DSCR Loan Rates 2026: What You'll Actually Pay (By Tier, Primary Sources Only)

Why this article exists

Most "DSCR loan rates 2026" articles do one of three things: (1) publish a generic 5-percentage-point band ("DSCR rates are 7-12%") with no scenario context, (2) republish 2024-era rate tables that haven't been updated since the cycle peak, or (3) quote a single aggressively low "par rate" without disclosing that it requires 2-3 points of buy-down to achieve. None of them cite primary-source data from SEC filings, earnings calls, securitization presale reports, or Federal Reserve / Treasury benchmark series.

This article fixes that. Every rate quote is cross-referenced against at least two independent sources. Hard data points come from Velocity Financial's Q1 2026 10-Q and earnings call, Angel Oak Mortgage REIT's Q1 2026 results, KBRA and DBRS presale reports on recent 2025-2026 securitizations, Federal Reserve and Treasury yield series, lender-published rate sheets, and aggregator survey data with sample sizes disclosed. Derivations are explicitly tagged [DERIVED]. Forecasts are explicitly tagged [FORECAST]. Single-source data points are tagged [SINGLE-SOURCE]. Unverified data is marked [UNVERIFIED].

Section 1: Current DSCR rate environment (May 17, 2026)

1.1 Primary-source anchor data — the verifiable record

The following data points are pulled directly from SEC filings, earnings calls, and rating-agency presale reports — none of which appear in any competing "DSCR loan rates 2026" article we surveyed in research:

Reading the data: Velocity Financial's portfolio WAC of 9.75% — verifiable from a 10-Q filed with the SEC — is the truest "average rate paid" across the broad BPL/DSCR universe because Velocity's portfolio includes RTL (residential transition loans), multi-family, and credit-flexible segments. CROSS 2026-NQM1's pool WAC of 7.15% reflects the typical investor-loan rate in a higher-quality non-QM securitization pool with mixed DSCR / bank-statement / asset-utilization underwriting. The 13-bps difference between Park Capital's A-1FCF coupon (5.44%) and A3 coupon (5.78%) reflects credit-tranche stratification — investors pay 13 bps more for thinner subordination, but the underlying borrower rate is significantly higher across the stack.

1.2 Lender-published retail rate ranges (May 2026)

Reconciliation note: The HomeAbroad "par rate" of 6.12% and Easy Street's 5.75% advertised floor are achievable only under specific best-case conditions (2-3 points of buy-down, 740+ FICO, 65% LTV, 1.25+ DSCR, 5-year prepayment penalty acceptance). Net effective rate after amortizing the buy-down points over a typical 5-year hold is closer to 7.0-7.25%. Borrowers receiving a sub-7% rate quote should always ask: "What is this quote with 0 points, and what is it with the typical broker comp baked in?"

1.3 Headline May 2026 triangulated DSCR rate ranges

After cross-referencing all the data above, the consolidated rate ranges for May 17, 2026 are:

Section 2: Macro rate drivers — what's pushing DSCR rates

2.1 Benchmark rates (May 15-17, 2026)

The critical reading: The 10-year Treasury yield closed at a 1-year high on May 15, 2026, driven by an April 2026 CPI print released May 12 showing +0.6% MoM headline / +3.8% YoY headline / +2.8% YoY core. This is the highest CPI reading since May 2023 and is materially above the Fed's 2% target. Inflation pressure is being driven by an Iran-war energy shock (+3.8% MoM), sticky shelter inflation (+0.6% MoM), and services ex-energy at +3.3% YoY. The Fed held the April 29 FOMC meeting with four dissents — the most since 1992 — suggesting internal disagreement on whether to maintain a "steady" stance vs pivot hawkish.

Implication for DSCR rates: Every 25 bps move in the 10-year Treasury flows approximately 20-25 bps into DSCR rates after the spread adjustment lag. The current 1-year high in the 10-year is likely to push DSCR best-tier rates 20-50 bps higher over the next 30-60 days unless the May CPI print (released ~June 12) surprises lower or the Fed delivers an unexpectedly dovish June FOMC statement.

2.2 Spread environment — the contrarian finding

The dominant trope in aggregator DSCR-rate content is that "DSCR loans price 200-400 bps over conventional mortgages." That has been true historically but is no longer true today.

• Freddie Mac 30-year fixed (owner-occupied benchmark): 6.30% (May 2026)
• DSCR best-tier 30-year fixed: 7.00-7.50%
• Current spread to DSCR best-tier: 70-120 bps

This is the tightest DSCR-to-conventional spread in 3+ years. The compression reflects:

1. Institutional comfort with DSCR collateral. Four years of strong DSCR loan performance have given AAA RMBS buyers confidence in the asset class.
2. Record non-QM issuance forecast. KBRA's 2026 RMBS Outlook projects $75.25B in non-QM issuance, +12% YoY from $67B in 2025. This supply absorption requires tight spreads.
3. AAA non-QM spread to UST. The Angel Oak Mortgage REIT AOMT 2026-2 deal priced AAA at +113 bps over the UST curve in early March 2026 — tight by historical standards (vs +180-220 bps during late 2023 stress).
4. Tight-to-stable KBRA outlook. KBRA explicitly characterizes 2026 non-QM RMBS spreads as "tight-to-stable through 2026" in their full-year outlook.

Practical investor implication: Borrowers shopping rates today should NOT assume the historical 200-400 bps DSCR premium. A 7.5% DSCR quote on a property where the owner-occupied conventional quote would be 6.3% is roughly the market — not a "gouge." Articles using old spread math systematically overestimate DSCR rates by ~150 bps.

2.3 ARM-vs-fixed spread is narrower than usual

The standard trope says "ARM is always 100 bps cheaper than the 30-year fixed." Currently the swap curve is unusually flat (5-yr swap 3.36% vs 10-yr swap 3.645% = only 28 bps of curve), which caps the ARM discount.

Real spreads in May 2026:

• 5/6 ARM vs 30-year fixed: ARM 0.50-1.25% lower (typical 50-75 bps; HonestCasa cites 0.625-0.75%)
• 7/6 ARM vs 30-year fixed: ARM 0.25-0.75% lower (smaller discount due to longer fixed period)

Concrete example (April 2026, per HonestCasa): Best-tier 30-year fixed 7.125-7.875% vs 7-year ARM 6.625-7.250% → spread of only ~50-65 bps.

Implication: With a flat curve, the ARM-vs-fixed trade is less attractive than in normal environments. Borrowers planning to hold a property 5+ years should evaluate whether the 50-65 bps ARM discount actually compensates for the post-reset rate-adjustment risk.

Section 3: Add-on costs that determine effective rate

3.1 Origination points

• Typical range: 1.0-2.0 points (1% of loan amount each) on retail DSCR
• Negotiable on loans >$1M: down to 0.75-1.0 points (per Lima One direct disclosure)
• Best-tier borrowers sometimes secure 0-1 points
• Lower-tier borrowers may pay 2-3 points

Points-for-rate math (industry standard): 1 point ≈ 0.25% rate reduction per Mo Abdel and multiple aggregator confirmations. Some lenders offer better promotional pricing of 0.30-0.375 bps per point.

Break-even analysis: 1 point at 0.25% rate reduction has a break-even of approximately 36-48 months. Worth it if holding period > break-even AND no plan to refinance or sell during the prepayment-penalty window.

3.2 Lender fees (the "junk fees")

• Underwriting fee: $995-$1,995 typical
• Processing fee: $495-$995 typical
• Doc prep: $250-$500
• Wire / credit / appraisal pass-throughs: ~$1,000-$2,000 combined
• Total junk fees typical: $2,500-$5,000 on a SFR DSCR loan

These do not show up in the headline rate but do show up in the APR. A 7.0% rate with 2 points and $4,500 in fees has an APR closer to 7.6-7.8%. Always compare APRs (not rates) when shopping between lenders.

3.3 Prepayment penalty schedules

Standard DSCR market norm: 5-4-3-2-1 step-down PPP (5% in year 1, 4% in year 2, 3% year 3, 2% year 4, 1% year 5, zero thereafter). This is the structure across Mo Abdel, Calcix, Lima One, and most non-bank DSCR lenders.

Alternative structures:

• 3-2-1 step-down (lower rate impact, lower prepay cost)
• 5-5-5-5-5 flat (less common; higher rate discount in exchange)
• 3-3-3-3-3 flat (5 years flat 3%)

Why DSCR loans can have 5-year PPPs while QM loans are capped at 3 years: DSCR loans are non-QM and not bound by the Fannie Mae Selling Guide constraints that limit QM prepayment penalties.

PPP buy-down: Most wholesale DSCR lenders allow borrowers to pay an upfront fee at closing to reduce a 5-year PPP to 3-year or eliminate it entirely. Typical cost: 0.5-1.5 points.

PPP rate benefit: Choosing a 5-year PPP vs no PPP typically saves 0.50-1.00% on the note rate per Mo Abdel data. The break-even on accepting the PPP is roughly 18-24 months if rates don't fall enough to trigger a refinance within the PPP window.

The carve-out most borrowers don't know: Most DSCR loans allow a 20% partial prepayment per year of original balance without triggering the PPP. Investors aggressively paying down principal can do so without penalty up to that threshold.

FDIC-insured-bank DSCR loans (CFBank, Quontic, NASB): NASB explicitly publishes no prepayment penalty as a relationship differentiator. CFBank and Quontic PPP schedules are not publicly disclosed and should be obtained in writing before signing.

3.4 Yield Spread Premium (YSP) — the broker conflict

Yield Spread Premium is the broker compensation paid by the lender for placing a borrower at a rate above the direct-to-consumer baseline. Most brokers earn 1.5-3.0% total comp (points + YSP), often blended.

The conflict: Brokers can quietly raise rates to capture YSP. A borrower quoted 7.5% from a broker may be a 7.0% direct-lender rate with 0.5% YSP layered on top, generating an extra $2,500 of broker comp on a $500K loan with no visible cost to the borrower.

The solution: Always request a lender-paid (LPC) vs borrower-paid (BPC) compensation comparison before signing the Lock Agreement. If the broker quote and direct-lender quote differ by ≥25 bps, YSP is being captured.

DSCR loans are particularly YSP-vulnerable because the loan balances ($500K-$3M+) generate larger absolute comp than vanilla mortgages, and brokers raise comp 5-10 bps when volume tiers are hit.

3.5 Lock periods and extension costs

• 15-30 day lock: typically free
• 45-60 day lock: included or +0.125 pts
• 75-90 day lock: +0.25-0.50 pts (extension fee)
• Extension cost: 0.125-0.25 points per week of extension
• Float-down option: Many DSCR lenders offer 30-45 day float-down provisions for +0.125-0.25 pts (lets you re-lock at a lower rate if rates fall during your lock window)

Section 4: Historical rate context

Net read: DSCR rates have declined approximately 100-150 bps over 18 months (peak 2024 → today), but the path has been choppy. The 2021-cycle-low rates of 4.5-5.5% are unlikely to return in a 3-5% Fed funds environment.

Section 5: Forward-looking rate outlook [FORECAST]

5.1 Analyst commentary synthesis

5.2 12-month forecast scenarios (May 2026 → May 2027)

• Base case (~50% probability): DSCR rates oscillate 7.00-8.25% for best-tier scenarios. Inflation moderates into Q3 2026; Fed delivers 1 cut Q4 2026. Spreads tight, supply absorbed.
• Hawkish case (~30% probability): Iran/energy shock persists, core PCE re-accelerates to 3.0%+, Fed delivers 0 cuts or 1 HIKE. DSCR best-tier rises to 7.75-8.75%.
• Dovish case (~20% probability): Energy shock fades, jobs data softens, Fed delivers 2-3 cuts H2 2026. 10-yr UST falls to 3.75-4.00%. DSCR best-tier falls to 6.25-7.00%.

5.3 Confidence statement (honest)

Forward DSCR rates are highly path-dependent on three variables:

1. 10-yr Treasury (every 25 bps move flows ~20-25 bps to DSCR after spread adjustment)
2. Non-QM RMBS AAA spread (currently tight; widening would push rates up faster than UST moves)
3. Fed policy revisions (next FOMC: June 17-18, 2026; new dot plot)

Multiple credible sources point in opposite directions. Anyone claiming a precise 12-month forecast is selling certainty that doesn't exist. The honest answer for borrowers is: rates are likely to oscillate in a 7.0-8.5% band for best-tier borrowers, with low-probability tails on both sides.

Section 6: What aggregators get wrong (7 corrections)

Error #1: "DSCR loans are 7-12% — pick your tier." Most aggregator listicles publish a generic 5-percentage-point band with no scenario context. A 720/65/1.25 borrower paying 11% is being ripped off; an 80%-LTV sub-1.0-DSCR ITIN borrower at 7.5% does not exist outside fantasy pricing.

Error #2: Outdated 2024 rates published as "current." Several first-page Google results still quote a "current range" of 8-10% that hasn't been true since late 2025.

Error #3: Confusing the Visio / HomeAbroad "par rate" (6.12%) with achievable rate. That 6.12% number requires 2-3 points of buy-down + 740+ FICO + 65% LTV + 1.25+ DSCR + 5-year PPP. Net effective cost is closer to 7.0-7.25% when amortized properly.

Error #4: Failing to disclose YSP. Most aggregator pages don't mention that broker comp can be buried in your rate via Yield Spread Premium.

Error #5: Wrong spread assumptions to conventional. The "200-400 bps over conventional" claim is historical, not current. Today's actual spread is 70-120 bps — the tightest in 3+ years. Articles using old spread math overestimate DSCR rates by ~150 bps.

Error #6: Ignoring ARM curve flatness. "ARM is always 100 bps cheaper" is repeated everywhere — currently it's only 50-65 bps cheaper because the swap curve is flat.

Error #7: Treating "rate" and "APR" interchangeably. A 7.0% rate with 2 points and $4,500 in fees has an APR closer to 7.6-7.8%. Always compare APRs across lenders, not headline rates.

Section 7: How to actually get the best DSCR rate

Step 1: Optimize your FICO score above 720 if possible. The FICO-tier breakpoints at 660, 700, 720, and 740 each unlock 25-75 bps of rate improvement. If your current FICO is 715, a 90-day push to 740 can save 50-100 bps.

Step 2: Right-size LTV to 75% or below. The LTV breakpoints at 75% and 80% are the most punitive. Going from 80% to 75% LTV by adding 5% more down payment saves 25-50 bps and unlocks better lender selection.

Step 3: Engineer DSCR ≥1.25 if cash flow permits. Multiple lenders offer best-tier pricing only at 1.25+ DSCR. If your property is at 1.10, adding a larger down payment or restructuring to interest-only can bump DSCR to 1.25+.

Step 4: Get three quotes — wholesale broker, direct lender, FDIC bank. Wholesale brokers can access the deepest pool but carry YSP risk. Direct lenders (Visio, Kiavi via direct channel) eliminate broker comp. FDIC banks (CFBank, NASB, Quontic) often carry 50-100 bps depository premium but offer no prepayment penalty trade-offs.

Step 5: Negotiate the points-for-rate trade explicitly. Ask: "What is this quote at 0 points, 1 point, and 2 points?" Run the break-even math against your expected hold period.

Step 6: Demand the lender-paid-vs-borrower-paid comparison. If the broker refuses to provide it, walk to a different broker.

Step 7: Lock when the 10-year Treasury is in the lower half of its 30-day range. Don't try to time the bottom; just avoid locking on a multi-week high.

FAQ

Related coverage

• Pillar: Best DSCR Loan Lender 2026
• Calculator: DSCR Loan Calculator (interactive)
• Comparison: Visio vs Velocity vs Kiavi DSCR
• Decision guide: DSCR Loan vs Conventional Mortgage 2026
• Refinancing: How to Refinance a DSCR Loan in 2026
• FDIC bank reviews: CFBank · Quontic Bank · NASB · 1st Security Bank · Farm Bureau Bank
• Wholesale lender reviews: Visio Lending · Kiavi · Lima One Capital · LendingOne · Velocity Financial · Constructive Loans · A&D Mortgage · Easy Street Capital