How the Analysis Works
Dash Decisions prices Hotel Management Agreement bids from the operator’s side of the table. The work rests on one instrument — the Master Deal Engine — and a set of disciplines around how it is calibrated, how its output is checked, and how the judgment is written down. This page describes both, at the level a reader evaluating the practice would want before a conversation.
The Master Deal Engine
The Master Deal Engine is a calibrated financial model that prices an operator’s HMA bid. It runs forward: set the deal’s inputs and it computes net present value, GOP margin, performance-test failure probability, and the fee economics over the full term on a month-by-month grid. It also runs backward — a solver finds the input change required to hit a stated target, which is the basis of the Operating Levers analysis described below.
Two things separate it from a spreadsheet.
Dimensional calibration. The model’s defaults are not single market averages. The engine carries seven operator archetypes, six property archetypes, five market tiers, and three property statuses — new-build, conversion, and takeover. A major branded operator bidding a full-service convention headquarters hotel in an urban gateway market on a new-build carries different calibrated assumptions than a mid-market third-party operator taking over a select-service property in a tertiary market. Most analysis collapses those dimensions into an average. Averaging hides the outliers, and the outliers are where both the money and the risk sit.
A calibration audit trail. Every default in the model traces to a source and a confidence grade. Fee bands, departmental cost ratios, discount rates, termination economics, and ramp curves were each triangulated across four independent research passes and checked against industry references — The HMA & Franchise Agreement Handbook (5th edition) from Jeffer Mangels Butler & Mitchell, the DLA Piper treatment of HMA economics, and HVS studies. Where the sources agree, the value is graded high-confidence. Where they diverge, the uncertainty is carried as a band rather than smoothed into false precision, and the engagement scope says so. Documented uncertainty survives a sophisticated owner-CFO’s scrutiny. Papered-over uncertainty does not.
How a Bid Gets Priced
The pricing output is three numbers, not one.
A floor — the fee at which the deal’s net present value reaches zero. Below it, the operator is bidding to lose money.
A target — the calibrated market fee for that operator-and-property type.
And the negotiating room between them, stated in basis points and dollars, so the operator knows exactly how much give it has before the deal stops paying.
Around those three numbers sits a four-case scenario band — bear, base, bull, and stress. A deal whose base case is healthy but whose bear case is already negative is a different bid than one that survives its stress case, even when the headline number is identical. The band makes that difference visible instead of leaving it buried in the point estimate.
Underneath all of it is one non-negotiable discipline: every number in the brief traces to a sourced assumption, and every load-bearing assumption is named. The brief tells the operator which assumptions are doing the quiet work — the comp-set positioning, the term length, the owner’s development-cost basis — and marks the ones that are analyst estimates rather than owner-disclosed facts. An operator should never be surprised later by an assumption it did not know it was relying on.
Monte Carlo Performance Test Exposure
Most bid analysis stops at a point estimate with a stress case set beside it. Performance-test exposure is not a point estimate. It is a distribution, and treating it as a single number understates the tail.
The engine runs ten thousand simulations per engagement, varying the drivers that actually move performance, and reports the result as a probability distribution: the tenth-percentile outcome, the median, and the ninetieth-percentile outcome, plus the probability the performance test fails and the conditional loss if it does. This is where a thin-looking deal reveals whether its risk is spread evenly across the term or concentrated in a single window — most often the opening ramp. That distinction determines whether the right response is a lower bid or a specific contract term, and the two are not interchangeable.
The simulation is termination-aware. A failed performance test is not automatically a lost contract. Owners exercise termination rights at very different rates depending on who they are and how replaceable the operator is — a public authority with few credible replacement operators behaves nothing like an independent owner of a fungible select-service asset. The model draws termination at a rate calibrated to the owner archetype rather than assuming every failure ends the deal, and it reports two distinct figures: the probability the test is failed, and the probability the owner actually terminates. Conflating the two overstates the exposure, sometimes by a wide margin. Keeping them separate is often the difference between a deal that reads as unbiddable and one that reads as manageable.
How the Numbers Are Checked
A model is only as trustworthy as the check on its output. Before any number is reported, the engagement runs a reconciliation pass: the model’s figures are checked against the deal’s own terms, and any finding that revises an earlier conclusion is traced back to the specific contract provision that drives it before it is trusted. The method also carries a dedicated contract read — a pass that looks for the provisions a numbers-first analysis misses: the carve-outs, contingencies, and negotiable blanks that change which term governs the economics without changing a single number. The failure a sophisticated operator most fears is not an unsophisticated model. It is a missed provision. This step exists to close that gap.
The Eight Deal-Killer Flags
When an operator receives proposed HMA terms, the engine screens them against eight flags — the terms that most reliably damage operator economics. Each is defined, graded, valued against its dollar impact on NPV, and paired with a remediation.
One — Performance test set above market. Thresholds pitched above the industry two-pronged norm — a RevPAR-penetration prong and a GOP prong — shift termination risk asymmetrically onto the operator.
Two — Termination for convenience without adequate liquidated damages. An owner right to terminate without cause, with damages that fail to cover the remaining fee stream, unrecovered key money, and transition cost, strands the operator’s investment.
Three — Approval rights out of balance. Owner approval rights over budget, capital, and key personnel that run past the norm create operational paralysis. The mirror failure — an owner who waives meaningful approval — signals absentee ownership and carries its own risk. Screened as budget-construction, fee-balance, and capital-threshold sub-flags.
Four — Key money structured to be unrecoverable. Back-loaded amortization or heavy late-term recapture turns an operator’s up-front investment into money it cannot get back if the deal ends early.
Five — Indemnification uncapped or mis-allocated. The defensible standard is gross negligence and willful misconduct with a hard liability cap. Exposure past that standard, or an absent cap, is a flag.
Six — Inadequate lender protection. Fee rights subordinated to debt service without a subordination-non-disturbance agreement in place at closing leave the operator’s revenue collection at the lender’s discretion.
Seven — Transfer and assignment without operator protection. An ownership change with no operator consent, or a notice-only requirement, exposes the operator to a counterparty it never underwrote.
Eight — PIP and pre-opening capital pushed onto the operator. Property-improvement and brand-standard capital is typically owner-funded. Operator contribution past the norm is a flag.
Each flag is graded across four severity bands — mild, moderate, severe, critical — against its modeled dollar impact, not against a generic checklist. Each carries the owner’s most likely counterargument, framed honestly rather than as a straw man, and a remediation pattern: operator-favorable language by default, with a compromise fallback held in reserve for the negotiation.
The flags roll up to one of four verdicts: clean bid, negotiate flags, proceed with caution, do not bid. The last one matters most. It is used rarely and on purpose, because a screen that never says walk is not a screen. When the analysis returns do not bid, it is defending the operator’s willingness to walk with a structural reason, not a feeling.
Operating Levers
Not every marginal deal should be walked. Some can be made bid-able by changing a single term.
The Operating Levers analysis runs the engine backward. Given a threshold the deal does not currently clear — a performance test it is likely to fail, or a net present value that does not yet justify the bid — the solver finds the smallest change to a single lever that gets there. What occupancy, what ADR, what comp-set position, what fee, what key-money level would move the deal from likely walk to bid-if-we-can-execute. It converts a binary walk-or-bid into a precise, negotiable ask: here is the one thing that has to be true, and here is how far it has to move.
How the Analysis Is Delivered
The output is a decision brief a CEO or CFO can read in five minutes. The judgment sits on top — the recommended bid, the walk-away floor, the room between them, and the two or three levers that move the answer — with the numbers underneath for anyone who wants to trace them. Every sentence has a number behind it. Every number traces to a sourced assumption. The brief is built to be defended across the table, not only read on the operator’s side of it.