The Mortgage Bankers Association (MBA) projects $806 billion in 2026 commercial and multifamily mortgage originations, a 27 percent increase from the 2025 estimate of $633.7 billion. What is driving the surge, and what does it mean for CRE investors? The headline answer is rate stabilization, near-term loan maturities, and pent-up transaction activity. The deeper answer is that AI-assisted underwriting is the operational reason the lending market can absorb this volume increase without a corresponding hiring surge at lenders. The CRE investors who position for the 2026 origination environment understand both the rate cycle and the AI-driven workflow shift inside lender shops. For the broader framework of AI in deal-side analysis, see our pillar guide on AI deal analysis real estate scoring.
Key Takeaways
- The MBA forecasts $806 billion in 2026 commercial and multifamily mortgage originations, a 27 percent increase over 2025.
- The volume surge is driven by rate stabilization, a wall of 2026 to 2027 loan maturities, and acquisition activity that is forecast to grow 15 to 20 percent.
- AI underwriting tools (Claude, GPT-5.5, Gemini) are how lenders absorb the volume without proportional headcount growth; the workflow shift is the under-reported part of the story.
- Borrowers who present AI-structured deal packages close roughly 30 percent faster and tend to negotiate tighter terms because their packages match lender intake systems.
- CRE investors should expect tighter underwriting on environmental, structural, and tenant-credit risk, because AI does not relax standards; it enforces them more consistently.
Why the $806B Forecast Matters
The 2025 origination total of $633.7 billion was a recovery year from the 2023 to 2024 trough, but it remained well below the 10-year peak. The 2026 forecast of $806 billion would be the highest origination volume since 2022 and would represent meaningful normalization in CRE debt markets. Three structural forces drive the number.
First, the maturity wall. Roughly $1 trillion in CRE loans mature in 2026 and 2027 combined, with a meaningful share originated at lower rates between 2020 and 2022. These maturities create refinancing demand whether or not the property has appreciated in value, and most of them go to origination rather than amortization-only extensions. Second, transaction volume recovery. CRE sales volume is forecast to increase 15 to 20 percent in 2026, with each acquisition typically requiring 60 to 75 percent loan-to-value financing. Third, rate stabilization. The 10-year Treasury has settled into a tighter band in early 2026, which makes both lenders and borrowers more willing to commit to 7- and 10-year fixed-rate paper.
The AI Underwriting Capacity Story
The MBA forecast assumes lenders can actually process $806 billion in originations. That assumption is doing more work than it appears. A 27 percent volume increase without a 27 percent headcount increase requires meaningful productivity gains, and the gains are coming from AI.
The AI underwriting workflow inside lender shops in 2026 looks roughly like this. The borrower's deal package (offering memorandum, rent roll, T12, leases, environmental, PCA) is uploaded to the lender's intake system. An AI layer (commonly Claude Opus 4.7 or Gemini Pro) extracts and structures every financial line, ties it to the lender's underwriting template, runs the initial sizing and DSCR sensitivity, and flags exceptions. A junior analyst spends 30 to 60 minutes verifying the AI output rather than 6 to 8 hours building the analysis from scratch. A senior credit officer reviews the structured analysis and exception list, then makes the credit call. The total elapsed time to a term sheet drops from 5 to 7 business days to 1 to 3 days for clean files.
What This Means for CRE Borrowers
The AI workflow has three direct consequences for borrowers in 2026.
1. Speed Premium for Clean Files
Borrowers who submit AI-structured deal packages, with each financial document pre-extracted into the lender's expected format, close roughly 30 percent faster than borrowers who submit raw documents. This is because the AI extraction layer at the lender works better with clean inputs, and clean inputs reduce manual rework. Borrowers can structure their own packages using the same Claude or GPT-5.5 models the lenders use; doing so essentially pre-runs the lender's intake.
2. Tighter Underwriting on Risk Categories
AI underwriting does not relax standards. It enforces them more consistently. Pre-AI, an underwriter might miss an environmental Phase 1 finding under time pressure; AI surfaces every flagged item every time. Borrowers should expect more consistent attention to environmental, structural, tenant-credit, and rent roll quality. Properties that previously closed despite soft findings will face harder questions in 2026.
3. Term Negotiation Edge
Borrowers who understand the lender's AI-driven sensitivity analysis can negotiate term more precisely. If the lender's AI is running 200-basis-point rate stress and identifying DSCR breakpoints, the borrower can come to the table with the same analysis and propose structure (interest-only periods, extension options, partial recourse) that the AI shows is acceptable to the lender's risk model.
What This Means for CRE Acquirers
For acquirers, the $806B environment combines with AI underwriting in three ways. First, financing availability is up; the maturity wall and rate stabilization create lender appetite. Second, the lender's diligence is faster; expect term sheets in days, not weeks, for clean files. Third, the deal cycle is compressed; sellers are running shorter LOI windows because AI on the buy-side is also faster. Acquirers who are not running their own AI workflow on the deal side will lose ground to those who are. Industry research from CBRE indicates that AI-assisted CRE workflows now contribute meaningfully to win rates in competitive bid situations.
For investors evaluating which AI tools to deploy on the borrower side, our benchmark guide on AI underwriting speed test: ChatGPT vs Claude vs Gemini compares the major models on CRE underwriting tasks. Investors looking to evaluate CRE debt fund opportunities can also reference our AI CRE debt fund analysis guide for the lender-side mirror of this workflow.
What This Means for Capital Markets and Brokers
Capital markets brokers are the most exposed to the workflow shift. The traditional broker function (assemble the package, take it to 8 to 12 lenders, negotiate term) is partially automated by AI on both sides. Brokers who embrace AI workflows for package assembly and lender targeting will preserve and grow their pipeline; those who do not will see fee compression and lower deal counts. The brokers winning in 2026 are running AI as their analyst layer, not their replacement.
Recommended Action for CRE Investors
Three concrete moves to position for the 2026 origination cycle:
- Build your own AI underwriting workflow. Use Claude Opus 4.7 or Gemini Pro to pre-structure your deal packages before submission. The cost is negligible (under $50 per deal in API charges); the speed gain is meaningful.
- Run the lender's likely sensitivity analysis on your own deal. Anticipate where DSCR breaks, where rate stress hits, and where exceptions will be flagged. Address them in the borrower's narrative.
- Tighten your environmental, structural, and tenant-credit diligence. AI underwriting is consistent, which means soft findings get caught. Clean files close; flagged files stall.
If you are ready to build a Claude-driven deal-package and underwriting workflow tuned to lender expectations, The AI Consulting Network specializes in exactly this.
Frequently Asked Questions
Q: Is the MBA $806B forecast realistic?
A: It is in line with the maturity wall (roughly $1 trillion across 2026 and 2027) and consistent with the 15 to 20 percent transaction-volume growth forecast. The downside risk is a rate shock; the upside is acquisition volume above the 20 percent forecast.
Q: Which lenders are furthest along on AI underwriting?
A: Large banks and insurance companies have moved fastest, with debt funds following. CMBS conduits are mid-pack; agency lenders (Fannie, Freddie) have integrated AI into intake but maintain heavier human-in-the-loop review.
Q: Does AI underwriting mean smaller deals get easier or harder financing?
A: Easier in absolute terms because lender capacity expands. Marginal deals (sub-$5M, untraditional asset class) may still face friction because AI is calibrated for institutional patterns. Owners of marginal deals benefit most from AI-structured packages.
Q: What if my deal package has gaps the AI cannot fill?
A: Run your own AI on the package first to identify gaps before submission. Common gaps: rent roll formatting, missing T12, environmental Phase 1 over 6 months old, missing tenant estoppels. Fix the gaps, then submit.
Q: How does this interact with lender appetite for specific asset classes?
A: AI underwriting reads the asset class signal hard. Multifamily and industrial are well-mapped; office is heavily AI-stressed on lease rollover and tenant credit; retail is mid-pack with attention to anchor exposure. Borrowers should match their narrative to the asset-class signals AI flags.