A rent miss is a permanent reduction in a growing stream; a delay is a year of full carry, funded by equity, sitting on top of a year of no income at all.
Yes, and the gap is measurable. On the illustrative 180 MW turnkey campus — $1.98 billion of cost, $215 million of stabilised net operating income, 60 per cent construction financing, a seven-year hold and a 7.75 per cent exit — the base case returns 18.56 per cent levered. An eight per cent rent miss takes it to 15.99 per cent, a fall of 257 basis points. A twelve-month energisation slip takes it to 14.56 per cent, a fall of 399. The delay costs 142 basis points more.
Worked in full in The Data Center Development Handbook by Julian R. Sterling, with every figure reproduced in a free workbook.See the book on Amazon →
The delay case is not a haircut on income. It is a year of full carry on a nearly fully drawn balance, followed by a year of income that never arrives. At 60 per cent loan-to-cost the drawn balance on this campus is $1,188 million. Twelve months of interest on that balance is $75 million in round terms — $74.8 million computed — which implies a construction rate of 6.3 per cent. Extended general conditions, insurance and overhead add roughly $20 million on top.
| Line | Amount |
|---|---|
| Additional construction interest on the drawn balance | $74.8 million |
| Extended general conditions, insurance and overhead | $20 million |
| Cost of the slip | $94.8 million |
The interest rate here is not invented. It is taken out of the deal rather than out of the air: the delay figure and the drawn balance together fix it at 6.3 per cent, and the same rate then runs the whole model. A model built on the lender's actual rate reproduces the memo the lender is reading.
Who funds the $94.8 million matters as much as its size. Lenders do not enlarge commitments for delay, so the money comes from equity. The rent miss, by contrast, is absorbed inside a capital structure that is already in place.
Run both on the same model, with no rent escalation so that the comparison is like for like, and the ranking is stable.
| Case | Levered return | Fall |
|---|---|---|
| Base case | 18.56 per cent | — |
| Eight per cent rent miss | 15.99 per cent | 257 basis points |
| Twelve-month energisation slip | 14.56 per cent | 399 basis points |
The reason the delay wins is structural rather than arithmetical. The rent miss is a permanent reduction in a growing stream. The delay is a year of carry funded by equity, plus a year of lost net operating income, plus every future dollar pushed one year further out, plus a year of a fifteen-year lease term consumed without collecting on it. And it carries a tail the rent case does not: a twelve-month slip is the kind of event that trips a lease outside date, and a delay that becomes a re-leasing problem is a different project.
There is a second-order point that cuts the other way and is worth pricing properly. An eight per cent rent miss is usually valued at first-year rent: eight per cent of $215 million is $17.2 million, capitalised at 7.75 per cent is $222 million of exit value. That figure assumes the shortfall never escalates.
A first-year shortfall is a shortfall in the rent that escalates. Seven years on, under a three per cent escalator, the shortfall at the exit is $21.2 million, and capitalised it is $273 million — twenty-three per cent larger. Price a rent miss at the exit, on the escalated rent, not in year one.
The escalator is worth stating in the same breath, because it moves the headline return more than most negotiators expect. On identical dirt, identical megawatts and identical capital, the levered return is 18.6 per cent with a flat lease and 22.6 per cent with a three per cent escalator. Net rent over a fifteen-year term is $281 million higher at three per cent than at two, or 7.5 per cent of the total, so matching a three per cent escalator with a two per cent one requires first-year rent of $112.93 per kW-month instead of $105.00 — $7.93 more. A tenant offering a dollar of first-year rent for a point of escalation is offering about a seventh of what it is asking for.
Sensitivities do not add. Take a twelve-month delay arriving, as delays do, with a five per cent cost overrun, an eight per cent rent miss and twenty-five basis points of exit cap widening. The four effects taken one at a time sum to 873 basis points; run together they cost 809 — 64 basis points less bad than the sum. The lesson runs in both directions: never add sensitivities to estimate a combined case, and never assume the combination is worse than the sum either.
Three things follow, and none of them are about rent.
And quote a levered return with its escalator attached. Four points of return sit between a flat lease and a three per cent one on the same asset, so a return quoted without its escalation assumption cannot be compared to anything.
On a $1.98 billion campus at 60 per cent loan-to-cost, the drawn balance is $1,188 million. Twelve months of carry on it is $74.8 million of interest, implying a construction rate of 6.3 per cent. Extended general conditions, insurance and overhead add about $20 million, for $94.8 million in total.
The bare slip costs $94.8 million, which is 12.0 per cent of committed equity, or $527,000 per MW. Add a five per cent cost overrun, an eight per cent rent miss and twenty-five basis points of cap rate widening and the call is $197.6 million, or 24.9 per cent of committed equity.
No. A twelve-month delay, a five per cent cost overrun, an eight per cent rent miss and twenty-five basis points of exit cap widening cost 873 basis points when taken one at a time, but 809 when run together — 64 basis points less bad than the sum. Levered returns do not decompose additively.
This article is one calculation from The Data Center Development Handbook. The book takes the same case from first principles to the decision, chapter by chapter, and every figure it prints is a live formula in the free companion workbooks.
Get the book on Amazon →Free companion files
Also on Amazon UK · Amazon Germany · Amazon France · Amazon Canada
Reading guide: infrastructure, data centres and energy → · All 1706 articles →
If this book helped — or didn’t — a few lines on Amazon are worth more than they look: they are what the next reader goes on. Write a review. The workbook stays free either way.