DCF for hotels — what changes vs. real estate
Every investment bank pitches the DCF — discounted cash flow — as the rigorous valuation method. For most asset classes that is true. For hotels it is the most fragile of the three primary methods, and the operator who walks into diligence treating it as authoritative gets pulled into assumption fights that swallow weeks.
Why DCF is harder for hotels than for office or retail
Office and retail produce contractual cash flows: a 10-year lease at a known rent escalation. The DCF inputs are deterministic for most of the forecast horizon. Hotel cash flows are reset every night. There is no contracted revenue beyond the rare block-purchase agreement; every roomnight is re-priced, every F&B cover is discretionary, every spa treatment is opt-in.
The result: the DCF for a hotel is a sequence of bottom-up operating forecasts compounding over 10 years. Each year compounds the noise of the prior year's assumptions. A 1.5% error in year-three RevPAR growth becomes a 6-8% error in terminal value, which is where 55-65% of the DCF value lives.
The three inputs that actually move hotel DCF value
Operators get into trouble pushing back on RevPAR growth assumptions year by year while ignoring the terminal cap rate, which is doing most of the heavy lifting in the buyer's number. The terminal cap is where buyers smuggle in their pessimism. Push there first.
The capex line buyers love and sellers hate
Every hotel DCF includes a capex line — usually broken into FF&E reserve (3-4% of revenue, ongoing) and major capex (lumpy renovation cycles every 6-8 years). Buyers model major capex aggressively; sellers model it lightly. A €12M renovation in year 5 vs. year 7 changes the DCF value by 2-3% just from timing.
The defensible position from the seller side is the actual renovation calendar from the brand, with FF&E inspection reports backing the timing. If the brand has approved a renovation deferral to year 7, you can defend year 7 in your model. If the brand audit says year 5 is the requirement, the buyer will price it at year 5 regardless of what your model shows.
Why I still build the DCF
Despite the fragility, I build a DCF on every transaction. Not because it produces the answer — the comp-trade and NOI-multiple approaches do that — but because the DCF surfaces the assumption fights that need to be had: when does stabilization arrive, what is the terminal cap, what is the renovation schedule. Those are real conversations whether the DCF closes the valuation gap or not.