The 28-Day Cliff: The Deposit Problem No Mid-Term Rental Software Solves
There is a technical fact about security deposits that determines whether a mid-term rental operation is protected or exposed, and it is almost never discussed in property management software marketing.
Credit card pre-authorisation holds expire after 28 days.
This is not a platform policy. It is not something Stripe, Airbnb, or any PMS vendor chose. It is a network-level limit set by Visa and Mastercard, and it applies to every operator taking deposits by card hold, everywhere.
Which means that for the entire mid-term rental category — every stay of 30 days or longer, by definition — the deposit expires before the tenancy does. Silently. With no notification to either party.
What the cliff actually looks like
Consider the arithmetic on a standard corporate housing stay.
Average corporate housing tenancy: 90 days. Hold placed on day 0, expires automatically on day 28. Days of the stay with zero deposit protection: 62.
That is 69% of the tenancy running uncovered, during the period when risk is highest.
The risk profile of a furnished rental does not stay flat over a stay — it climbs. Wear and tear accumulates faster in furnished units than unfurnished ones. Unauthorised subletting tends to appear in weeks five through eight, when a corporate client moves in a partner. Property condition disputes get harder to resolve as time passes, because damage could plausibly have occurred at any point across several weeks. Utility overages and noise complaints surface late.
The deposit is supposed to cover all of that. The card network says it covers 28 days.
The Amsterdam case
A corporate housing operator in Amsterdam rented a serviced apartment to a consulting firm for 90 days. Damage deposit: €1,500, collected as a card pre-authorisation on day 0. Standard practice, professionally executed.
On day 29, the hold expired automatically. Neither party received a notification. Nothing in the operator's dashboard changed.
On day 78, the guest caused €2,400 in damage.
There was no enforceable hold, no captured deposit, and no practical recourse. Small claims court would have cost more than the damages were worth. The operator absorbed the loss entirely.
This is not an edge case or a cautionary hypothetical. It is a routine failure mode in an industry where most operators do not know the cliff exists.
The three architectures that actually work
There are exactly three approaches that hold up past day 28, and each fits a different operating context.
Rolling re-authorisation. Authorise €1,500 on day 0. On day 25, authorise a fresh €1,500 and release the first. Repeat on a 25-day cycle for the length of the stay. This works cleanly but requires the tenant to have enough available credit to cover a three-day overlap window, and it requires a deposit system with automated re-authorisation scheduling. Most PMS platforms do not have one.
Captured and refunded. Actually charge the deposit on day 0 and refund it at the end of the stay, minus any deductions. This is the default in SEPA, iDEAL and bank-transfer markets — the Netherlands, Germany, Belgium — and for corporate clients generally. The accounting requirement here is strict: a captured deposit is a liability, not income, and must sit on the books as something like "Tenant Deposits Payable." Operators whose systems record it as rent received end up receiving tax-collected-on-deposits notices from their state revenue department.
B2B deposit invoice with a portal. A formal deposit invoice delivered through a customer portal with visible refund status. Cleanest option for operators serving corporate clients and relocation companies, and the portal doubles as the dispute trail if part of the deposit is withheld.
Airbnb's pre-authorisation model works correctly for stays under 30 days and fails silently over that line. Furnished Finder has no platform-mediated deposit handling at all — rent moves by ACH or wire and the deposit is the operator's problem.
The accounting layer underneath
The deposit lifecycle has four distinct accounting states, and most systems handle at most two of them correctly:
- Pre-auth hold — not on the books at all. The tenant's bank has reserved funds; the operator's ledger shows nothing. A contingent position, not an asset.
- Captured deposit — a current liability. Not income. This distinction is where most systems break.
- Refunded deposit — zero balance, zero P&L impact.
- Applied deposit — the deducted portion becomes income or offsets an expense and is taxable; the remainder is refunded. DoorLoop and AppFolio have trust accounting modules that handle bank-held deposits competently. Very few handle the captured-card-deposit-then-applied-to-damage flow as a discrete event with correct general ledger treatment.
For operators working across borders the divergence widens further. EU mid-term operators run captured-and-refunded by bank transfer as standard; US operators run card holds. A US operator serving European corporate relocations is running two incompatible deposit architectures simultaneously, usually without a system that knows the difference.
The reconciliation tax on top
The deposit cliff is one symptom of a broader pattern: mid-term rentals inherited software built for either nightly stays or annual leases, and live in the gap between them.
The clearest measure of that gap is time. Connecting Yardi or AppFolio to QuickBooks eliminates a manual export-import cycle that costs multi-property accounting teams 8 to 15 hours per month in duplicate entry and reconciliation. For a 30-unit operator, that is $400 to $1,500 per month in bookkeeper time, every month, indefinitely.
The AppFolio-QuickBooks integration exists but is one-way and incomplete. It does not sync owner draws, trust account movements, vendor 1099s, custom GL accounts specific to furnished rentals, mid-stay partial refunds, or security deposit interest accrual in states that require it. Books running on both systems diverge within 30 to 60 days, at which point reconciliation takes longer than the original data entry would have.
The structural problem is sizing. Yardi is enterprise software — typically 500+ units, $30,000+ per year, implementation consultant required. AppFolio is mid-market, with a sweet spot above 50 units, and charges per unit in a way that makes furnished rentals more expensive than long-term ones. QuickBooks alone has a generic ledger with no trust accounting, no per-entity consolidated reporting, and no owner statements.
Which leaves operators with 10 to 50 furnished units in a gap: too small for AppFolio's return on investment, too complex for QuickBooks alone. Moving a 30-unit operation from QuickBooks to AppFolio costs roughly $8,000 to $15,000 in implementation time and lost productivity across three months.
And then there is the entity problem. Most mid-term operators hold properties across multiple LLCs for liability segregation — often one per building or one per state. QuickBooks Online charges per company file, so four LLCs runs $360 per month before a single property management licence. The consolidated view across those entities, which is the only view that answers "how is the portfolio doing," typically does not exist.
None of these are exotic problems. They are the daily operating conditions of the fastest-growing segment in rental housing.
What purpose-built means here
The mid-term category has specific financial architecture requirements that follow directly from its defining characteristic: stays long enough that everything designed around nightly turnover or annual tenancy breaks.
Deposits that survive past day 29, through rolling re-authorisation or captured-and-refunded handling with correct liability treatment. Recurring monthly billing with grace periods and retry logic. Owner statements that reconcile across entities and currencies. Mid-stay adjustments — extensions, partial refunds, rate changes — handled as first-class events rather than manual journal entries.
RentOS is built for that category specifically. Not a nightly-rate platform with a longer minimum-stay setting, and not annual-lease software with a booking calendar bolted on.
If you operate furnished mid-term units and any of the failure modes above sound familiar, book a demo at rrentos.com.
FAQ
Why do credit card holds expire after 28 days? It is a network-level limit set by Visa and Mastercard, not a policy of any individual platform or payment processor. Pre-authorisation holds automatically release after 28 days regardless of the booking length, which means every stay in the mid-term rental category (30+ days) outlasts its own deposit protection.
How can mid-term rental operators secure deposits on long stays? Three approaches work: rolling re-authorisation on roughly a 25-day cycle with a brief overlap window; capturing the deposit as an actual charge and refunding it at checkout with correct liability accounting; or issuing a formal deposit invoice through a B2B portal with a documented dispute trail. Which fits depends on the market — captured-and-refunded is standard in SEPA and bank-transfer markets and for corporate clients, while rolling re-authorisation suits card-based consumer markets.
Is a captured security deposit income? No. A captured deposit is a current liability — typically recorded as "Tenant Deposits Payable" — and must not be booked as rental income. If a portion is later applied to damages, that portion becomes income or offsets an expense and is taxable; the remainder is refunded with no P&L impact. Systems that record captured deposits as income can trigger tax notices from state revenue departments.
Why doesn't AppFolio sync properly with QuickBooks? AppFolio has a native QuickBooks integration, but it is one-way and incomplete. It does not sync owner draws, trust account movements, vendor 1099s, custom GL accounts, mid-stay partial refunds, or security deposit interest accrual. Operators running both systems typically find their books diverge within 30 to 60 days, and reconciliation ends up taking longer than manual entry would have.
What does property accounting cost mid-term operators in time? Industry reporting puts manual export-import reconciliation between property management and accounting systems at 8 to 15 hours per month for multi-property teams. For a 30-unit operator, that is roughly $400 to $1,500 per month in bookkeeper time — a recurring cost that scales with portfolio size and is largely invisible in unit economics until someone measures it.