How to compare commercial real estate lenders

What a lender comparison is and why six to twelve is the right number

A bake-off is a structured competitive bid process: a single deal package goes simultaneously to a curated list of lenders, who each respond with an indicative letter of interest within a defined window. The placement firm normalizes the responses, presents the borrower with a side-by-side comparison, and runs a final negotiation with the top two or three.

Six to twelve lenders is the working range. Fewer than six and the borrower is not getting full pricing discovery, there is statistically too much variance in any one lender's pricing on any given week. More than twelve and the deal starts to feel shopped; lenders learn the deal is going wide and price defensively or pass on the opportunity.

The selection of which six-to-twelve is the work. Random outreach to a long lender list wastes the borrower's time and the firm's reputation. The right list is built around the deal's specific characteristics: property type, sponsor profile, loan size, leverage, business plan, and timeline.

Step 1: translate the deal into a lender-fit profile

Property type drives capital-source eligibility. Stabilized multifamily over $5M opens the agency bench (Fannie DUS, Freddie Optigo) plus life cos and CMBS. Industrial stabilized opens life co and CMBS heavily, with bank perm a third pillar. Hospitality narrows to specialty hotel debt funds, CMBS, and a thin life-co bench. Office in the 2026 cycle has dramatically thinned the eligible bench, only specific lenders are open at all.

Leverage and business plan drive the bridge vs. perm split. Stabilized at 60% to 70% leverage is a perm conversation; stabilized at 75%+ or non-stabilized at any leverage is a bridge conversation, and the lender list changes completely.

Sponsor profile filters the bench further: first-time CRE sponsors with limited net worth need lenders willing to underwrite to property cash flow with a key principal designation; institutional sponsors with $50M+ liquidity get a wider bench and tighter pricing.

Step 2: build the lender list

From the eligible-capital-source list, the placement firm's job is to identify the specific lenders within each category that are actively quoting the deal type today. Lender appetite shifts quarter to quarter, a regional bank that was actively quoting industrial in Q1 may have hit its CRE concentration limit by Q3 and pulled back. The Federal Reserve's Senior Loan Officer Opinion Survey (SLOOS) tracks this at an aggregate level; current desk-level intelligence comes from active conversations with each lender.

A typical bake-off list for stabilized $20M multifamily refi might be: three agency lenders (one Fannie DUS, two Freddie Optigo), two life cos, two CMBS conduits, one regional bank, and one debt fund as a high-leverage cross-bid. Total nine lenders.

A typical bake-off list for $30M value-add multifamily bridge might be: four debt funds (varying in their leverage and pricing positioning), two banks (relationship-eligible), one mortgage REIT, and one specialty multifamily bridge platform. Total eight lenders.

Step 3: package the deal once, send simultaneously

The deal package, financing memorandum, T-12, current rent roll, sources and uses, sponsor bio and SREO, supporting exhibits, is built once to institutional standard and sent to all selected lenders the same day with the same submission deadline (typically 7 to 14 days for indicative LOI).

Q&A is centralized. Lender questions come to the placement firm, not the borrower. The firm responds in writing to all lenders simultaneously where the answer is generic, or one-to-one where the answer is lender-specific. This prevents one lender from getting an information advantage and prevents the borrower from being on twelve separate phone calls.

Step 4: normalize the LOIs into apples-to-apples comparison

Lenders write LOIs in their own format with their own terminology. The placement firm's normalization sheet puts every LOI on a single page with the same line items: loan amount, all-in rate (calculated where the LOI quotes spread plus index), loan term, amortization, IO term, prepayment menu, recourse, key reserves, origination fee, exit fee (if any), expected closing timeline, and any unusual conditions.

All-in cost calculation: the headline rate is misleading without exit fees, origination fees, lender legal cap, mandatory reserves, and the cost of the prepayment-flexibility delta. A 10-bps wider LOI with a one-year-shorter prepay lockout might be the better deal on a 5-year hold. A 20-bps wider life-co LOI without an exit fee might beat a tighter CMBS LOI with defeasance over a 5-year hold.

Certainty of close is not a quantifiable line item but it belongs on the comparison. A lender's track record on retrade, the specificity of their LOI conditions, and the tightness of their MAC language are all signals.

Step 5: pick the winner on the combination, not the headline

The cheapest LOI is not always the right one. The right LOI is the one where the combined economics, structure, and execution risk best fit the sponsor's business plan and risk tolerance. A 25-bps tighter LOI from a debt fund with a soft commitment and aggressive MAC language is a worse outcome than a 25-bps wider LOI from a bank with a tight MAC and a track record of closing.

Common selection patterns: institutional sponsors who plan a refi at year three or four often pick the LOI with the cheapest break-cost rather than the lowest headline rate. First-time sponsors often pick the lender with the tightest closing-conditions list. Heavy value-add sponsors often pick the lender with the most sponsor-friendly construction-draw mechanics rather than the cheapest pricing.

Step 6: run the final negotiation and lock the deal

Once the borrower selects a winner, the placement firm goes back to that lender for a final round of negotiation: tighten the MAC, narrow the springing-recourse triggers, lower the exit fee, push for IO. The credible threat that the runner-up LOI is sitting in the file is what produces the final round of concessions.

Some borrowers leave a backup LOI hot (deposit paid, term sheet signed conditional) until rate-lock or commitment on the winning lender. The cost of holding two LOIs hot is the second deposit; the benefit is preserved optionality if the winner retrades during diligence.

When one LOI is enough

There are situations where a parallel bake-off is not the right tool. Repeat business with an active relationship lender on a clean stabilized refi where the lender's pricing is known to be competitive, quoting a single LOI from that lender saves the borrower's time and the firm's reputation with the lender options.

Sub-$3M deals where the marginal pricing improvement from a parallel process does not justify the friction cost. The borrower's existing bank relationship is usually the answer.

Highly specialized property types or situations where the eligible lender pool is so small (e.g., three lenders nationally) that running a process is the same as direct outreach. Even then, three quotes is better than one.

Syndicated and club deals: when the lender comparison changes shape

On loans large enough to syndicate (typically $100M+) or in capital structures that need a club (large bank lines, hospitality construction, complex multi-property), the bake-off shifts from competing complete LOIs to selecting a lead lender who will arrange the syndication or club. The lead's relationship with the participating bank network becomes the differentiator more than the lead's own pricing.

The placement firm's role shifts to evaluating the lead's syndication track record, the proposed economics across the syndicate, the inter-lender voting and amendment mechanics, and the agent role and fees. The deal points still matter; the bake-off process is structurally different.

Lender types and financing options

Agency multifamily (Fannie DUS and Freddie Optigo): the deepest perm bench for stabilized multifamily over $1M. Pricing is highly transparent because the index spread is published in lender pricing sheets. Bake-off competition primarily moves the structural lines (IO term, supplemental availability, springing-recourse triggers) more than the spread.

Life cos: roughly 40 active life-co lenders quote CRE perm at meaningful volume. Each has a defined property-type and geographic appetite, a defined leverage band, and a defined deal-size sweet spot. The right bake-off list selects three to five life cos that fit the deal profile, not all 40.

Banks: regional and money-center banks quote both relationship-driven and competitive perm. Bake-off positioning differs: relationship banks expect to win on the relationship; non-relationship banks compete on price. Mixing the two on the same bake-off requires careful management, relationship banks dislike learning they were used as a price benchmark.

Debt funds: the bridge market is fragmented across roughly 200 active debt funds with widely varying capital structures, leverage capacity, and pricing discipline. Bake-off curation is hardest here because the lender list has to balance price-aggressive funds against execution-reliable ones.

HUD: the FHA 223(f) and 221(d)(4) programs are deep but slow. HUD is its own bake-off category, typically two to three lenders quoting against the same MAP-lender approval, and the timeline is structurally 6 to 9 months from package to closing.

Bridge lender comparisons vs. perm lender comparisons: process differences

Timeline: bridge bake-offs typically run 4 to 6 weeks from package send to signed term sheet; perm bake-offs run 6 to 8 weeks. Bridge LOIs are typically less specific than perm LOIs (the diligence path is heavier) so the soft-commitment risk is higher and the bake-off normalization is more nuanced.

Selection criteria: on perm, certainty of close and structural fit usually outweigh modest pricing differences. On bridge, lender flexibility on draw mechanics, cash-control language, and modification rights often matters more than pricing because the borrower will live with the loan through a value-add cycle.

Reading lender appetite signals before going wide

Pre-bake-off intelligence: before the package goes out, the placement firm typically calls each prospective lender to confirm appetite, current allocation availability, recent comparable closings, current pricing benchmarks. Lenders that haven't quoted the property type recently are usually polite passes; including them in the bake-off wastes their time and the firm's reputation.

Allocation signals: agencies publish quarterly volume caps and YTD volume; tracking shows where each agency is relative to its annual cap. Life cos signal allocation availability through their CRE allocation officers. Banks signal through their CRE concentration ratios as reported to regulators (FDIC Quarterly Banking Profile is the public reference).

Pricing signals: recent comparable closings (closed within the last 60 days on the same property type, geography, and capital source) are the most reliable pricing benchmark. Stale comps (closed 6+ months ago) are misleading because spread environments move.

Decision-maker continuity: lenders with high turnover at the originator level have higher bake-off variance. The deal champion who pitched the deal to credit committee is the deal champion who signs the term sheet; if that person changes mid-process, the deal often resets.

Managing relationship lenders inside a competitive process

Relationship lenders, the bank that holds the sponsor's operating account, the life co that wrote the prior loan on the same property, the agency lender that has financed three of the sponsor's prior deals, have legitimate claim to first look. Excluding them from the bake-off creates relationship friction; inviting them carelessly often leads to a soft quote that is uncompetitive on the comparison sheet.

Standard practice: include the relationship lender in the bake-off with a specific note that the bake-off is competitive and that the relationship lender is expected to bid sharpest. Most relationship lenders accept this because the alternative (no shot at the deal) is worse. The bake-off process produces a competitive quote without burning the relationship.

When the relationship lender quotes uncompetitively: take the better LOI from a different lender. Most relationship lenders accept losing a competitive bake-off; the relationship survives. Repeat losses are a signal that the relationship has structural pricing problems and the sponsor should re-evaluate where they keep their relationship business.

When the relationship lender wants exclusivity: rare and rarely worth granting unless the relationship lender's quote is materially superior to the expected market and the sponsor can document this from prior bake-offs. Granting exclusivity foregoes price discovery; sponsors who routinely grant exclusivity tend to underperform on financing economics.

What 'apples-to-apples' means line-by-line

Loan amount: convert all-in proceeds to net-of-reserves cash to the borrower. A lender quoting $20M with $1M of upfront reserves delivers $19M of cash; a lender quoting $19.5M with $200K of upfront reserves delivers $19.3M of cash. The headline loan amount is misleading without the reserve profile.

All-in rate: convert spread plus index to a comparable date and index. CMBS quoted as a spread over swaps must be converted to a spread over Treasury for comparison with life-co perm; SOFR-based bridge must be converted to a fixed-rate equivalent at the forward curve.

Term: extension options at borrower's discretion (subject to defined tests) extend the effective term meaningfully. Two one-year extensions on a five-year initial term effectively make the loan a seven-year loan if the borrower controls the exercise.

Recourse: full vs. partial vs. non-recourse should be compared on equivalent dollar exposure. A partial-recourse loan at 25% recourse on $20M is equivalent dollar exposure to a $5M personal guaranty; the LOI comparison should reflect this.

Model prepayment charges on the balance and date specified in the loan documents. Include them in the projected cash flows over the expected holding period.

Reserves: monthly reserve funding is a working-capital drag. Convert monthly reserve funding to bps of all-in rate (the carry cost on the trapped cash) and add to the comparison.

Confidentiality, NDAs, and how to keep a deal from being shopped

The dual-signaling problem: the placement firm wants the lender options to know the deal is well-shopped (so the bake-off is taken seriously), but the broader market should not know the deal is being shopped (because that signals weakness and depresses pricing). The discipline of running bake-offs through a curated, identified list, never through a broad email blast, is the practical defense against signaling weakness.

NDAs: standard institutional practice on competitive deals. Each lender on the bake-off list signs an NDA before receiving the package. Lenders that decline to sign are off the list. The NDA covers the package contents and the existence of the bake-off; it does not constrain the lender's ability to underwrite the deal internally.

Information control: the package is delivered through a secure data room, not as an email attachment. Q&A is centralized through the placement firm; lenders do not have direct access to the borrower or to property management. This both protects confidentiality and prevents one lender from getting an information advantage.

Post-bake-off control: when the borrower selects a winner, the runner-up lenders are notified that they did not win without disclosing who did. The package is removed from the data room. This preserves the option to re-engage runners-up if the winner falls through.

What happens when the lender comparison produces a poor field

Poor field defined: every LOI back is materially wider than expected, the structural terms are uncompetitive, or the certainty signals are weak across the field. This sometimes happens on property types under capital-source stress (office in many submarkets in 2026), on sponsors with limited track record, or on deals with structural complexity that lenders dislike (highly leveraged, partial cash-out, complex business plan).

Diagnostic step: a poor field on a clean deal often signals a packaging problem. Rebuild the package with a sharper financing memo, a cleaner T-12 normalization, and a better-documented sponsor profile, then re-run the bake-off with the same lender options plus two or three replacements. Often produces materially better LOIs.

Hold-and-rerun: in some cases, holding the deal for 60 to 90 days and re-running the bake-off when market conditions shift produces better results than forcing a closing in a cold market. The opportunity cost of the delay should be weighed against the pricing improvement.

Lender comparison mistakes sponsors make on their own

Mistake 1, the email blast. Sending the deal to 20+ lenders simultaneously without curation signals indiscriminate shopping. Lenders that receive blast emails frequently typically pass without quoting; the bake-off field comes back thin and unrepresentative.

Mistake 2, the wrong submission window. Submission windows under 7 days produce rushed quotes; windows over 14 days lose lender attention and the deal goes stale. Seven to ten business days is the working range for indicative LOI on a clean perm; 10 to 14 for bridge.

Mistake 3, answering lender questions one-on-one without centralizing. Lenders that get different answers to the same question lose confidence in the package; lenders that learn other lenders are getting better information opt out of the bake-off entirely.

Mistake 4, over-disclosure of competing LOIs. Telling each lender exactly where the competing quotes sit produces convergence to the lowest acceptable quote rather than discovery of the best quote. The placement firm's discipline on what to disclose and to whom is a meaningful source of pricing improvement.

Mistake 5, selecting on headline rate without normalization. The all-in cost calculation (rate plus fees plus reserves plus prepay-flexibility delta) often inverts the headline-rate ranking. Sponsors that select on headline often discover at closing that the chosen LOI was not the best deal.

Mistake 6, failing to keep a backup LOI hot until the winner is rate-locked or commitment-issued. The cost is one extra deposit; the benefit is preserved optionality if the winning lender retrades during diligence.

Closing the lender comparison: the final-round negotiation

Once the borrower selects the winning LOI, the placement firm runs a final round of negotiation to lock in incremental concessions. The credible threat that the runner-up LOI is sitting in the file is what produces this final round of improvement. Typical concessions extracted in the final round: 5 to 15 bps of rate improvement, MAC tightening, exit-fee reduction or waiver, springing-recourse trigger narrowing, IO-term extension.

Discipline matters: the final-round ask should be focused on three to five high-value items, not 20 small items. Lenders that get a 20-item ask after term-sheet selection often reset to their original term-sheet position; lenders that get a focused ask typically respond constructively.

Documentation: every final-round concession should be reflected in writing in the term sheet or in a side letter signed by both parties. Verbal concessions made in the final-round negotiation that do not get into writing typically do not survive into the loan documents.

Runner-up management: when the winner is locked, notify the runners-up promptly that they did not win without disclosing who did. Most runners-up appreciate the courtesy and remain available for future bake-offs. Runners-up that are kept hot too long without being told they lost lose trust in the placement firm and quote less aggressively on future bake-offs.

Frequently asked questions