Independent boutique-hotel CFO playbook: GOP, NOI, FF&E reserves.

01 GOP and NOI, separated cleanly

GOP — gross operating profit — is the operating-management number. NOI — net operating income — is the ownership number. USALI 11th Revised Edition (HFTP / AHLA) is the standard that defines both, and the most important single thing to internalise is that USALI 11th formally replaced the line called "Net Operating Income" with "EBITDA." Market participants still say NOI; conceptually they mean EBITDA before reserve for replacement. In independent groups, GOP and NOI almost always sit in the same column of the same spreadsheet, and the FF&E reserve, the self-management fee (even when self-managed), and the property tax accrual all leak from one to the other. The first piece of CFO work in any new engagement at this scale is to rebuild the income statement so that GOP and EBITDA are physically separated, with a named bridge between them.

The USALI walk is mechanical. Total departmental profit plus undistributed operating income (Administrative & General, Information & Telecom, Sales & Marketing, Property Operations & Maintenance, Utilities) less management fees equals GOP. GOP less owner's expense (real estate taxes, property and liability insurance), less rent or ground lease, plus or minus other non-operating items, equals EBITDA. EBITDA is before interest, income taxes, depreciation, and amortisation. That is the line valuation, debt sizing, and cap rate conversations attach to. GOP is the line your operating team should be benchmarked against — STR HOST and CBRE Trends in the Hotel Industry both report GOP margin as the primary operating KPI.

The margin bands to anchor against

For US upscale and upper-upscale full-service hotels in stabilised years, GOP margin (GOP as a percent of total revenue) typically runs 32–38%, per STR HOST and CBRE Trends. For independent urban boutiques, the band is 28–35%, depending on F&B intensity and the local labour market. EBITDA margin sits 6–10 percentage points below GOP after property tax, insurance, and ground rent. So a 34% GOP property is usually a 24–28% EBITDA margin on a fee-simple basis. That spread between GOP and EBITDA is the line that the owner needs to see weekly. Otherwise the operator gets credit for GOP performance while the owner watches the EBITDA margin compress under property tax assessments the operator does not see.

02 The FF&E reserve, sized to actual cycle

The standard branded number — 4% of total revenue into FF&E reserve — is a planning floor that the industry calls a benchmark. HVS has been explicit about this for a decade: 4–5% covers traditional FF&E, but the modern hotel needs an additional 2–3% for technology refresh and 1–3% for major permanent components (building systems, façade, roofing). The true all-in long-term replacement reserve requirement is 7–11% of revenue. An academic study referenced in the Wiley accounting journal in 2010 found FF&E reserves run roughly 40% under-funded relative to long-term needs, and the gap has not closed since.

For boutique groups with high-design rooms, short refresh cycles, and meaningful F&B fit-out, 4% is roughly half of what the actual capex calendar will demand over a seven-year cycle. We see groups that have run on 3.5–4.5% accruals for a decade arrive at a refresh year, look at the reserve, and short-fund the refresh by 30–50%. That short-fund shows up two years later as room rate elasticity that the operator cannot explain.

The cycle reality: soft goods, case goods, heavy renovation

For boutique hotels the reserve should be modelled as a stack of replacement cycles, not a single percentage. Soft goods — carpets, drapery, upholstery, bedding, decorative lighting, wall coverings — wear out on a 4–5 year cycle and wear faster in lifestyle and boutique properties because the guest expects an Instagram-current environment. Case goods — desks, nightstands, wardrobes, millwork, integrated joinery — run on a 7–9 year cycle. Heavy renovation, the PIP-style refresh that hits bathrooms, corridors, public areas, F&B outlets, and the MEP-related lift tied to repositioning, lands on a 12–15 year cycle. The 4% rule averages across these without acknowledging that years three, seven, and twelve are not the same year.

The dollar-per-key benchmarks tighten the picture. The ISHC CapEx Report puts overall capex spending at roughly 9% of revenue and $5,147 per available room on a multi-year average; $6,440 per key excluding the pandemic distortion years. REITs averaged $6,542 per key in 2023. HVS's Hotel Cost Estimating Guide puts upper-upscale guestroom soft-goods renovation at $9,182–$12,551 per key and total construction-plus-FF&E renovation at $12,187–$16,315 per key. CBRE Trends data shows IT department expenditures growing 4.6% from 2023 to 2024 — faster than the 4.0% growth in total hotel expenses — and that gap is structural, not cyclical: every boutique repositioning now pulls "FF&E-like" spend into guest-facing technology, PMS, casting, and Wi-Fi infrastructure that the 4% rule was never designed to absorb.

The math on a 100-key urban boutique makes the gap concrete. Assume $250 ADR, 75% occupancy, $6.8M rooms revenue, and ~$9.0M total revenue. A 4% reserve generates $360k per year — $2.52M cumulative over seven years. A seven-year capex event at a high-design boutique runs $4.7–8.0M ($47–80k per key): $1.2–2.0M for guestroom soft goods, $2.5–4.0M for public area and F&B redo, $0.5–1.0M for technology and back-of-house, $0.5–1.0M for MEP catch-up. The shortfall at 4% is $2.2–5.5M per property. Aggregate across a 5-property portfolio and the seven-year under-funding lands between $11M and $27M before you have sourced a single dollar of refinance or partner capital. That is the number you take to the owner. The 4% is the benchmark for what a third-party manager has to fund; 6.2% is the number that actually services the asset.

03 Rate versus occupancy, weekly, by property

Two paths to the same RevPAR carry very different operating profiles. For boutique groups operating in luxury and upper-upscale, the rate-driven path is almost always the right answer at the level of the operating model — but the occupancy-driven path is what most GM incentive plans reward. That is a CFO problem. The fix is two layers: rebuild the GM incentive so that the rate path is compensated, and report rate and occupancy contribution to RevPAR separately, every Monday, at every property.

The 2024 cohort numbers from Highland Group's Boutique Hotel Report (citing STR / CoStar) anchor where boutique product sits versus traditional hotels. Upper-upscale lifestyle: ~70% occupancy, $227 ADR, $160 RevPAR. Luxury lifestyle: mid-60s% occupancy, $380 ADR, $262 RevPAR. Luxury independent boutiques: ~67% occupancy, $440 ADR, $295 RevPAR. The boutique premium against comparable traditional hotels is 5–20% on ADR, similar to slightly higher on occupancy, and 10–25% on RevPAR. That ADR premium is the entire point of the boutique positioning, and it is fragile — the moment the GM starts trading rate for occupancy the comp-set ranking degrades and the premium collapses.

The decomposition math, and why the flow-through diverges

RevPAR = ADR × Occupancy. RevPAR growth decomposes approximately as gADR + gOccupancy. For a luxury boutique moving from 67% occupancy and $380 ADR to 68% and $399 (a +5% ADR move and a +1.5% occupancy move), RevPAR rises 6.3% and the ADR path accounts for 79% of the move. PwC's read of STR data through August 2025 has luxury chain-scale RevPAR up 5.3% with ADR up 5.0% — luxury and upper-upscale were the only chain scales with positive year-to-date RevPAR growth, and the move was almost entirely rate. STR's 2026 forecast (per Amanda Hite at the 2025 Hotel Data Conference) has overall US RevPAR roughly flat at +0.8%, with luxury and upper-upscale modestly positive on rate. CBRE's 2026 outlook is more constructive at +6% RevPAR, but the directional read is the same: rate does the work, occupancy does not.

The flow-through math is where the CFO discipline pays. Rate-driven RevPAR growth lands at 60–70% flow-through to GOP — the only incremental variable cost is OTA commission, credit card fees, and loyalty cost. Occupancy-driven RevPAR growth lands at 30–40% flow-through because every incremental room sold brings housekeeping labour, laundry, utilities, amenities, and incremental F&B variable cost. On a 200-key luxury boutique, a $0.9M rate-driven rooms revenue lift produces ~$630k of incremental GOP; the same dollar of occupancy-driven revenue produces ~$400k. That is a real number, and it is the basis for rewriting the GM incentive plan around BAR integrity and segment mix instead of headline occupancy.

F&B as a margin signal, not a revenue centre

In boutique groups, F&B is often where the brand experience lives but the margin lies. F&B share of total revenue ranges from 12–15% for lean independents to 20–25% for lifestyle boutiques to 30%+ for soft-brand boutiques with meaningful banquet space. The daypart mix is the diagnostic. Breakfast typically runs 20–25% of F&B revenue at a 10–25% margin — high labour, included rates, and buffets compress profit; it is a guest-experience and ADR-support amenity, not a profit centre. Dinner runs 30–35% of F&B at a 30–40% margin. Bar and lounge runs 15–20% of F&B at 40–50%+ margin — the highest-margin slice. Banquets run 15–25% at 25–35% margin with the labour discipline as the swing variable. We model F&B contribution by daypart at every property and run it weekly; the same hour spent reviewing the dinner cover count and the bar revenue per occupied room tells the operator more about brand health than any monthly P&L summary.

04 The capital stack at boutique scale

Independent groups in the 3-to-20 band almost always have a heterogeneous capital stack: original owner-financed property at one yield, a portfolio-debt refinance at another, an SBA loan on one acquisition, a partner-equity carry on another. We see groups carrying a blended cost of capital in the 10–12% range because nobody has aggregated the stack and looked at it as a stack. The fractional CFO's second-quarter project, after the GOP/NOI rebuild and the reserve sizing, is almost always a refinance plan that takes 200–400 basis points off the blended cost.

Where the senior rate market sits in 2026

Base rates in May 2026: 10-year Treasury (FRED DGS10) is trading around 4.4–4.6%, with a May 21 print of 4.57%. SOFR sits in the high-3s — the Q1 2026 30-day SOFR average per the FSA quarterly schedule is 3.86%. Bank prime is 6.75% (Fed H.15). Stabilised select-service or extended-stay hotels in strong MSAs price at SOFR + 275–375 bps — roughly 6.5–7.5% all-in — at 55–65% LTV. Full-service, lifestyle, and resort hotels with volatility price at SOFR + 325–450 bps — 7.0–8.5% all-in — at 50–60% LTV with underwritten DSCR of 1.40–1.50x. Transitional and value-add boutique deals sit in private credit at SOFR + 500–800 bps — 8.5–11.5% all-in — at up to 70–75% loan-to-cost. Hotel CMBS at Treasury + 250–425 bps clears at 6.8–8.8%, primarily for stabilised branded product; independent boutique CMBS is harder to source and prices wider.

The SBA 504 lever, materially undervalued

For owner-operated boutique hotels, SBA 504 is the single most under-utilised piece of the capital stack. The May 2026 SBA 504 debenture rates are 5.61–5.67% for 10-year, 5.78–5.98% for 20-year, and 5.72–5.95% for 25-year — all-in fixed inclusive of SBA and CDC fees. That is 100–250 bps below conventional CMBS for the same risk. The standard 504 structure is 50% bank first mortgage at conventional hotel rates plus 40% SBA 504 debenture at the rates above plus 10% sponsor equity (often pushed to 15–20% for hotel-specific underwriting). On 18 May 2026 the SBA doubled the combined 7(a) and 504 cap to $10M per borrower — material if you are using SBA across multiple properties in the portfolio. The owner-occupied test is the gating question; in independent boutique groups where the operating company owns the property, the SBA 504 lever almost always survives the test.

Worked stacks for a $10–12M property

Three stacks we model regularly. Stack 1 (clean stabilised, non-recourse): senior bank at 60% LTV / 7.0% + mezz at 15% / 13.0% + common equity 25% / 17% IRR target = blended WACC ~10.4%. Stack 2 (SBA 504 owner-user): bank first 50% / 7.5% + SBA 504 40% / 5.9% + sponsor equity 10% / 18% target = blended WACC ~7.9%. Stack 3 (mixed with seller carry): senior bank 55% / 7.25% + SBA 504 on PIP 15% / 5.9% + seller note 10% / 6.5% interest-only + partner equity 20% / 16% = WACC ~8.7%. The 250–300 bps spread between Stack 1 and Stack 2 — for the same asset, just a different way of putting the capital together — is the prize.

The refinance opportunity windows are specific. Legacy floating-rate loans originated 2022–2023 at SOFR + 450–600 bps are paying 9–11% today; refinancing those to SBA 504 for owner-occupied properties, or repricing to SOFR + 250–350 bps for stabilised assets, takes meaningful basis points off the blended stack. CMBS originated 2016–2018 with balloon maturities coming into 2026–2028 should be refinanced for balloon de-risking even if the coupon does not drop — the gating question is whether the new LTV proceeds clear the existing balance plus transaction cost. Mezz-heavy stacks (12–15% all-in) are routinely cleaned up by replacing mezz with seller carry (6–8%) or incremental senior. The case where we tell groups to do nothing: long-term fixed CMBS or life-co at 4–5% from the 2018–2021 origination window. Hold that paper; do not refinance for rate. Add SBA 504 layered behind it only for new PIP or expansion eligible under the program.

05 What the lodging price index is telling you

For CFOs running a boutique group, the BLS Producer Price Index for Traveler Accommodation (FRED series PCU72117211) is the single best macro check on whether the rate-driven RevPAR thesis is real or imagined. The series is the wholesale-output price index for NAICS 7211 — the lodging industry's revenue line measured on the BLS schedule. The April 2026 index reading was 207.1, against 197.2 in April 2025 — a 5.0% year-over-year lift. The two-year comparable (April 2024 at 195.8) is +5.8%. That is the rate path showing up in the official price index, and it lines up with PwC's reading of STR luxury ADR through August 2025 at +5.0%. The labour side closes the loop: BLS accommodation industry employment (FRED CES7072100001) was 1,919K in April 2026, essentially flat against 1,923K in June 2024 and 1,910K in June 2023. The lodging industry is pricing up, not staffing up. The rate-driven thesis is structural, and the operating model that does not capture the rate-flow-through gets left on the table.

The implication for boutique groups planning into 2027: do not assume occupancy recovers to pre-pandemic peak. Build the operating model around the rate path, build the GM incentive around BAR integrity and segment mix, and protect the F&B price ladder against discount pressure during shoulder seasons. The macro is on the side of the operator who holds rate; it is against the operator who chases occupancy.

06 Three questions for this quarter

The CFO work breaks down to three questions you can take to the next operating review. None of them need a consultant; they all need a CFO sitting in the operating seat for one quarter.

1. What is the FF&E reserve as a percent of revenue across the portfolio, what does the actual seven-year capex calendar require by property, and where does the cumulative under-funding sit today in dollars? The number you walk into the owner conversation with is the under-funding line, not the reserve percentage.
2. How is RevPAR uplift attributed in the GM incentive plan — rate, occupancy, or blended — and is that the path you would want the GMs to choose? The 60–70% rate flow-through vs 30–40% occupancy flow-through math should be in every GM's comp letter.
3. What is the blended cost of capital across the portfolio, when was the stack last looked at as a stack rather than as a series of loans, and what is the SBA 504 capacity that has not been deployed? A 200–400 bps compression on portfolio WACC is the second-largest line a fractional CFO produces in the first year.