The number nobody wants to publish
A four-bedroom property in Tulum generates $280,000 in gross booking revenue. The owner’s pro forma — built by the previous manager — projected $185,000 in net operating income after a 20% management fee and “miscellaneous expenses.”
The actual net: $118,000.
That $67,000 gap wasn’t theft or mismanagement. It was the cost of every line item the pro forma didn’t include. Turnover labor. Linen replacement. Platform commissions calculated against the gross, not the net. A pool pump that failed in August. The CapEx reserve that didn’t exist until the outdoor furniture disintegrated eighteen months in.
This is the pattern we see repeatedly across our operations in Miami, Tulum, and Houston. Gross revenue is visible. The expense architecture required to produce it is not — and the gap between what investors model and what the property actually costs to operate runs 15–20 percentage points.
In The Yield Gap, we documented the conditions under which short-term rentals outperform traditional leases. That analysis assumed competent expense management. This piece examines what competent expense management actually costs.
Fixed costs: the number that doesn’t care about your occupancy
Fixed costs are the floor. They run whether the calendar shows 95% occupancy or sits empty for a month during shoulder season.
Property taxes and insurance
This is where market selection creates the widest variance in the entire expense stack.
A $1.2 million property in Houston’s Inner Loop carries an annual property tax burden of $26,000–$34,000 — roughly 2.2–2.8% of assessed value. Texas has no state income tax; it makes up the difference in property tax. For a property generating $200,000 in gross revenue, that single line item consumes 13–17% before anything else is paid.
The same asset value in Tulum faces a predial tax under $1,000. But the fideicomiso — the bank trust required for foreign ownership in Mexico’s restricted zone — adds $1,500–$3,000 annually in trustee fees. Legal compliance, RFC registration, and SAT reporting introduce costs that don’t appear in any template financial model built for the U.S. market.
Miami sits between. Property tax runs lower than Houston on a percentage basis, but insurance is the outlier. STR policies in South Florida — factoring hurricane exposure, flood zone premiums, and the liability profile of short-stay hospitality — cost 40–70% more than equivalent long-term rental coverage. A property that costs $4,800 to insure as a long-term rental may cost $7,500–$8,200 as a short-term rental.
Licensing and compliance
Miami-Dade requires STR registration, resort tax collection (6% on top of the 13% state and county transient taxes), and periodic renewals with inspection requirements. Non-compliance carries fines that can exceed the annual licensing cost by an order of magnitude. We detail the full Miami property management compliance architecture — including building-level zoning distinctions — in our market overview.
Houston’s framework is lighter — no STR-specific licensing requirement at the city level as of this writing — but HOA restrictions, deed restrictions, and evolving municipal code create compliance costs that are less visible and harder to forecast.
Tulum requires municipal permits, tax registration with SAT, and compliance with evolving Quintana Roo state regulations that have tightened significantly since 2023. Foreign owners operating without proper RFC and fiscal domicile documentation face withholding penalties that erode returns faster than any operational expense.
Budget 1–3% of gross revenue for regulatory compliance across markets. The variance reflects complexity, not optionality. These are not costs you choose to pay.
Management fee
Full-service management for a property at this tier runs 15–25% of gross revenue. The range reflects market norms, service scope, and property complexity.
This is the most visible expense line — and the one that most often gets confused for the total cost of operations. It isn’t. The management fee covers the operator’s team, systems, guest communication, and coordination overhead. It does not typically include the direct costs the operator is coordinating: cleaning crews, maintenance vendors, supplies, platform commissions, or capital expenditures.
When an investor sees a pro forma with “20% management fee” as the only operating expense below the gross revenue line, the model is missing 30–45 additional percentage points of cost.
Utilities and connectivity
High-speed internet, smart locks, security cameras, climate control between bookings, landscape irrigation, pool circulation. These systems run continuously. A four-bedroom property with a pool in a warm climate draws $400–$700 per month in base utilities regardless of whether anyone is sleeping there.
Fixed cost total: 25–35% of gross revenue. Houston sits at the high end due to property tax. Tulum’s fixed costs are lower in absolute terms but carry more administrative friction. Miami’s insurance burden is the distinguishing factor.
Variable costs: the margin compression nobody models
Here’s the structural difference between a long-term rental and a short-term rental that most financial models fail to capture: variable costs in a long-term rental are approximately zero. The tenant pays utilities, handles their own cleaning, and buys their own toilet paper. The owner’s cost doesn’t change whether the tenant is home or traveling.
In a short-term rental, every booking generates direct expense. And those expenses don’t scale linearly — they compress margins at precisely the moments when revenue is highest.
Turnover
This is the single largest variable cost and the one that separates hospitality operations from residential property management.
A turnover at this tier isn’t a cleaning. It’s a 4–6 hour process: deep clean of every surface, linen strip and replacement, kitchen audit and restock, bathroom amenity reset, outdoor furniture wipe-down, pool check, technology verification (WiFi, smart TV, locks, thermostats), damage inspection, and photo documentation for the property condition log.
A three-bedroom property in Miami runs $350–$450 per turn. A five-bedroom in Houston with a pool house: $500–$650. Tulum properties with outdoor living spaces or rooftop areas push above $500 due to the additional square footage and tropical debris management.
At 80–100 turns per year, this single line item costs $28,000–$65,000 annually. For a property grossing $250,000, that’s 11–26% of revenue — from cleaning alone.
The compression effect matters: peak season produces the most bookings, the shortest gaps between guests, and the highest turnover frequency. July in Tulum might require 12–15 turns. That’s $6,000–$7,500 in a single month — precisely the month the owner is watching gross revenue spike and assuming the margin is expanding.
It isn’t.
Guest supplies and consumables
At the rate tier where these properties compete — against boutique hotels and members-only travel clubs — consumable quality is a rate-integrity cost. Generic soap and a Keurig signal a price point. Aesop amenities and a Breville espresso machine signal a different one.
Coffee, welcome provisions, cooking oils, cleaning supplies for guest use, pool chemicals, toilet paper, paper towels, garbage bags, dishwasher pods. The list is mundane. The cost is not: $40–$80 per booking, scaling with guest count and length of stay.
Annual spend for a well-stocked property running 80+ bookings: $3,200–$6,400.
Laundry and linen service
Separate from the turnover clean. Commercial laundering of king sheets, duvet covers, pillow protectors, bath towels, hand towels, pool towels, and robes. Properties with pools and outdoor showers run heavier loads.
Two options: own the linens and pay for commercial laundering ($2.50–$4.00 per pound), or use a linen rental service with per-turn pricing. Either way, the annual cost for a 3–5 bedroom property runs $5,000–$14,000.
Linen replacement is a separate line item that falls under CapEx — a set of hotel-grade king sheets lasts 80–120 wash cycles before it needs to be retired, not because it’s torn but because the hand degrades below the standard guests at this tier expect.
Dynamic pricing tools
Revenue management software — PriceLabs, Wheelhouse, Beyond — runs $300–$1,200 annually in subscription costs. The tool itself is cheap. What’s not cheap is the operational intelligence required to calibrate it: setting floor rates, event-driven overrides, length-of-stay adjustments, last-minute discount thresholds, and orphan-day strategies.
The software provides the mechanism. The operator provides the judgment. Both are costs — one shows up on a subscription invoice, the other is embedded in the management fee.
Variable cost total: 14–22% of gross revenue. The range is driven primarily by turnover frequency and property size. Higher occupancy means higher variable costs — a relationship that every yield projection should model explicitly and almost none do.
Platform fees: the silent partner
Airbnb charges hosts 14–16% of the booking subtotal on its host-only fee model. Vrbo charges 5–8% to hosts with additional guest-side fees that affect conversion. Direct bookings eliminate the platform commission but introduce website maintenance, booking engine software, payment processing (2.5–3.5%), and the marketing spend required to drive traffic.
For properties in our portfolio, the blended platform cost — across Airbnb, Vrbo, and direct channels — runs 10–15% of gross revenue. Properties with established direct booking programs and repeat guest relationships push toward 8–10%. Properties fully dependent on platforms sit at 14–16%.
The economics of channel diversification are real but slow. Building a direct booking channel takes 18–36 months of consistent guest experience, email capture, and follow-up. During that transition period, the platform fee is a fixed cost of distribution, not a variable you can optimize away with a better website.
Platform fee total: 10–15% of gross revenue.
CapEx reserve: the line item that determines whether your yield is real
This is where most pro formas quietly lie.
A long-term rental tenant uses one set of sheets, one set of towels, one kitchen’s worth of equipment — and replaces them on their own schedule with their own money. A short-term rental guest uses all of those things, and so do the next 80–100 guests that year. Every fixture, surface, and appliance depreciates on a hospitality timeline, not a residential one.
An investor who doesn’t reserve for this isn’t earning 18% net yield. They’re earning 12% net yield and consuming 6% of the asset’s physical condition to bridge the gap. The returns look strong for two years. Then the property needs $40,000–$60,000 in refreshment, the reserve doesn’t exist, and the owner funds it from outside the operation — retroactively converting what looked like income into deferred maintenance.
The replacement cycle
Year 1–2 (annual). Linens, towels, pillows, kitchen smalls (can openers, wine glasses, cutting boards), welcome baskets, consumable décor items. These are the components guests touch most frequently. Cost: 2–3% of gross revenue annually.
Year 2–4 (mid-cycle). Mattresses, sofa cushions, accent chairs, bathroom hardware, small appliances (coffee machines, blenders, toasters), exterior paint touch-up, grout and sealant maintenance. Cost: 3–5% of gross revenue annually, or a single allocation of 6–10% at the midpoint.
Year 4–7 (major refresh). Full furniture replacement in primary living areas, flooring in high-traffic zones, fixture upgrades, outdoor furniture, pool equipment, HVAC service or replacement, technology refresh (smart TVs, locks, thermostats). Cost: 10–18% of a single year’s gross revenue, amortized across the cycle.
Climate as a cost multiplier
In Tulum, the replacement cycle compresses by 30–40%. Salt air corrodes metal hardware. Humidity warps wood furnishings and degrades mattresses faster. UV exposure bleaches and weakens outdoor textiles. Outdoor furniture that lasts five years in Houston lasts eighteen months on a Tulum rooftop terrace.
A Tulum CapEx reserve should sit at 10–12% of gross revenue. An investor underwriting a Tulum property at the Houston CapEx rate will face a capital call within three years.
In Houston, the cycle is more predictable but the tail risk is concentrated. A single hailstorm can trigger $25,000–$50,000 in unplanned repairs. Insurance covers structural damage, but deductibles ($5,000–$15,000 on a property of this value), claim-related premium increases, and uninsured losses — landscaping, pool equipment, exterior cosmetics, patio furniture — are real costs that don’t appear in actuarial models of operating expense.
Miami combines both profiles: hurricane exposure creates Houston-style event risk while coastal humidity and salt air create Tulum-style accelerated degradation.
An investor who asks their manager “what’s the CapEx reserve?” and gets a blank look has the most important piece of information they need about that operation.
CapEx reserve total: 8–12% of gross revenue, set aside annually regardless of whether it’s deployed that year. The reserve is not a cost — it’s a recognition that the cost already exists and will present itself on its own timeline.
The full stack
Adding all four categories:
Fixed costs, 25–35%. Variable costs, 14–22%. Platform fees, 10–15%. CapEx reserve, 8–12%.
Total: 57–84% of gross revenue.
The 40–55% figure we referenced in The Yield Gap represents the operational core — properties with optimized channel mix, disciplined CapEx planning, and favorable tax positioning. It’s achievable. It is not the default.
The default — a property on platform-only distribution, in a high-tax jurisdiction, with tropical climate exposure and no CapEx reserve — operates closer to 65–75%. The owner sees a gross revenue number that implies a 15% yield and nets an 8% yield after the expense stack they didn’t model.
That 8% may still outperform a traditional lease. But it’s a different conversation than the one the pro forma started.
Three rules for underwriting STR expenses
Build from the bottom, not the top. Start with the expense stack. Populate every category with market-specific assumptions. Subtract the total from gross revenue. The number that remains is the yield you can actually expect — and the number you should use to decide whether the short-term rental strategy outperforms a long-term lease for this specific asset.
Require a CapEx schedule. Any management company that doesn’t maintain a rolling replacement calendar and a funded reserve account is deferring costs into the future and reporting them as current-year returns. Ask to see the schedule. If it doesn’t exist, you’ve identified the gap between the projected yield and the real one.
Stress-test the variable stack at peak occupancy. The instinct is to model the upside: “If we hit 85% occupancy, revenue reaches $X.” Model the cost side of that scenario with equal rigor. Each additional booking adds $300–$600 in direct variable cost. The marginal contribution of the 90th booked night is materially lower than the 50th. Peak revenue does not mean peak margin — and the model should reflect that.
The yield gap between short-term and long-term rentals exists. We’ve documented it. But it exists on the other side of every cost in this breakdown. Investors who underwrite from the gross revenue line find the gap on paper. Investors who underwrite from the expense stack capture it in practice.
