A cost shock, and who pays for it

Second of three on management agreements. The first is where the short-let fee came from. The third is what I read first in the contract.

Arguments about fee structure go in circles until someone puts a P&L on the table. So here is one.

The building

Twenty-five studios outside the centre of Madrid. Occupancy of 80 per cent and €125 a night including VAT, so €113.64 net. An average stay of 3.2 nights, which gives 7,300 occupied nights and about 2,281 stays a year.

Inside that €125 the operator allocates €30 a stay to a check-in fee, meant to cover cleaning, linen and amenities. The guest never sees the split. It is an internal pricing decision, and it is the line to watch.

The owner owns the building, so there is no rent, and this is the building's NOI rather than anybody's business. The operator's own people and systems are not in it: those come out of whatever fee they charge, under either structure. Every figure excludes VAT.

The bad year is a cost shock and nothing else. Cleaning, linen and amenities go from €28 a turn to €38. Energy rises 40 per cent, maintenance 20 per cent. Not one euro of revenue changes and occupancy holds. This is the year the building performed and the costs did not behave.¹

The P&L

Look at the two cleaning lines before anything else, because that pair is the one everybody models wrong.

In theory the check-in fee is a pass-through. Money in, cleaning out, nothing left over and nothing missing. In the good year it collects €68,438 and the cleaning costs €63,875, which is what a pass-through set to cover itself looks like: €4,563 of slack, near enough to zero that every model treats the pair as self-cancelling.

In the shock year it collects the same €68,438 against €86,688 of cleaning. It is €18,250 short, and it stops being a pass-through the moment it is short.

Nobody decided that. The allocation did not move, because moving it means raising the one price the guest actually sees, which is a pricing decision that has to survive the market rather than an internal adjustment. A line the spreadsheet treats as self-cancelling becomes a cost, and the cost has an owner in it.

Worth asking your operator which way this runs, because it is not always a pass-through. Some allocate well above what a turn costs, and then it is not a pass-through at all, it is margin. Whose margin is a question the agreement should answer and often does not.

The fee, two ways

Same building, same NOI. Only the fee changes.

The two structures cost almost exactly the same in a normal year: €165,909 against €162,481, a difference of €3,428 on a building making €588,596. That is deliberate, and it is the only honest way to compare them. A restructured fee that happens to be much cheaper is not a restructured fee, it is a price cut, and the operator will find that money back somewhere you are not looking.

So this is not an argument about paying less. Both take a little over 27 per cent of NOI.

Now read the shock column. The 20 per cent is €165,909 in both years. The building lost €34,513 and the fee did not notice, because nothing that happened touched the number it is calculated on. Under the split fee the operator gives up €8,363, and the owner loses €26,150 instead of €34,513.

Five years, and who feels them

A cost shock is one of the ways a year goes wrong. Here are five, against the same normal year.

Start with the first row, because it is the one that makes the structure signable. In a genuinely good year, revenue up and costs held, the operator earns more under the split fee than under the 20 per cent: €17,926 against €16,591. This is not a proposal to pay an operator less. It is a proposal to pay them for something different, and to pay them better when they deliver it.

Second row, revenue up ten per cent with costs up ten per cent to produce it. The 20 per cent pays the same €16,591 it pays for the good year, because it cannot tell the two apart. The split fee pays €15,351.

Third row is the cost shock in one line. Under the 20 per cent the operator's income does not move while the owner loses €14,981.

Fourth row is a bad market. Here the revenue fee behaves well, falling €16,591 with the revenue. The split fee gives up more, €20,163, and the owner is €3,572 better off. Whatever else is true, a fee on profit does not abandon the owner in a downturn.

The row I would put in front of an owner

The last one.

Revenue up ten per cent, bought by paying the portals 20 per cent instead of 15. The kind of year that looks like growth in every report and in no bank account.

Under the 20 per cent the operator earns €16,591 more for a year that leaves the owner €1,267 better off. Thirteen euros to the operator for every euro to the owner, for a year of buying revenue at a price that did not work.

Under the split fee the operator earns €4,887 and the owner keeps €12,971.

Nothing in a fee on room revenue can see a channel commission, because commission is deducted after the number the fee is measured on. Move the base onto net revenue and the operator pays for that decision twice: once in the base, which falls when commission rises, and again in the incentive, which is measured after every cost.

That is the difference between an operator who is paid to fill a building and one who is paid to run it well.

Which risk are you handing over

A fee on revenue shares demand risk and ignores cost risk completely. If the market turns, your operator's income turns with it. If your costs run away while the market holds, they never find out.

A fee on profit shares both, because profit is what is left after the costs.

Short-term rentals earned the revenue fee, and it is worth saying, because the industry is not usually the one credited with rigour. In a properly automated asset almost every cost moves with revenue. No front desk standing idle. Cleaning bought per check-in rather than staffed by the month. Access, messaging and pricing running without a person per building. Push that far enough and there is very little left that does not flex, and a fee that flexes with revenue fits a P&L that flexes with revenue.

The building above is not quite that asset, and most are not. The community, the IBI, the insurance and most of the energy sit still while revenue moves. The more of those you have, the worse a pure revenue fee fits, because your operator's pay flexes and your costs do not.

What this does not settle

The percentages are mine, not a market standard. Eight and twenty is a pairing that happens to be fee-neutral on this building, which is the property worth copying rather than the numbers. Run it on yours before you propose anything: the levels that make it neutral depend on your margin, and an operator asked to take a pay cut dressed as an alignment will say no, and should.

And an incentive measured on NOI is only as good as the definition of NOI, which is the next piece, along with the clauses that decide whether any of this survives contact with year three.

Notes

1. The building is illustrative and so are the figures, though the shape is one I have seen many times: a check-in fee that nets to zero in a model and does not in a bad year, and a management fee that never notices either way. Annual figures, excluding VAT. ↩