Haute Lumière
Commerce · IV.09 · MMXXVI · daylight
One page each. A reader who reads only these ten pages has the chapter.
The idea. How long a thing is meant to last is a specified engineering quantity, not an aspiration, and it is written down in codes that are already on your engineers' shelves.
Eurocode EN 1990, Table 2.1, sets indicative design working lives: 10 years for temporary structures, 10–25 for replaceable structural parts, 15–30 for agricultural structures, 50 years for buildings and other common structures, and 100 years for monumental buildings, bridges and other civil engineering works. US bridge practice designs to 75 years under AASHTO LRFD; UK practice to 120. The Hong Kong–Zhuhai–Macao Bridge states 120; Canada's Confederation Bridge states 100. ISO 15686 gives the method for planning service life asset by asset.
What it decides. How many times you build the thing. Over a two-hundred-year study period: a 25-year life means 8 builds, 40 years means 5, 50 means 4, 100 means 2, and 200 means 1.
Worked example. Two proposals for the same service. One is designed for 40 years at £40.0m; one for 200 years at £52.0m — a 30 percent durability premium. Over two centuries the first is bought five times and the second once.
Why it matters. Almost no organisation can answer how long is this meant to last for any asset it owns. Design life belongs on the asset register in years, with its source, beside the depreciation schedule — and the two are allowed to disagree, because when they do, something interesting is true.
You already know this because you have replaced a kettle three times while the same cast-iron pan sat on the shelf, and you never once thought of it as a financing decision.
The idea. The principle is real, it is Haudenosaunee, and the version most often quoted is not the version that is documented. Both halves matter.
What is recorded. The Great Law of Peace, Kaianere'kó:wa, governs the Haudenosaunee Confederacy. The most widely reproduced English rendering is Arthur C. Parker's The Constitution of the Five Nations (1916), New York State Museum Bulletin 184. Among the duties of the Confederate Lords:
Look and listen for the welfare of the whole people and have always in view not only the present but also the coming generations, even those whose faces are yet beneath the surface of the ground — the unborn of the future Nation.
It is a clause about unborn generations and it does not say seven.
What is living teaching. The seven-generation formulation is genuinely Haudenosaunee and is stated by Haudenosaunee leaders in their own voice — most influentially by Oren Lyons, Faithkeeper of the Turtle Clan of the Onondaga Nation, who carried it into international forums from the 1970s, including the 1977 Geneva NGO conference.
What is later attribution. From 1988 a Vermont company printed a version on cleaning products. The phrase then entered corporate and sustainability literature, usually credited to "the Great Law of the Iroquois" with no citation that resolves to a primary text. The separate claim that the Great Law shaped the US Constitution has a congressional acknowledgment (1988) and an unresolved scholarly dispute.
The rule. Quote the clause from Parker and name the version. Attribute the seven-generation formulation to living Haudenosaunee teaching and to Oren Lyons. Say once that the Confederacy is a government, not a metaphor.
Why it matters. A reader who checks one citation and finds a slogan will not check the second one. Cited properly, the principle gets stronger.
The idea. The same two structures give opposite answers depending on which instrument you measure them with, and the two instruments are both correct.
The case. Case A, 40-year design: £40.0m capital, £0.240m a year routine maintenance, £8.0m refurbishment at year 20 of each cycle, rebuilt at years 40, 80, 120, 160. Case B, 200-year design: £52.0m capital, £0.208m a year, £7.80m refurbishment at years 50, 100, 150.
rate Case A Case B cheaper
0.0% (cash) 288.00 117.00 the durable one
2.0% 93.14 66.58 the durable one
3.5% 65.69 59.63 the durable one
5.0% 54.93 56.90 rebuild it
7.0% 48.51 55.25 rebuild it
Green Book schedule 68.97 60.48 the durable one
The reading. In cash, durability saves £171.00m and costs 2.46 times less. At a 7 percent hurdle it loses by £6.74m. The reversal is worth £177.74m and no physical fact changed between those two lines.
Why it matters. Present value is the frame in which this argument has already been lost. Present the cash column too — not instead, as well — because the cash column is the one a maintenance budget actually experiences.
You already know this because you have watched a landlord choose the cheap boiler four times in twenty years, and you knew exactly what was happening, and you could not have written the reason down.
The idea. For any two designs of different length there is one discount rate at which they cost the same, and finding it converts an argument about values into an argument about a parameter.
For the two cases above it is 4.510 percent a year, computed by bisection. Above it the four-times-rebuilt structure is cheaper in present value. Below it the durable one is. Nothing about concrete changes at four and a half percent.
How to use it. Do not argue that the durable option is better. Compute the indifference rate and put it on the front page beside your organisation's hurdle rate. The question in the room becomes is our hurdle rate above or below 4.51 percent, and why is it that number — which is Chapter III.05's question, and it has a published answer.
Worked example. An organisation using a 10 percent hurdle sees Case A cost £44.52m against Case B's £54.15m. The same organisation adopting the Green Book declining schedule for anything beyond thirty years sees £68.97m against £60.48m. One policy amendment, one page, and the answer inverts.
Why it matters. An indifference rate is falsifiable and a conviction is not. It also tells you when to stop: if your hurdle is 12 percent and the indifference rate is 4.5, the durability case cannot be won on cost, and you should move to the funding structure instead of arguing harder.
The idea. Discounting is usually blamed for shrinking the distant benefits of lasting. That is the smaller half. It shrinks the cost of not lasting by more.
PV of Case A's four rebuilds, 4 x GBP 40.0 m
at 0.0% GBP 160.000 m 100.00% of face
at 3.5% GBP 13.462 m 8.41% of face
at 7.0% GBP 2.862 m 1.79% of face
At 7 percent, buying the same bridge four more times costs £2.862m in today's money — less than the paint on the durable one.
Why this happens. A replacement is always later than the thing it replaces, so it always sits further down the exponential. Any rate high enough to make the long-lived asset's benefits vanish makes the short-lived asset's liabilities vanish faster.
What follows. Stop presenting durability as a benefit to be discounted. Present the replacement cycle as a liability to be provisioned — which is the sinking fund in Brief 7 — because a provision is a present-tense number and a benefit is not.
Why it matters. This is the single most common misunderstanding in long-horizon capital argument, and correcting it changes which column you present, not merely how loudly.
The idea. Convert a promise about the future into a sum of money today, and the discount rate stops being able to touch it.
corpus = annual real maintenance cost / real spending rate
Case B needs £0.208m a year. At a 3.0 percent real spending rate the corpus is £6.933m — 13.3 percent of the structure's own capital cost. At 2.5 percent it is £8.320m; at 4.0 percent, £5.200m.
The identity, and it is exact. The corpus costs more than the discounted value of the same flow, and the ratio is nothing but two rates:
corpus / PV = hurdle rate / spending rate
at 3.5% hurdle, 3.0% spend 1.167x
at 7.0% hurdle, 3.0% spend 2.333x
at 10.0% hurdle, 3.0% spend 3.333x
That ratio is the honest price of making a perpetual promise credible, and it is the only number anyone needs to argue about.
The precedents. The National Trust's Chorley Formula (1946): a property is accepted only with an endowment sufficient to maintain it in perpetuity. US cemetery perpetual care trusts, required by statute, income only. Norway's fiscal rule, capping transfers at the expected real return — cut from 4 percent to 3 in 2017.
You already know this because you understand why a charity that is given a building without money is being given a problem.
The idea. When the liability has a date rather than a horizon, accumulate it instead of endowing it.
annual payment = target x i / ((1 + i)^n - 1)
To accumulate one £40.0m rebuild by year 40:
| Real return | Level payment | Total paid in |
|---|---|---|
| 0% | £1.000m /yr | £40.00m |
| 1% | £0.818m /yr | £32.73m |
| 2% | £0.662m /yr | £26.49m |
| 3% | £0.530m /yr | £21.22m |
At 2 percent real the fund earns £13.51m of the £40.0m, and the payment is 1.66 percent of the asset's capital cost per year.
The point of the table. The gap between the 0 percent row and the 3 percent row is the whole argument for starting early rather than large. A sinking fund begun in year one is a rounding error on an operating budget; the same fund begun in year thirty is a crisis.
Where it is already law. Condominium and strata reserve funds. Florida's post-Surfside regime requires structural integrity reserve studies with mandatory funding from 31 December 2024, removing owners' previous right to waive reserves.
Why it matters. A dated liability that is not provisioned is a decision to hand the bill to whoever is in the chair that year, and that person is not in the room when it is taken.
The idea. A maintenance obligation stated as a flat cash amount has an expiry date that nobody wrote down.
a flat GBP 0.208 m/yr covenant buys, in real work:
2.0% inflation year 30 55.2% year 50 37.2% year 100 13.8%
2.5% inflation year 30 47.7% year 50 29.1% year 100 8.5%
3.0% inflation year 30 41.2% year 50 22.8% year 100 5.2%
At 2.5 percent inflation the covenant is buying under a third of its original work by year fifty.
The fix. One sentence naming the index. That is all. The only honest alternative is to name the year the covenant expires in the document, so that somebody has to renew it deliberately rather than watch it evaporate.
Where to look for one. Concession and PPP agreements carry lifecycle obligations and a handback condition — a residual-life test at transfer that makes the final payment depend on the asset's state. That mechanism is the one worth copying, because it converts a future condition into a present cash flow for the party who controls it.
Why it matters. Covenants are drafted by people optimising the first five years. The indexation clause costs nothing at signature and is worth more than every other clause combined by year fifty.
You already know this because you have seen a service charge that has not moved since 2004 and a building that shows it.
The idea. A short-lived design carries a free option: at every rebuild you may build something else. A durable one does not. That option has a price and it belongs in the comparison.
PV(rebuilds and maintenance avoided) > durability premium
+ value of the option given up
Worked, with declared assumptions. A 40 percent chance that a materially better answer exists by year 50; a gain of 35 percent of capital if it does; a switching cost of 25 percent. Net gain £4.00m; Green Book factor at year 50 is 0.19726; option value £0.316m. Against Case B's Green Book advantage of £8.48m, the durable design still nets £8.17m.
Now ask it the other way. How large must the foreclosed option be to overturn the durable case? 8.48 / (0.40 × 0.19726) = £107.5m — 2.69 times the structure's own capital cost. That sounds impossible.
It is not. Elwha Dam, 1913, and Glines Canyon Dam, 1927: concrete gravity structures, 28 MW combined, no fish passage although Washington State law of 1890 required fishways. Salmon runs estimated near 400,000 adults a year fell below 4,000. Elwha stood 99 years and was removed in 2012; the National Park Service puts the whole restoration project at about USD 325 million, or USD 11.61m per MW — roughly four times what comparable capacity costs to build.
Why it matters. A permanent structure is an option written against the future and exercised by someone who does not get a vote. Price it in the paper, or someone will price it with a demolition contract.
The idea. The instruction is not build something that lasts. It is leave the next builders the means and the money to build it again.
The case that proves it. Ise Jingu is rebuilt entirely every twenty years in the rite of shikinen sengu; the sixty-second rebuilding was 2013, the sixty-third is 2033. Tradition dates the first to 690 CE. The institution is about 1,336 years old and no timber in it is more than twenty. In 1923 the shrine began a two-hundred-year forestry plan to grow the hinoki cypress for rebuildings from about 2130 — now 103 years old with 97 still to run.
Why the interval is the design. A skill exercised every twenty years is transmitted by people: every carpenter has worked with someone who did it last time and will work with someone who does it next. A skill exercised once a century has to be reinvented at full cost. Choose the renewal interval so the knowledge travels in people, not in documents, wherever you have the choice.
The succession note. Every long-lived asset carries one short document naming four things the next builder will need:
Ise's succession note is a forest and a twenty-year rite. Yours can be a drawing archive, a spares contract, an apprenticeship and a sinking fund.
And the corollary. Any asset whose design life exceeds the confidence interval on its own purpose should be designed for removal, with the removal costed and provisioned in the same paper. Write the demolition into the birth certificate. It is the cheapest insurance in the chapter and almost nobody buys it.
All figures in these briefs are computed in lib/verify/IV_09.py and sourced in the chapter's Works Cited. The discount rate itself is Chapter III.05.