Haute Lumière
Commerce · IV.09 · MMXXVI · daylight
For the person working inside a gainshare arrangement — where a defined share of verified improvement returns to the people who created it. This chapter is the hardest one for a gainshare and the most valuable: it is about work whose gain arrives after the scheme's measurement period, and how to get paid for it anyway.
A gainshare has four parts: a baseline, a measure, a share, and a period and verifier. The period is the one that breaks here.
Almost every gainshare measures annually. Almost everything in this chapter pays out in decades. A maintenance regime that adds forty years to a structure's life, a specification change that removes a rebuild, a drawing archive that saves the next generation a survey — none of these show up inside a twelve-month measurement window, and a scheme that only measures twelve-month windows is structurally unable to pay you for the most valuable work you will ever do.
That is not a reason to stop doing it. It is a design problem with a known answer, and the answer is the same one the chapter gives the finance function: convert the future flow into a present sum. The firm does it with an endowment corpus. You do it with a capitalised avoided liability — and the arithmetic is identical.
what you claim = the annual cost you removed / a published rate
If your work removes £40,000 a year of real maintenance cost permanently, and the firm's own published spending rate is 3.0 percent, you have created £1,333,333 of present value in one year. At the firm's 7 percent hurdle the same flow is worth £571,429. Both numbers are defensible and you should present both, but you should know, and be able to say, that the ratio between them is exactly 0.07 / 0.030 = 2.333 and that it is a choice of rate rather than a fact about your work.
Exercise 1.1 — Read your scheme for its horizon (two hours)
Answer five questions in writing.
Question five is the one nobody has ever been asked. If the answer is no, you have found the thing worth changing this year, and Part Three is how.
Exercise 1.2 — The four invisible gains (one week)
Go looking in four places. Every one of these is work people in your position do routinely and are never paid for, because it shows up as an absence.
Output: four candidates, one page, roughly quantified.
Exercise 1.3 — Find who already gets paid this way (two hours)
Somewhere in your organisation, or in a contract it holds, somebody is already paid for a future condition. Look for a concession handback clause — a final payment tied to the measured residual life of an asset at transfer. Look for decommissioning provisions, warranty reserves, pension funding.
The mechanism you want already exists in your organisation's documents. You are asking for it to be pointed at you rather than at a counterparty, which is a far easier conversation than asking for a new one to be invented.
Exercise 2.1 — Capitalise one gain properly (one afternoon)
Take the strongest of your four candidates. Establish, in writing:
M — the annual real cost removed, or the annual real cost of the obligation you have extinguished. Be conservative. A number you can defend beats a number you can justify.s — the published real spending rate, if your organisation has one, or 3.0 percent with Norway's fiscal rule as the citation. permanent: value = M / s
dated: annual provision = R x i / ((1 + i)^n - 1)
For a dated one: £40.0m at year 40 costs £0.662m a year at 2 percent real, of which the fund itself earns £13.51m. If your work moved that date out, the gain is the reduction in the required annual provision, and that number lands inside an annual measurement period. That is the trick: a deferred liability becomes a present-period saving the moment somebody is provisioning for it.
Exercise 2.2 — Build the two-column claim (one week)
Never present one number. Present four columns, exactly as the executive workbook does, because a single figure invites a single objection.
| Short-lived path | With my change | |
|---|---|---|
| Cash over the study period | £288.00m | £117.00m |
| PV at the hurdle rate (7%) | £48.51m | £55.25m |
| PV at the public appraisal rate (3.5%) | £65.69m | £59.63m |
| PV on the declining schedule | £68.97m | £60.48m |
Then say the sentence out loud, because it is the one that wins: "Two of these four columns say my change loses. The physical asset is identical in all four. The disagreement is entirely about a rate, and the indifference rate is 4.510 percent."
You are not asking anyone to be generous. You are pointing at a parameter.
Exercise 2.3 — Know the identity (fifteen minutes)
Somebody will say the endowment number is inflated because the flow is only worth its present value. The answer is exact:
corpus / PV = hurdle rate / spending rate
7.0% / 3.0% = 2.333x
The difference is not a margin, a risk loading or an ask. It is a ratio of two published numbers, and the only question on the table is which of the two the organisation intends to keep its promises with.
Exercise 2.4 — Do the honest half (one session)
Before you claim, price what your change forecloses. If the durable specification removes the organisation's ability to change course at the next rebuild, that has a value: probability a better answer exists, times the net gain if it does, discounted to today. In the chapter's worked case that is £0.316m against an advantage of £8.48m — small, but present, and putting it in your own claim before anyone asks is worth more than the £0.316m ever will.
The claim that contains its own strongest objection is the claim that gets believed.
Exercise 3.1 — Propose the capitalised-gain clause (one week)
One paragraph, added to your scheme document:
Where a verified improvement removes or defers a cost arising beyond the measurement period, its value shall be computed as the annual real cost removed divided by the organisation's published real spending rate, or as the reduction in the required annual provision for a dated liability, and shall be treated as a gain of the period in which it is verified.
That is the whole change. It does not alter the share, the baseline or the verifier. It alters what counts, and it is the difference between a scheme that pays for patching and one that pays for lasting.
Exercise 3.2 — Ask for the four numbers to be published (one session)
Design life basis. Discount schedule beyond year thirty. Real spending rate. Indexation basis. You benefit from every one of them being public, because each is a number that is currently chosen inside the paper that spends it, and every one of them is currently chosen against you.
Ask in the least confrontational available form: "Is our spending rate published anywhere I can cite?" The answer is usually no, and the question usually produces one.
Exercise 3.3 — Get the baseline protected (one week)
Life-extension work has a specific ratchet problem. If your maintenance regime extends an asset's life and next year's baseline assumes the longer life, you have handed over the gain permanently and been paid for one year of it.
Ask for one clause: where a gain is capitalised, the baseline holds for the stated term rather than resetting to the improved level. Three to five years is typical. Get the term written down.
Exercise 3.4 — Write the succession note and log it (one session)
Four items for anything you will hand on: drawings in a format that will still open; materials and where they come from; skills and who still has them; money and its instrument.
Then do the thing almost nobody does: log it as a deliverable with a date. An undocumented handover is unclaimable. A logged one is a verified improvement with your name on it, and it is the only form in which knowledge-keeping has ever been paid for.
Exercise 4.1 — Get one clause into a standing document (one week)
A claim you win once is a bonus. A clause is income. Rank in this order: the scheme document; the asset policy page; the standard form of contract; the condition survey template. The lowest-status document you can reach is usually the most durable, because nobody renegotiates a template.
Exercise 4.2 — Recruit the second owner (one session)
One person carrying a long-horizon argument is a hobby. Two is a practice. Recruit whoever runs the asset register or the condition surveys — they hold the data your claim depends on, and they have never been paid for it either. Give them the credit for the first result.
Exercise 4.3 — The indexation sweep (one afternoon)
Read every long agreement you can reach for a fixed cash figure that is not indexed. This is the highest-yield hour in the workbook, it requires no authority, and one catch is worth a decade of ordinary gainshare.
Exercise 4.4 — The unsupervised improvement (ongoing, no reporting)
Some of this work will never be countable and you should do it anyway. The extra cover. The joint that can be opened. The copy of the instructions in the drawer. The label on the valve. Somebody in forty years will find it and be addressed by you.
A scheme that paid for everything would be a scheme that had measured everything, and nobody would want to work inside it. Claim what is claimable, ruthlessly and with arithmetic. Then do the rest because the work is good.
Tick what you can evidence in writing.
Fewer than five ticks is normal and is not a problem statement. It is a list of clauses, each of which is one sentence, and each of which is worth more to you than a pay round.
You: I have a verified improvement that does not fit the measurement period, and I would like to propose how it should be valued rather than argue about whether it counts.
Them: Go on.
You: The change removes £40,000 a year of real maintenance cost, permanently. Capitalised at our spending rate of 3.0 percent that is £1,333,333. At our hurdle rate of 7 percent it is £571,429. The difference between those is exactly seven over three — it is a choice of rate, not a dispute about the work. I am content with either, provided we use the same rate on the next decommissioning provision.
Them: Why would we use the lower one?
You: Because that is the rate at which we would have to fund it if we actually intended to pay for the maintenance rather than to intend to. That is what the National Trust has done since 1946 and what Norway does with its fiscal rule.
Them: And what does this cost us if you are wrong?
You: Here is the honest side. If a better answer appears at the mid-point, the durable specification forecloses it, and I have priced that at £0.316m against an advantage of £8.48m. I have also computed what it would take to overturn the case — £107.5m, which is 2.69 times the asset's own capital cost. It is not nothing; the Elwha dams cost about USD 325 million to remove after 99 years. I would like the removal cost provisioned in the same paper.
Them: What are you actually asking for?
You: One clause, and it is not about my payment. Capitalised gains counted at a published rate, and the baseline held for the stated term. Then this is a standing part of the scheme and nobody has to have this conversation again.