For the Corporation

Volume VII — Planetary and Cosmic · Extension II of III

Nine movements, one board paper.

THE PLATE

A man in a white shirt seated against a tree trunk, reading a blue book, sunlight on the water behind him. — Plate VII.E2The Appendix Nobody Reads.Every number that decided anything today was chosen by somebody, once, and written down in a place the meeting did not reach. She has gone back for it.

THE LETTER

A board can approve four things at planetary scale: a disclosure, a measurement, a covenant, and a consequence. It is routinely offered the first two, occasionally the third, and almost never the fourth — and this chapter is about the arithmetic of that gap, because the gap is where the money is.

Volume VII has already done most of the work. Chapter VII.01 grades the evidence behind the diagram every board has seen, and finds that of the six transgressed boundaries exactly 1 is Grade A while 3 are Grade C or D — 16.7 percent against 50.0 percent. Chapter VII.02 establishes the social cost of carbon as a floor with a known sign of error, because its omissions all run one way. Chapter VII.08 supplies the one disclosure that turns a long-dated liability from an assertion into an arguable number: the real return the funding is implicitly asserting. And Chapter VII.04 finds, on the high seas, that the thing everybody called a surveillance problem was a pricing problem the whole time.

That last finding is the one this chapter carries into a board room, because it is not about fish. Run Becker's line on an ordinary sustainability-linked note and it fails the same test the longliner fails, by an order of magnitude less badly and by a factor that is still greater than one in the wrong direction. The arithmetic is in the fourth movement and it is three divisions long.

So what follows is a schedule of what to stop claiming, stated as precisely as what to start measuring, each with the number that retires it; a single instrument a board can actually approve, sized by arithmetic rather than by ambition; and an honest negative that is about the instrument's own price — sized at group scale it costs 4.67 percent of EBITDA and no board will approve it, and sized at one covenant it costs 0.23 percent and clears. The difference between those two numbers is the whole of the implementation advice.

— The Editors

DISCOVERY

What is already working

Begin with what firms and ministries have already built, because it is more than the commentary admits and it is all public.

Finance ministries wrote the balance sheet first. In 2021 HM Treasury commissioned and published an argument that nature is a portfolio of assets and that the fundamental problem is a stock-and-flow one. In the same year the United Nations Statistical Commission adopted ecosystem accounting as an international statistical standard — the same status as the national accounts themselves — so that extent, condition and services now have an agreed treatment sitting alongside the accounts rather than in a supplement behind them. Neither of those is an NGO document. They are the machinery of public finance, and they arrived before most corporate reporting did.

The disclosure standard starts with a question about geography. The first letter of the Taskforce on Nature-related Financial Disclosures' core method is L, for Locate: before a firm evaluates anything it is asked where its assets and dependencies physically are. Chapter VII.01 reads that correctly and the reading is worth repeating in a board room. The ecologists' oldest objection to a global biodiversity number is that loss is local and its consequences are local. The disclosure standard's first instruction is to find out where you are. Those are the same sentence, arrived at from opposite ends of the building, and nobody had to persuade anybody.

The firm-level evidence on carbon pricing is now clean. Colmer, Martin, Muûls and Wagner matched regulated French manufacturing firms against comparable unregulated ones and found emissions 14 to 16 percent lower with no detectable contraction in output or employment and no evidence of outsourcing. The abatement came from targeted investment in emissions intensity. Whatever a board has been told about competitiveness, the best-identified study of the largest system says the trade-off it feared did not appear.

Two independent teams converged on the same damage number. Rennert and colleagues computed 185 dollars a tonne; the Environmental Protection Agency, on a different assembly of the same open-source components, computed 190. A ratio of 0.97. In a field whose public reputation is for irreconcilable numbers, two routes built by different people landed within three percent of each other, and both published their code.

And one regulator wrote a covenant for a period forty times longer than recorded history. The standard for Yucca Mountain now runs to one million years, and it does so not by choosing a very low discount rate but by not using one — a dose ceiling, a compliance test, a monitoring interval. Chapter VII.08 shows why it had no alternative. For a board, the transferable part is narrower and more immediately useful: where a duty outruns the instrument, replace the price with a standard and optimise cost subject to it.

Five things working, and the pattern under them is the one this chapter builds on: in every case somebody separated what is measured from what is judged, and put each into the statement where it belongs. That is an accounting move, not an environmental one, and it is the only move in this chapter a board has not already approved a hundred times in another context.

THE ARITHMETIC

What the board is being asked to accept, and what it costs

First, the grades, because half of the most-quoted sentence in this field rests on them.

Of 9 proposed boundaries, 6 are declared transgressed. Chapter VII.01 grades them on three tests — is the system well-mixed, is the control variable directly measured, is the threshold demonstrated rather than inferred — and the answer is that 1 of the six is Grade A and 3 are Grade C or D. 16.7 percent against 50.0 percent.

That is not an argument from suspicion, and it can be checked from the framework's own published tables. Take each boundary's uncertainty band and divide it by the boundary value it hangs off:

  ozone                                        5.3 %
  aragonite saturation                        12.4 %
  CO2 concentration                           28.6 %
  radiative forcing                           50.0 %
  biosphere intactness                        66.7 %
  extinction rate                            900.0 %
  ----------------------------------------------------
  climate and chemistry, mean                 24.1 %      median  20.5 %
  biosphere integrity, mean                  483.4 %
  the ratio, on means                          20.08 x
  the ratio, on medians                        23.58 x

About twenty times either way, which is the point of printing both. A band nine times wider than the boundary it surrounds is a published statement that nobody knows, and it is to the authors' credit that they printed it. And 6 of the 9 have had their boundary value or control variable revised since 2009 — 66.7 percent of the framework — while the three that have never been revised are exactly the three resting on a demonstrated global threshold in a well-mixed system.

The board consequence is one rule: grade before you consolidate, and never covenant on a Grade D. A covenant written against a number whose published band is 900.0 percent of the number is unenforceable, both parties know it, and writing one buys nothing and costs credibility.

Second, the ratchet, measured against the thing it claims to price.

The market's existing answer is the sustainability-linked loan, and the market-norm margin ratchet is 25.0 basis points. On a drawn facility of 400,000,000 dollars that is 1,000,000 dollars a year, against a covenant whose breach makes 400,000,000 repayable:

  the ratchet as a share of the consequence     0.25 %
  out by a factor of                              400 x

A ratchet is not a price on a threshold. It is a rounding error wearing one.

Third — and this is the cut — your sustainability-linked note is the longliner.

Chapter VII.04's arithmetic is Becker's, and it is one line: an operator complies when p · F ≥ G, where p is the probability a violation is both detected and sanctioned, F is the penalty, and G is the gain from violating. On a representative distant-water tuna longliner, at 5 percent observer coverage and 50 percent sanction follow-through against a penalty of 100,000 dollars and a gain of 180,000:

  p  =  0.050 x 0.50                         0.0250
  expected penalty                          $ 2,500
  p . F / G                                   0.014
  the violation pays                             72 x what it costs

Now run the identical sum on a board's own instrument. A 2,000,000,000-dollar ten-year note at a coupon of 5.00 percent, with a 25.0 basis-point two-way step, verified annually by a named verifier. The step is 5,000,000 dollars a year, 50,000,000 undiscounted, and at an annuity factor of 7.7217 its present value — which is F — is 38,608,675 dollars. Put the probability of detection and step at 0.90, which is generous, and the gain from missing the target at a present value of 250,000,000 dollars of avoided abatement capital:

  p . F / G, the note                         0.139
  better than the longliner by                10.01 x
  and short of deterrence by                   7.19 x

Ten times better than the tuna boat, and still an order of magnitude short. Then invert it exactly as Chapter VII.04 does. What verification probability would deter at this step?

  p*  =  G / F  =  250,000,000 / 38,608,675  =  6.475

Greater than one. There is no assurance regime of any frequency — not a quarterly verifier, not a continuous data feed, not a second auditor — that deters this, because certainty is the ceiling and certainty is not enough. And the other way round, what step would deter at the verification we actually have? F* = G / p is 277,777,778 dollars of present value, which over the tenor is 35,973,493 a year, which on the note is:

  the deterrent step                          179.9 bp
  against the market norm                        7.19 x

One hundred and eighty basis points, not twenty-five. Forty years of ocean governance was argued as though the high seas were ungovernable because they are too big to watch; Chapter VII.04 shows it was unenforced because the fine was too small. The corporate version of that sentence is: a board that responds to a weak covenant by buying more assurance has bought coverage where the missing term was a consequence, and the two are not substitutes at any price.

Fourth, the floor with a known sign of error.

Chapter VII.02's instruction is to quote the triple, never the point. The Environmental Protection Agency's estimates are 120 dollars at a 2.5 percent near-term rate, 190 at 2.0, and 340 at 1.5, and the agency states in its own report that data and modelling limits implicitly assign a value of zero to the omitted damages. A quantity whose omissions all run one way is not an unknown. It is a bound with a known direction.

On the worked borrower — 250,000 tonnes of scope 1, 0.60 of it under a compliance scheme at 70.40 dollars, against EBITDA of 120,000,000:

  cash carbon cost                       $10,560,000 /yr      8.8 % of EBITDA

  at $120/t   exposure $30,000,000       25.0 % of EBITDA     unpriced $19,440,000
  at $190/t   exposure $47,500,000       39.6 % of EBITDA     unpriced $36,940,000
  at $340/t   exposure $85,000,000       70.8 % of EBITDA     unpriced $74,440,000
  -------------------------------------------------------------------------------
  the same firm, low to high                                          2.83 x

One parameter nobody in the room chose moves the same firm's exposure from a quarter of EBITDA to seven tenths of it. That is the reason a social cost of carbon quoted without its discount rate is a sentence fragment rather than a figure — and it is also why the cash cost, at 8.8 percent, is managed by procurement while the exposure, at 39.6 percent, is managed by nobody.

Fifth, the disclosure that turns a funding ratio into an argument.

A funding ratio is read aloud in board meetings as though it were a fact. It is a discount rate wearing a percentage sign. Chapter VII.08's reverse disclosure fixes it in one line: state the real return the funding is asserting, beside the rate the provision uses.

Take a decommissioning duty of 1,200,000,000 dollars undiscounted over 40 years — a level schedule of 30,000,000 a year — against assets of 250,000,000:

  the sentence a board hears                   20.8 % funded
  implied real return, perpetuity form        12.00 %
  implied real return, annuity form           11.86 %      the two agree to 1.011 x
  realised real return of such funds           4.00 %
  the restatement gap                           7.86 points
  corpus required at the realised rate    $593,783,217
  the shortfall                           $343,783,217

"Twenty percent funded" is not a fact. It asserts 11.86 percent real for forty years against a fund realising four. That is the sentence the paper should open with, and the two routes — a perpetuity, which is the convention Chapter VII.08 prints, and a forty-year annuity, which is the honest form for a finite duty — are computed separately and agree, which is what makes either believable.

Sixth, what your successor will do, correctly.

Chapter VII.08 measures the one thing that defeats a long-dated corporate commitment, and it is not bad faith. Standing at year thirty of a two-hundred-year plan, the original plan's own continuation value is 17,423 pounds and the successor, re-applying the identical published table from its own present, values it at 11,244:

  the successor values it at                   64.5 %
  the plan over-states by                       1.550 x

Nobody has cheated. A horizon-indexed schedule re-bases with every board, so a commitment you do not entrench is repriced down by a third on schedule, by people applying your own table. That is an argument for the covenant, the trust and the calendar-form disclosure, and against the memorandum of intent.

Seventh, what to stop claiming, each with the number that retires it.

Stop claiming alignment with the planetary boundaries as a set. 16.7 percent of the transgressions are Grade A and 50.0 percent are Grade C or D. Claim the Grade A line and disclose the rest.
Stop reporting a group total for a local flow. 190.0 megatonnes of nitrogen over 1,600 megahectares is 118.75 kilograms a hectare spread evenly and 593.75 concentrated on a fifth — 5.0 times — and both worlds report the same total. One has anoxic estuaries and one does not.
Stop discharging an inside-boundary tonne with an offset. 85 percent of Clean Development Mechanism projects and 73 percent of potential supply had a low likelihood of being additional, against 2 and 7 percent with a high likelihood. Of 62,000,000 forest credits issued, about 6 percent were genuinely additional — 3,720,000 real tonnes — and of the 14,600,000 already used to offset, 13,724,000 tonnes of claimed offsetting did not occur. An offset that is not additional is an emission with a receipt.
Stop reporting a funding ratio as a fact. One public pension liability moved from 2,631 billion pounds to 1,375 billion in twelve months — a movement of 1,256 billion, 47.7 percent of the opening figure — and not one promise was changed.
Stop quoting an internal carbon price without its basis. A quarter of firms set one below 20.00 dollars, which is 10.5 percent of the central social cost and below the month-to-month movement in the fuel prices they already manage.

And start measuring four things, in this order: locate every material asset and dependency by site; grade every metric already in your reporting; publish the implied required real return on every long-dated duty; and compute the expected-consequence ratio on the covenants you have already signed. The first three are a quarter's work by people you employ. The fourth is three divisions.

Eighth, the honest negative, and it is about this chapter's own instrument.

Size the consequence properly at group scale and it is unaffordable. F* was 277,777,778 dollars; at a surety premium of 2.00 percent of face, a 280,000,000 face costs 5,600,000 dollars a year — 4.67 percent of EBITDA. No board approves that, and no board should be asked to.

Size it to one Grade A covenant with a bounded gain and it clears. At a gain from missing one physical covenant of 12,000,000 dollars and the same 0.90 verification, F* is 13,333,333, a face of 14,000,000 costs 280,000 a year — 0.23 percent of EBITDA — and the expected-consequence ratio comes out at 1.050. The two premiums differ by 20.0 times.

The sequence therefore follows from the ratio and not from ambition: one Grade A covenant, or nothing. A firm that takes the group frontier to its board first will have the affordable version refused afterwards, and the refusal will be correct on the evidence it was shown.

There is a second negative and it is about the benefit. The graded schedule and the implied-return disclosure are a cost with a governance benefit and no priced benefit this chapter can compute. Nobody has demonstrated that a firm publishing an honest implied required return is rewarded in its cost of capital, and this chapter will not pretend otherwise. The exit is sequencing rather than argument: do them where a lender is already asking for them, at which point the cost is already sunk and the disclosure is the cheapest half of a conversation you are having anyway.

DREAM

What becomes ordinary

In the firm that has absorbed this, the annual pack has one appendix everybody reads, and it is the schedule at the back with the assumptions in it.

Every physical metric on the face of the report carries its evidence grade in the same typeface as the number. Nobody finds this strange; it is the same discipline applied to a change in accounting estimate. A Grade A line is covenanted, tested quarterly, and its breach clause is in the borrowings note with the rest of the covenant schedule rather than in a sustainability section — because putting it in a sustainability section is what makes it optional. A Grade D line is disclosed and not tested, and the disclosure carries its published band beside it, so a reader can see at once that the band is wider than the boundary.

Every long-dated duty publishes the real return its funding is asserting. It sits on the first page of the trust's report, not the funding ratio, and it can be compared to something — the fund's own realised return over twenty years — which is what makes it usable. A trust asserting 11.86 percent real forever is asked about it in the same tone as an unusual depreciation life, and the asking is routine rather than hostile.

Every covenant the firm signs carries its expected-consequence ratio on the front page. Not the target, not the verification frequency, not the assurance provider — the ratio. Somebody in the room knows that a 25.0 basis-point step is 0.25 percent of the consequence it claims to price, and says so, and the conversation moves to what would make it bite. The answer is usually a bond rather than a bigger number, because a bond is already in hand and a step is a promise about a future coupon.

The carbon column is in every capital paper, at three prices, with the discount rate stated. The board decided the basis once, minuted it, and put a review date on it. A parameter is restful; a position has to be defended every time.

And there is one thing on the register with no number at all. The items nobody can measure reliably are listed, named, and explicitly not totalled — because a zero is a claim and a blank is not, and the firm would rather be seen to have an unmeasured line than be caught having quietly set it to nothing.

DESIGN

The four statements a board can actually approve

Chapter VII.01 supplies the taxonomy and it is the whole design. Every environmental exposure a firm holds is one of four things, and putting them on one face is a category error with a price.

A covenant — a level test with a step loss function. You are inside it or you are not; there is no version in which being slightly over costs slightly more. One to three Grade A metrics, tested quarterly, with a defined breach clause and an equity cure. This is the only class that carries a step, and it is the only class the instrument in the last movement bonds.

A register — an asset with a carrying value and an impairment test. Proportional loss gets a proportional consequence. Forest extent by biome, root-zone soil moisture by basin, land condition by holding, with gross, accumulated impairment and net. The treatment already exists for property, plant and equipment and for living assets, and the location layer is built once under the Locate step and then reused, because the location data is the expensive part and it is a fixed cost paid once.

A flow account, at the place of effect. Nitrogen, phosphorus, water and aerosols by catchment or airshed against that place's own allowable input. Never a group total. The 118.75 against 593.75 is the reason, and it is one division.

A disclosure schedule for everything not measurable enough to recognise. The provisions standard already distinguishes a provision, recognised because it can be measured reliably, from a contingent liability, disclosed precisely because it cannot. A band of 900.0 percent fails the measurement-reliability test by a distance. Disclosure is not demotion — a contingent liability is a live, legally consequential statement read by exactly the people who need to read it.

Then two governance moves that cost nothing and decide everything.

Grade before you consolidate, and never sum across grades. A total spanning a covenant and an expert judgement is not a total. When somebody asks for one number for the cover — and they will, and they are not being foolish, because a cover needs a number — give them the covenant schedule's worst headroom, which is a real number about a real threshold, and refuse the composite.

Fix the consequence before buying the coverage. The order is F first, then p, and it is not a preference. At a 25.0 basis-point step the required detection probability is 6.475, so no amount of assurance closes it. Coverage is procurable and consequences are political, which is exactly why the procurable one gets bought first, and why the order has to be written into the policy rather than left to the budget cycle.

Sequence. Locate, then grade, then covenant one Grade A metric, then bond it. Each step is cheaper than the one after it and each produces the evidence the next one needs. A firm that starts at the bond has nothing to bond; a firm that stops at the grading has a very good appendix and no consequence anywhere in the system.

DESTINY

How it holds when the sponsor is promoted

Three things keep this alive and all three are structural rather than motivational.

The number is filed against a formula the firm did not choose. The nuclear decommissioning regime gets one thing unambiguously right: a licensee reports biennially, publicly, against a published formula minimum. A firm cannot quietly drift against a number somebody else set. Where a regulator publishes a floor, use it; it is free, defensible, and nobody in the building gets to argue with it.

The duty and the money are held by different institutions. The German operators transferred the fund and the state took the liability, because a utility is not an institution that can be relied on to exist a century from now. Separating the party that benefits from the party that owes is the oldest device in trust law and it is the one most often skipped.

A counterparty loses money if the reading is wrong. This is the least romantic of the three and the most durable. A regime maintained only by officials decays at the speed of a posting cycle. An underwriter with money at risk against the firm's conduct reads the data on the days nobody cares, which are most days — and creates, without anybody legislating it, a paid private monitor with a balance-sheet interest in accuracy.

Now the failure modes, named.

It fails by grade inflation, because A is the grade that gets a meeting. The defence is that the three tests are written down, a named person applies them, and the grades are auditable against the framework's own published bands.

It fails by grade deflation, which is more common in practice: everything is pushed to D so that nothing has to happen. The same arithmetic runs the other way. Climate's band is 28.6 percent of its boundary and ozone's is 5.3, and no honest reading puts either in the same class as 900.0.

It fails by re-aggregation, when somebody builds a composite index across four evidence grades and puts a decimal point on it. That is the diagram again, with a decimal point.

It fails when the covenant is written as a reporting covenant rather than a financial one and then never stepped up. Expect the first lender to want exactly that; it is a reasonable place to start and an unreasonable place to stop, and the term to fight for is the second-facility step-up once a compliance history exists.

And it fails — most expensively — when the firm buys assurance in place of consequence, year after year, each purchase defensible and the ratio never moving. p rises from 0.90 to 0.95 and the expected-consequence ratio moves from 0.139 to something still far below one, and the pack gets thicker, and nothing changes. The ratio is the thing to put on a standing agenda, precisely because it is the number that does not improve when you spend money on the wrong term.

DELIGHT

What it feels like

There is a particular relief in a document that says what it does not know.

Anyone who has sat through a presentation of nine numbers of nine different qualities, all drawn at the same weight, knows the specific fatigue of it — the low-grade suspicion that you are being asked to accept the weakest line in order to get the strongest, and the resentment that follows. A statement that grades itself takes that away entirely. You can put your weight on the graded lines and let the rest be exactly what they are, and nobody has to pretend.

Then the better pleasure, which arrives in a negotiation. The covenant schedule goes in front of somebody who has spent twenty years reading covenant schedules, and none of it needs explaining. They read the headroom, they read the trend, they ask one question about the breach clause, and the conversation is technical inside ninety seconds. No translation, no persuasion, no prior agreement about what kind of world this is.

And the quietest one, which is about the ratio. Somebody in the room divides three numbers on the back of the agenda and says the step is 0.25 percent of the thing it prices. Nothing dramatic happens. The paper goes back, and comes again a month later with a bond in it, and the person who did the division watches it arrive and says nothing — which is the correct response and the most satisfying one available in corporate life.

OPERATIONALIZE THIS

At the level of finance

The instrument: a consequence-sized compliance surety on one Grade A covenant, with the graded exposure schedule as its appendix.

The precedent is not environmental. It is wreck removal and oil pollution, where the industry long ago stopped relying on judgments against foreign-flagged vessels and moved to collateral already in hand. Chapter VII.04 applies it to catch; this applies it to one physical covenant a firm has signed.

The mechanics.

What is bonded. One Grade A metric. Well-mixed or single-site, directly measured, with a demonstrated threshold and a defined breach. Never a Grade C or D item: a bond on a number with a band of 900.0 percent is a licence fee.
Face, set by arithmetic and not by negotiation. F* = G / p, where G is the firm's own gain from breach — the capital or operating cost avoided — and p is the probability a breach is detected and determined under the agreed verification. On the worked covenant that is 13,333,333 dollars, rounded up to a face of 14,000,000.
Premium. A surety on a clean risk prices at one to three percent of face. At 2.00 percent: 280,000 dollars a year, 0.23 percent of EBITDA.
Trigger. Forfeiture on a determined breach, resolved from the verification record rather than by litigation. Released annually against a clean record, which is what makes it an asset the firm is building rather than a charge it is paying.
The counterparty. The surety, not the regulator and not the lender. This is the elegant part: the underwriter now has money at risk against the firm's conduct, so the underwriter reads the data. You have created a paid, private, permanent monitor with a balance-sheet interest in accuracy, and you did not have to legislate one or hire one.
The appendix. The graded exposure schedule — every metric with its grade, its band, its class and, for Grade A, its headroom and breach clause. It is the evidence the surety underwrites against and the document the board approves.

The balance-sheet treatment. The premium is an operating expense. Where cash collateral is posted it is restricted cash — an asset, not an expense — and specifically not a provision, because a provision can be released by an estimate and restricted cash cannot. Absent a present obligation there is no liability to recognise, and a firm that books one will be corrected; the honest treatment is the appropriated reserve plus disclosure that Chapter VII.02 sets out. Take it to the auditors in the first month rather than the ninth: it is a disclosure conversation, not a recognition one, and disclosure conversations are quick.

The number that decides it. One line, on the front page:

            expected consequence          p · F
          ------------------------  =  ----------   >   1.00
            gain from the breach            G

  today, unbonded, at a 25 bp step               0.139
  bonded, one covenant                           1.050

If that ratio is below one, every other line in the pack is decoration. And the total cost of moving it above one — the premium plus the assurance and location data at 340,000 dollars a year — is 620,000 a year, 0.52 percent of EBITDA, against a pairing that Chapter VII.01 already prices: avoided impairment of 2,400,000 and a margin benefit of 1,000,000 against verification and a restoration programme, returning 213.8 percent and a net 1,810,000 dollars a year.

The first ninety days.

Day	Action	Artifact
1–15	Locate: every material asset and dependency, by site	The location layer
16–30	Grade every metric already in your reporting against the three tests	The graded register
31–45	Compute the expected-consequence ratio on every covenant already signed	The ratio schedule
46–55	Publish the implied required real return on every long-dated duty	The implied-return page
56–70	Choose one Grade A covenant; estimate `G`; agree `p` with the verifier	The sizing memo
71–80	Surety indication at the frontier face; auditors on the treatment	Quotation and accounting note
81–90	Board approves the face, the trigger and the review date	The signed board minute

What to refuse. A covenant on a Grade D metric. A face set by negotiation rather than by G / p. An assurance upgrade offered in place of a consequence. A funding ratio presented as a fact. And a composite index across four evidence grades, however elegantly it is drawn.

APPRECIATIVE QUESTIONS

Twelve, for a room

Discovery — what is already working

Where in our reporting do we already grade our own evidence — formally or informally — and who was it who insisted on that?
Which of our long-dated duties already has money segregated against it, and who set the amount, and against what?
When has a covenant we signed actually changed a decision here? What was in it that made it bite?

Dream — what becomes possible

If every number we published carried its evidence grade in the same typeface as the number, what would change about which numbers we chose to publish?
Imagine our long-dated funds each printing the real return they are asserting on their first page. What conversation would the audit committee have next year that it cannot have now?
If an underwriter had money at risk against our conduct on one physical metric, what would they ask to see — and what would we be glad to show them?

Design — what we build

Which of our metrics is a covenant, which is a register, which is a flow — and where have we been managing one as though it were another?
What is the single Grade A exposure we would be willing to bond this year, and what is our honest estimate of G?
Where are we currently buying assurance when the missing term was a consequence, and what would the ratio be if we computed it this week?

Destiny — how it holds

What would have to be true for the graded register to survive the first person who wants one number for the cover?
If a future board wanted to undo this quietly, what is the most likely route — and what disclosure would make that route visible instead?
Which of our funded duties has money but no delivery covenant, and who would notice first if the money were there and the thing never happened?

WORKS CITED

Becker, G. S. (1968). "Crime and Punishment: An Economic Approach." Journal of Political Economy, 76(2), 169–217.

Cames, M., Harthan, R. O., Füssler, J., Lazarus, M., Lee, C. M., Erickson, P. and Spalding-Fecher, R. (2016). How Additional Is the Clean Development Mechanism? Öko-Institut, study prepared for DG CLIMA, European Commission.

Colmer, J., Martin, R., Muûls, M. and Wagner, U. J. (2024). "Does Pricing Carbon Mitigate Climate Change? Firm-Level Evidence from the European Union Emissions Trading System." The Review of Economic Studies, rdae055.

Dasgupta, P. (2021). The Economics of Biodiversity: The Dasgupta Review. HM Treasury, London.

Environmental Protection Agency (2008). Public Health and Environmental Radiation Protection Standards for Yucca Mountain, Nevada; Final Rule. 40 CFR Part 197.

IFRS Foundation. IAS 16 Property, Plant and Equipment · IAS 37 Provisions, Contingent Liabilities and Contingent Assets · IAS 41 Agriculture · IFRS S2 Climate-related Disclosures.

Nuclear Regulatory Commission. 10 CFR 50.75, Reporting and recordkeeping for decommissioning planning. United States.

Rennert, K., Errickson, F., Prest, B. C. et al. (2022). "Comprehensive Evidence Implies a Higher Social Cost of CO₂." Nature, 610, 687–692.

Richardson, K., Steffen, W., Lucht, W. et al. (2023). "Earth beyond six of nine planetary boundaries." Science Advances, 9(37), eadh2458.

Steffen, W., Richardson, K., Rockström, J. et al. (2015). "Planetary boundaries: Guiding human development on a changing planet." Science, 347(6223), 1259855.

Sutinen, J. G. and Andersen, P. (1985). "The Economics of Fisheries Law Enforcement." Land Economics, 61(4), 387–397.

Taskforce on Nature-related Financial Disclosures (2023). Recommendations of the Taskforce on Nature-related Financial Disclosures. TNFD, September 2023.

United Nations et al. (2021). System of Environmental-Economic Accounting — Ecosystem Accounting (SEEA EA). Adopted by the UN Statistical Commission.

United States Environmental Protection Agency (2023). Report on the Social Cost of Greenhouse Gases: Estimates Incorporating Recent Scientific Advances. EPA, Washington DC.

West, T. A. P., Wunder, S., Sills, E. O. et al. (2023). "Action Needed to Make Carbon Offsets from Forest Conservation Work for Climate Change Mitigation." Science, 381(6660), 873–877.

Weitzman, M. L. (1998). "Why the Far-Distant Future Should Be Discounted at Its Lowest Possible Rate." Journal of Environmental Economics and Management, 36(3), 201–208.

HM Treasury. Whole of Government Accounts, years ended 31 March 2021 and 31 March 2022.

Note on figures. Every figure in this chapter is computed in lib/verify/VII_E2.py and printed there with its inputs, its units and its source by python3 lib/verify.py VII.E2. Figures carried from VII.01 to VII.11 are recomputed here rather than quoted. The borrower's EBITDA, the gain from missing a covenant, the verification probability, the surety rate and the decommissioning schedule are stated assumptions, labelled as such, and every conclusion is stated as a ratio so that it survives a different set of them. The implied required real return is computed twice, by a perpetuity and by a forty-year annuity, and both are printed rather than reconciled silently.