduced the bill Loeffler named as her preferred mechanism for a cap increase. Here in Part 2, we go deeper into what that bill actually does, where it currently sits in the legislative process, and — critically — why "MAMFA passes" and "the general 7(a) cap rises to $10 million for every business" are not the same event, even though they get talked about interchangeably in casual coverage.

What MAMFA actually is

The Made in America Manufacturing Finance Act (MAMFA) — introduced simultaneously as H.R. 3174 in the House and S.1555 in the Senate on May 1, 2025 — was drafted by the House and Senate Small Business Committees working directly with SBA (NAGGL). Per the bill text itself, MAMFA amends Section 7(a) of the Small Business Act to create a manufacturer-specific carve-out — raising the guaranteed-exposure ceiling for qualifying "small manufacturer" borrowers (NAICS sectors 31, 32, and 33, with all production facilities located in the United States) to $7.5 million where the gross loan amount would otherwise exceed $3.75 million, and up to $10 million in gross loan amount under the manufacturer-specific 504/working-capital provisions (Congress.gov, S.1555 full text). Read that carefully and the scope becomes clear: this is a targeted expansion for one sector, not a blanket increase for every 7(a) borrower regardless of industry.

That distinction is where public conversation about "the $10 million cap" gets muddled. Loeffler's personal position — "I absolutely would" support raising the general cap — is broader than what MAMFA, as drafted, delivers. If you're a manufacturer, MAMFA is your bill to watch. If you're outside NAICS 31-33, MAMFA doesn't directly raise your ceiling — the July 4, 2026 decoupling covered in Part 1 remains the mechanism that already benefits you, and any further general increase would require separate legislation that hasn't been drafted or introduced.

Sponsors, committee status, and where the bill sits today

On the House side, H.R. 3174 has already cleared the highest bar a bill in this Congress can clear short of full passage into law: it passed the House floor unanimously and on a bipartisan basis, a rarity for anything touching SBA policy in the current environment (SBA.gov). The Senate companion, S.1555, has moved through committee hearings at the Senate Small Business and Entrepreneurship Committee (chaired by Sen. Joni Ernst, with Sen. Ed Markey as ranking member) as of May 2025, on a bipartisan committee vote — but this desk could not independently confirm a full Senate floor vote as of July 28, 2026 (Congress.gov). In plain terms: MAMFA is further along than almost any comparable SBA bill in recent memory, but it is not law yet, and the remaining step — full Senate floor passage, followed by reconciliation with the House text and a presidential signature — is not a formality. It's a real, sequential process that can stall for reasons entirely unrelated to the bill's substance, from floor-time scarcity to unrelated procedural fights.

One number materially strengthens MAMFA's odds relative to typical legislation: the Congressional Budget Office scored the cost impact of raising manufacturer loan guarantee caps to $10 million as "insignificant" (CBO, H.R. 3174 cost estimate). That score matters procedurally because it means the bill likely doesn't trigger PAYGO offset requirements — one of the more common ways worthwhile legislation quietly dies in committee. A bill that scores as budget-neutral has a meaningfully easier path than one that requires Congress to find new revenue or cut something else to pay for it.

The NDAA rider hypothesis

In her Forbes interview, Loeffler said the administration is "hoping for passage as part of defense funding" — a reference, almost certainly, to the FY2027 National Defense Authorization Act, running as H.R. 8800 in the House (with a CBO score of $1.1 trillion in FY2027 defense authorizations) and S.4784 in the Senate, filed by Chairman Roger Wicker and Ranking Member Jack Reed (CBO; Senate Armed Services Committee). Attaching a popular, bipartisan, budget-neutral small business bill to a must-pass defense authorization bill is a well-worn legislative strategy — NDAA bills move every year, almost never fail outright, and frequently become vehicles for unrelated riders that might otherwise struggle to find independent floor time.

Here's the honest caveat, and we'd rather tell you this directly than let you assume the rider is a done deal: this research did not find confirmation that MAMFA's loan-cap language has been formally attached to either chamber's FY2027 NDAA text as of publication. The House Rules Committee's amendment tracker for H.R. 8800 does show a related but distinct SBA-adjacent amendment — a bipartisan measure from Reps. Cisneros, Van Orden, Olszewski, and Harris that "simplifies and streamlines single-source contracts across SBA programs by implementing a $10 million threshold" — but that provision governs federal contracting thresholds, not lending caps, and should not be conflated with MAMFA's manufacturer loan-cap language (House Rules Committee amendment tracker). Whether Loeffler's stated hope becomes an actual amendment remains an open question worth tracking through the rest of the NDAA markup process, not a settled outcome you should plan your funding timeline around.

The historical cap-raise cadence — why the 16-year gap is the whole argument

Loeffler leans on one number more than any other in every venue she's spoken in this cycle: 16 years since the last cap increase. Here's the full historical cadence, laid out so you can see exactly how rare these increases actually are:

Sources: American Land Title Association; Journal of Accountancy; Congress.gov.

There's an inflation-adjustment wrinkle worth surfacing here too, because it complicates the "just catching up" framing that sometimes gets applied to the $10 million figure. If the $5 million cap set in 2010 had simply tracked inflation since then, it would sit at roughly $7.2 million today (Forbes/Kochkodin). A move to $10 million, in other words, isn't merely restoring lost purchasing power — it's a real, above-inflation expansion of capital access. That's a meaningful distinction for the political feasibility conversation: proponents can frame $10 million as more generous than a pure inflation catch-up would require, while critics of the zero-subsidy math (covered in Section 7) can point to the same fact as evidence the risk modeling needs to be unusually careful at that size.

Political feasibility, read plainly

Several factors favor eventual passage. MAMFA has demonstrated genuine, repeated bipartisan support — unanimous House passage and bipartisan committee votes in both chambers, which is close to unheard of for anything SBA-adjacent in the current Congress. The CBO's "insignificant" cost score removes the single most common procedural obstacle. And notably, the loudest Democratic opposition to Loeffler's broader agenda — from Reps. Velázquez and Sens. Murray and Markey — has concentrated on the citizenship eligibility rules and the proposed 67% agency budget cut, not on the manufacturer cap increase itself (MMCG Invest). That suggests the cap-raise concept specifically carries less partisan risk than other pieces of the current SBA policy agenda.

Section 6 — The Manufacturer Segment: Where Loeffler's Signals Land Hardest

If there's one segment of the small business economy that should be reading this entire two-part series with a highlighter, it's manufacturing. Every policy thread we've covered so far — the cap raise, the July 4 decoupling, even the underwriting rollback's practical effects — lands on manufacturers with more force, and more favorably, than on almost any other industry currently eligible for SBA financing.

200 manufacturers already hit the ceiling

Per Loeffler's own Congressional testimony, as recapped by the Coleman Report, roughly 200 loans have hit the existing $5 million ceiling "in the last couple of years," and manufacturing has become one of SBA's largest and fastest-growing lending categories by her account (Coleman Report). That 200-loan figure is the more precise, citable number from her testimony — the Forbes interview's looser phrasing of "by the hundreds" is directionally consistent but less specific. Either way, the signal is the same: a meaningful and growing population of manufacturers is running into a hard ceiling that hasn't moved in sixteen years, and that population is exactly who MAMFA and the July 4 decoupling are designed to serve.

MARC — the first SBA loan program built specifically for manufacturers

SBA's newest lending product, the MARC program (Manufacturer's Access to Revolving Credit), is SBA's own description of it: the agency's "first-ever loan program dedicated to America's small manufacturers," offering up to $5 million in working-capital revolving credit specifically for manufacturing businesses (SBA.gov). Its launch was disrupted by the 43-day government shutdown covered in Part 1, which delayed the program's initial rollout window (Manufacturing Dive). Worth knowing before you get too excited about it: some lenders have expressed real skepticism about MARC specifically, because its revolving-credit structure cannot be sold on the secondary market the way term loans can, and it requires heavier ongoing borrower reporting than a standard 7(a) facility — meaning not every PLP lender is equally enthusiastic about originating MARC loans, and you should ask directly about a lender's MARC experience and appetite before assuming it's a straightforward addition to your capital stack.

Alongside MARC, SBA has introduced a 90% Made in America Loan Guarantee specifically for small manufacturers — a meaningfully higher guarantee than the standard 75-85% tiers most 7(a) borrowers receive (SBA.gov). A higher guarantee percentage matters practically because it reduces the lender's unguaranteed exposure on the loan, which typically translates into more lender willingness to approve borderline files and, in some cases, better pricing.

Fee waivers on the table

SBA has already waived upfront guarantee fees entirely for qualifying small manufacturers on loans up to $950,000, effective October 1, 2025 (SBA.gov). Given the standard fee schedule runs from 0% on the smallest loans up through the mid-3% range on larger guaranteed portions, a full waiver up to $950,000 is a real, immediately usable cost reduction for manufacturers financing equipment or facility expansion at that scale — not a hypothetical future benefit contingent on legislation passing.

Small manufacturer unlimited 504 access — already in effect

This is the piece of the current framework that's easiest to overlook, and it's already live: small manufacturers have long had access to unlimited 504 financing across multiple distinct projects, provided each project is genuinely separate (a new facility, a new equipment line, and so on don't count against each other the way they would under a single-project cap). Layer the July 4, 2026 decoupling on top of that unlimited 504 access, and a qualifying manufacturer can now combine unlimited 504 project financing with a full $5 million 7(a) facility for working capital or equipment — a materially more favorable stacking position than almost any other industry currently has available under existing law, with zero additional legislation required (SBA.gov).

Stacking Section 179 and bonus depreciation with SBA 504

For manufacturers financing equipment specifically, there's a tax-side lever worth coordinating with your accountant alongside the SBA financing conversation: Section 179 expensing and bonus depreciation both allow a business to deduct the cost of qualifying equipment in the year it's placed in service, rather than depreciating it over its useful life. Financing that equipment through an SBA 504 loan — which is purpose-built for major fixed-asset purchases like machinery, real estate, and heavy equipment — doesn't disqualify you from claiming those deductions; you're financing the purchase, not paying cash, but the tax code generally allows the deduction based on when the asset is placed in service regardless of how it was paid for. That combination — 504 financing for the capital outlay, Section 179/bonus depreciation for the tax treatment — is one of the more underused pairings we see manufacturer clients overlook, largely because the SBA conversation and the tax-planning conversation happen with two different advisors who never talk to each other.

What actually counts as a "manufacturer" for SBA classification

This matters more than it might seem, because eligibility for MARC, the 90% guarantee, the fee waivers, and MAMFA's eventual cap increase all hinge on the same classification: NAICS sectors 31, 32, and 33, generally with production facilities located in the United States. That's a broader category than people often assume — it includes not just heavy industrial operations but food processing, apparel and textile production, furniture manufacturing, printing, plastics and rubber products, fabricated metal products, machinery manufacturing, and electronics assembly, among others. Per Loeffler's Senate testimony, 99% of all U.S. manufacturers qualify as "small manufacturers" under the relevant statutory definition — meaning if your business genuinely produces a physical product under one of those NAICS codes, you are very likely eligible for this entire favorable policy stack even if you don't think of yourself as a traditional "factory" (Senate testimony, sbc.senate.gov).

Manufacturing-adjacent sectors: the trucking connection

Trucking and logistics operators (NAICS 484) fall outside the strict manufacturing NAICS codes, but they're worth mentioning here because the underlying lesson about lender compliance applies with equal force across every capital-intensive, equipment-heavy industry adjacent to manufacturing. We've told this story before because it's one of the clearest illustrations we have of how a single compliance error can quietly sink an application that has nothing wrong with it financially: a trucking client came to us after being denied by two prior funding companies, with no clear explanation from either one about why. Our Bankable Scan — the 20-program compliance check we run on every new file — found the root cause in about five minutes: a PO box listed as the business address on his Experian Business file. Lenders read a PO box as a signal of instability or, worse, a shell operation, regardless of how strong the underlying business actually is. We fixed it that same day. The point for manufacturer and manufacturing-adjacent readers specifically: none of the favorable policy tailwinds in this section — MARC, the fee waivers, unlimited 504 access — will save an application with a lender-compliance problem sitting underneath it. Compliance is leg one of four for a reason; it's usually the first thing an underwriter's automated system checks, and it's invisible to the business owner until someone actually runs the scan.

Section 7 — Zero-Subsidy Framing: What It Actually Constrains

We introduced the zero-subsidy standard in Part 1 as the legal backbone behind Loeffler's caution around every cap-raise claim she's made. Here, we walk through exactly how it works mechanically, what it costs borrowers today, and why it's the single biggest constraint on how fast — and how far — any future cap increase can actually move.

The statutory basis: SBA has operated at zero subsidy since 2013

Congressional appropriations reform established that the 7(a) program must operate at zero subsidy to taxpayers — meaning the program's expected credit losses have to be fully covered by the fees SBA collects from lenders and borrowers, with no ongoing federal appropriation subsidizing operations (House Small Business Committee). Rep. Roger Williams (R-TX), Chairman of the House Small Business Committee, put it plainly: "This program is intended to be self-sustaining, ensuring proper support for small businesses nationwide while safeguarding taxpayer funds." That single sentence is the frame through which every cap-raise conversation inside SBA has to pass. It's not a political talking point layered on top of the policy — it's the actual legal ceiling on what the agency can do without going back to Congress for new appropriations authority, which is a far higher and slower bar than an administrative rule change.

The current fee schedule funding the program

Zero-subsidy is achieved mechanically through the upfront guarantee fee schedule, reinstated by SBA effective March 27, 2025 after being waived or discounted between 2021 and 2025. Here's the FY2026 schedule as it currently stands:

Sources: MMCG Invest; House Small Business Committee.

Fee waivers during the 2021-2025 window cost the program an estimated $460 million in uncollected revenue, contributing directly to the roughly $400 million negative cash flow the 7(a) program posted in FY2024 — its first deficit in thirteen years, and the financial event that catalyzed the entire underwriting reset covered in Part 1 (House Small Business Committee). That history is exactly why Loeffler treats fee discipline and underwriting discipline as two sides of the same coin, and why any future cap raise gets evaluated through the same lens: does the fee revenue this loan size generates actually cover the expected losses this loan size produces?

Why zero-subsidy directly constrains cap-raise politics

This is where the data from Part 1's underwriting section becomes directly relevant to the cap-raise conversation rather than a separate topic. Independent loan-level analysis shows the $3 million-plus loan bucket has gone from the safest size band in the portfolio in 2016 to the third-riskiest by 2025 — a 4.4% default rate, up 4.1 times over that period, driven largely by leveraged acquisitions and partner buyouts on floating, prime-linked rate structures that got repriced unfavorably as benchmark rates climbed (Lumos SBA lending report). If that's the risk profile of loans in the current $3M-$5M range, a program extending its ceiling to $10 million has to solve a real actuarial problem before it can responsibly do so: either the fee revenue collected on the largest loans needs to rise, the underwriting standard applied specifically at that size tier needs to tighten further, or some combination of both — otherwise the math tips the program into an actual taxpayer subsidy, which isn't just a political problem, it's a legal one under the current statutory framework.

This is precisely the caveat Loeffler herself attached to every expansion claim in the Forbes interview: "we're testing the default rates at every size. We're modeling what that looks like to go forward. So we would just want to make sure that it comports with our commitment to be cost-neutral to taxpayers" (Forbes). That's not hedging for its own sake — it's an accurate description of the actual constraint she's operating under.

The alternative: differentiated fee tiers

One plausible path that would let SBA raise the ceiling without violating zero-subsidy is a more sharply differentiated fee structure at the top of the loan-size distribution — essentially, a higher marginal fee specifically on the portion of a loan above the current $5 million mark, priced to reflect the higher observed default rate in that size band rather than applying a flat percentage across the whole range. This is speculative on our part; SBA has not published a specific proposed fee schedule for a $10 million ceiling as of this writing. But it's the logical mechanism given everything else we know about how the program is required to fund itself, and it's worth watching for in any future MAMFA implementation guidance or NDAA rider text.

What a fee bump would mean for your effective borrowing cost

Here's the part that's easy to miss if you're only thinking about the headline loan-size ceiling: SBA guarantee fees are typically financed directly into the loan amount rather than paid out of pocket at closing, which means a higher fee doesn't just cost you more in absolute dollars — it raises your effective APR slightly, since you're now paying interest on a larger financed balance to cover the same net proceeds. If SBA does introduce a higher fee tier to make a $10 million ceiling actuarially sound, borrowers accessing that top tier should expect their effective cost of capital to rise modestly relative to what a $5 million loan costs today, even before accounting for whatever the prevailing interest-rate environment looks like at the time you close. That's a fair tradeoff for access to double the capital — but it's worth planning for rather than assuming a bigger cap arrives at the same effective price as today's $5 million ceiling.

Section 8 — PLP Landscape Q3 2026: Which Lenders Are Best Positioned For The Tightening

Part 1 closed its underwriting section by noting that Preferred Lender Program (PLP) status — the delegated authority certain banks hold to close 7(a) loans without waiting for SBA's own pre-approval on every file — no longer functions as a way to shop for a looser standard. SOP 50 10 8 applies the same restored baseline to every PLP lender regardless of size or reputation. What PLP status still determines is speed, sophistication in walking your file through the new documentation requirements, and — increasingly, per the Q2 2026 earnings data below — the lender's own financial incentive to keep originating quality SBA loans at all.

Top FY2026 SBA 7(a) lenders by volume

Sources: NerdWallet; SBAlenders.com; Bankrate; GoSBA Loans.

Worth naming directly: Chase, one of our five core Tier 1 stacking banks for 0% business credit cards, maintains a comparatively limited direct SBA 7(a) lending footprint relative to its dominance in the business credit card space — most of Chase's small business relationship focus runs through its Ink card lineup rather than SBA origination volume. That's not a knock on Chase; it simply means when you're ready for SBA financing specifically, you're more likely to be routed toward a dedicated SBA lender or a different Tier 1 relationship bank with a deeper SBA desk, even if Chase remains central to your earlier-stage 0% credit card stacking.

Q2 2026 earnings commentary: the top lenders are adapting, not retreating

Live Oak Bancshares (NYSE: LOB), reporting Q2 2026 results on July 22, 2026, posted net income of $34.7 million ($0.74 per diluted share, up 45% year-over-year and ahead of the $0.68 consensus estimate). Loan originations reached $1.55 billion across 33 industries in the quarter, with total loans growing 16% year-over-year to roughly $13.1-13.2 billion. Notably, government-guaranteed loans now represent 29% of Live Oak's total portfolio, down from 33% a year earlier — a deliberate strategic shift toward fee-based originate-to-sell activity via Live Oak Express, its small-dollar SBA program, which hit record originations of $82 million in the quarter, up 63% year-over-year, and generated the bank's highest-ever quarterly gain-on-sale figure at $5 million (Live Oak IR / Nasdaq press release; Investing.com). Live Oak also touts a 10-year average SBA net charge-off ratio of just 0.4%, against an industry average of roughly 1.2% — a 3x credit-quality outperformance the bank explicitly leans on as evidence its underwriting discipline holds up against the broader portfolio deterioration described in Part 1.

Byline Bancorp (NYSE: BY), reporting Q2 2026 on July 24, 2026, posted record profitability: net income of $40.2 million ($0.90 per diluted share, beating the $0.79 consensus by roughly 14%), on revenue of $117.7 million. Its combined government-guaranteed lending portfolio — SBA 7(a) plus USDA — totaled $492.8 million, with unguaranteed exposure down to 5.4% of the total portfolio from 14.6% back in 2016, reflecting sustained reliance on the guarantee mechanism as its core risk-mitigation strategy. Gains on sale of government-guaranteed loans reached $6.1 million on $78.1 million of guaranteed loans sold during the quarter (Byline Bank IR / Investing.com).

Read together, these two results tell a coherent, non-alarmist story: the top-tier PLP lenders are not shrinking away from SBA lending under the tighter underwriting regime — they're recalibrating profitably around it, originating fewer but better-qualified loans and monetizing them more efficiently through secondary-market sales, exactly consistent with Loeffler's "quality over quantity" framing from Part 1. That's a meaningful, practical signal: the biggest, most sophisticated SBA lenders in the country are betting their own balance sheets on this underwriting regime holding, which should give you real confidence that it isn't a temporary political posture likely to reverse.

What the tightening means for lender selection

Since PLP lenders can no longer differentiate on underwriting latitude, the practical differentiators that remain are speed, sector expertise, and — per the earnings data above — financial durability. Lenders with strong underwriting discipline and profitable secondary-market operations, like Live Oak and Byline, are positioned to keep originating through the tightening. Marginal PLP lenders who built volume during the "Do What You Do" era on thinner underwriting practices face a real risk of losing delegated authority or exiting the program voluntarily — consistent with the nearly 500 lenders Loeffler said have already walked away from SBA lending over lost confidence in the guardrails, cited in Part 1.

This desk's July 24 coverage of Amex's Q2 2026 earnings covered how the card issuers positioned for Round 1 stacking activity heading into H2 2026, and our July 27 SBA Advocacy piece laid out the prime-rate and business-formation backdrop this lender landscape sits inside. Read together with this section, the three pieces give you a complete, current picture of where the card-stacking side, the macro backdrop, and the SBA lender landscape all stand as of late July 2026 — which matters because none of these three pieces should be read in isolation if you're actually sequencing a real capital stack rather than just researching one product in a vacuum.

Section 9 — What Business Owners Should Do Right Now

Everything in this two-part analysis converges on one practical question: what should you actually do with this information? We built our answer as a 30-60-90 day framework, extending the same structure we used in our July 27 SBA Advocacy article, because the underlying discipline doesn't change based on which policy headline prompted you to start — it changes based on where your file actually stands today.

Days 1–30: Audit

Start by finding out, precisely, where your four legs actually stand — not where you assume they stand. Run a full lender-compliance audit across your Secretary of State filing, IRS records, and all three business bureaus (Experian Business, D&B, Equifax Business), looking specifically for the kind of silent killer that sank our trucking client's file: mismatched addresses, PO boxes, inconsistent phone numbers, or wrong industry codes. If you haven't already, open business checking accounts at any of the five Tier 1 banks you're not currently banking with — Chase, American Express, U.S. Bank, Wells Fargo, and Bank of America — since banking footprint and depth is one of the four legs and takes months to season properly. Pull all three business credit bureau reports and get a current read on your personal FICO across all three personal bureaus as well, since SBA underwriting under SOP 50 10 8 scrutinizes both business and personal credit closely.

Days 30–60: Address the gaps

Once you know exactly what's weak, fix it deliberately rather than all at once. If your trade line count is thin, seed it with vendors that actually report to the business bureaus — Uline, Grainger, and Nav Prime Tradelines are common starting points — and understand that these need to season for six or more months before they meaningfully move your business credit scores; there's no way to compress that timeline artificially. If your personal utilization is elevated, pay it down below 30% — remember, utilization has no memory, which means the moment you bring the ratio down, the negative signal it was sending starts clearing immediately, it doesn't linger just because it was high last month. If your books are inconsistent or incomplete, this is the window to get two full years of clean financials in order, because SOP 50 10 8's DSCR-driven underwriting standard has zero tolerance for financials an underwriter can't immediately verify.

Days 60–90: Prepare the file and identify your lender

In the final stretch, assemble the actual financial package an SBA underwriter will scrutinize: a current P&L, a balance sheet, and your most recent two years of business tax returns, alongside a real DSCR calculation you can defend rather than one you're hoping clears the 1.10:1 floor by accident. This is also the point where a Bankable Blueprint consultation is most useful — not at day one, when you don't yet know what your gaps are, but once you have a real picture of your file and need a strategist to sequence the remaining steps and identify which PLP lender profile — Express-focused, traditional PLP, or a larger relationship bank — actually fits your specific deal size and industry.

The anchor case: Frank

We've referenced Frank's file before because it's one of our proudest case studies, and it's exactly the profile that thrives under the tighter underwriting environment described throughout this article rather than getting squeezed by it. Frank, a real estate investor with roughly $2 million in revenue and an 800 FICO score, worked with us across three funding rounds totaling roughly $1 million. His third round included a $350,000 SBA Express facility specifically structured to refinance 0% balances that were approaching the end of their introductory period into long-term, lower-cost debt — exactly the kind of intentional, sequenced use of SBA financing we described in the order-of-operations framework, not a first-resort emergency loan. Midway through that round, Frank's score dropped from the 800s into the 600s because of a cosigned student loan going late — a real, live crisis that could have derailed the entire approval. Our team fixed it mid-round. The lesson for readers navigating SOP 50 10 8's stricter standard: a properly-prepped file with a real relationship behind it can survive a genuine credit shock that would sink an unprepared applicant outright. That resilience is exactly what "becoming bankable" is supposed to buy you.

If your personal FICO needs work first

For owners whose personal credit needs meaningful repair before any of the above becomes realistic, Patrick built creditblueprint.org as a free, do-it-yourself resource specifically for fixing personal FICO issues — disputing inaccuracies, understanding utilization mechanics, and building the kind of clean personal profile that underlies everything else in this article. It's not a substitute for the full advisory engagement, but it's a legitimate starting point if your Days 1-30 audit reveals your personal credit is the actual bottleneck standing between you and a bankable file.

Section 10 — Cross-Referencing The Fed + SBA Policy Stack

SBA policy doesn't move in isolation from the broader interest-rate environment, and this desk's July 29 FOMC coverage laid out the Fed's positioning heading into its policy decision: hike odds running 34-38% for the immediate meeting, jumping to roughly 82% by September, and climbing further to 96%-plus by October, per that piece's data. Read alongside everything in this article, the combination matters because SBA loan pricing is directly tied to the prime rate, which moves in lockstep with Fed policy.

A compressed underwriting environment coming in H2 2026

Put the two policy tracks side by side and the picture for the second half of 2026 becomes clearer: SBA underwriting tightening under SOP 50 10 8 on one axis, and a real possibility of the Fed raising rates further on the other. If both materialize together, borrowers face a genuinely more demanding environment than either factor alone would produce — stricter documentation and equity requirements at the same time borrowing costs may be rising. That's not a reason for alarm, but it is a reason to take the timing of your own application seriously rather than treating "whenever I get around to it" as a neutral choice.

Reconciling the tailwinds against the headwinds

The timing implication for your specific file

If your file will genuinely be ready in 60-90 days — meaning your four legs are close to complete and you're mostly finishing documentation rather than starting from a thin or damaged profile — there's a real case for prioritizing speed to lock in current SBA 7(a) pricing before a potential September rate hike takes hold. That's not urgency for its own sake; it's a legitimate rate-timing argument grounded in the actual FOMC probabilities our July 29 coverage documented. If, on the other hand, your file needs six or more months of genuine preparation — repairing personal credit, seasoning trade lines, cleaning up two years of financials — that's equally fine, and you shouldn't rush a weak file into an application just to beat a rate hike. The MAMFA cap raise, per Section 5, is unlikely to become usable law within that same six-month window regardless of what happens with the NDAA, so there's no meaningful cap-raise timing pressure pushing against a properly paced preparation period. Rate timing and cap-raise timing are simply not the same clock, and conflating them leads to bad decisions in both directions.

Why the macro variables matter less than you'd think

We'll close this section with an anecdote we return to often, because it's a useful corrective against over-indexing on macro headlines. Patrick often tells the story of a 16-year-old martial arts student he mentored years ago — a kid who started adding authorized users to his credit profile and using secured credit-building strategies before he'd even graduated high school, well before any of the big-picture economic conditions around him were particularly favorable or unfavorable. The point of that story isn't nostalgia. It's that the fundamentals of bankability — clean compliance, seasoned trade lines, a real banking relationship, defensible financials — matter more to your actual funding outcome than any single macro variable, whether that's a Fed hike, a government shutdown, or a Congressional cap-raise vote. Rates go up, rates go down, caps rise sixteen years apart, shutdowns happen and end. The businesses that keep getting funded through all of it are the ones that treated bankability as a discipline rather than a reaction to whatever happened to be in the news that week.