Haute Lumière
Commerce · III.02 · MMXXVI · daylight
For the executive with a P&L, a board, and a capital committee. Applied to a business unit. The language of the firm without apology, and the places where the firm's own numbers already support the move.
Your business already holds an unhedged position in energy that does not appear as a position anywhere in your accounts. It appears as a cost line, which is a different thing: a cost line is a historical fact, and a position is an exposure with a distribution attached.
The 2022 gas year made this expensive for a great many firms who believed they were hedged and discovered they were hedged against the wrong index. TTF gas peaked near €339.20 per megawatt-hour in August 2022 while Henry Hub sold at about $8.80 per MMBtu — $30.03 per megawatt-hour, €29.39. A factor of 11.54 on the same molecule in the same month. Any hedge constructed on the assumption that gas is gas was, for that month, a hedge against a different commodity.
The proposition of this workbook is narrow and commercial: add an energy column to the accounts you already produce, weight it by exergy, index it to time and place, and use it to size a position you are already carrying. It requires no new philosophy and no change in reporting standard. It requires a conversion table, a meter, and one signature.
Exercise 1.1 — The energy line, disaggregated (one week, finance + facilities)
Take last year's energy spend for one site or one unit and split it four ways:
| Quantity (MWh) | Spend | Implied £/MWh | |
|---|---|---|---|
| Electricity, on contract | |||
| Electricity, out of contract or pass-through | |||
| Gas or other fuel, on contract | |||
| Fuel, out of contract |
Then the question that produces the finding: what share of the quantity was bought at an unhedged price, and what share of the cost did it represent? In almost every firm these two percentages differ sharply, and the gap is the position nobody has sized.
Exercise 1.2 — Where you already settle in energy (half a day)
You are more energy-denominated than you think. Look for:
Write the list. It is the evidence that this is an extension of existing practice rather than an innovation, and that distinction is worth a great deal in front of an audit committee.
Exercise 1.3 — The appreciative question, asked upward (45 minutes)
Ask your operations leadership, in these words:
"When has knowing the physical quantity rather than the price actually made us money here? What made that possible, and what would it take to have more of it?"
Record conditions, not anecdotes. You are assembling the repeatable causes.
Exercise 2.1 — Your intensity, benchmarked honestly (2 hours)
Compute energy per unit of gross value added for your unit, and compare it to the national figure for where you operate. For reference, in megajoules per 2017 PPP dollar: Ireland 1.6, United Kingdom 2.6, Germany 3.0, Japan 3.4, United States 4.6, world 4.6, China 6.4, Canada 6.9, Russia 8.2, Iceland 15.6.
The widest spread is 9.75 times, and it is mostly structure. Iceland smelts aluminium with stranded geothermal; Ireland's GDP carries profits of firms burning energy elsewhere. Strip the artefacts and the defensible national spread is 3.15 times. If your unit differs from its national benchmark by less than that, you have found nothing; if it differs by more, you have found something and should establish which way before anyone presents it.
Exercise 2.2 — Exergy-weight the consumption (half a day)
Weight every stream by what it can actually do, against a stated 20 °C ambient:
| Grade | Factor |
|---|---|
| Electricity, shaft work | 100.00% |
| Process steam, 250 °C | 43.96% |
| Hot water, 80 °C | 16.99% |
| Hot water, 40 °C | 6.39% |
Worked, at the scale of the instrument below: of 14.0 GWh, if 60 percent is electricity and 40 percent is 80 °C heat, the exergy-equivalent is 14.0 × (0.60 × 1.00 + 0.40 × 0.1699) = 9.35 GWh. The flat-joule figure overstates the tradeable quantity by 49.71 percent.
This one calculation usually reorders an efficiency pipeline, because projects that save low-grade heat and projects that save electricity have been sitting in the same ranking as though a joule were a joule.
Exercise 2.3 — Find the 409 (one week, needs interval data)
Join a year of your half-hourly consumption to a year of day-ahead or real-time prices and compute the share of annual cost sitting in the most expensive 1 percent of hours.
Scale, so the exercise is not abstract: in ERCOT the annual average real-time price was around $22 per MWh in 2020; in February 2021 it sat at the $9,000 offer cap for roughly 96 hours — 409.1 times the average, against a floor of negative $251, a span of $9,251 per MWh. One megawatt held at the cap for those hours earns $864,000 against $2,112 at the annual average. The cap was subsequently reset to $5,000, which is still 227.3 times the average.
Exercise 2.4 — Restate your emissions on a consumption basis (one week)
If you report Scope 1 and 2 only, you are reporting the territorial account of your own firm. Run the Scope 3 categories that dominate, and put both series on one page.
The national analogue is the scale check: 6.2 GtCO₂ was embodied in international trade in 2004 — 22.96 percent of global fossil emissions. A UK-shaped 52.6 percent territorial cut becomes a 28.40 percent consumption cut on the bounding assumption that imports did not move: 54.00 percent of the headline survives. Expect a similar shape inside a firm that has outsourced its manufacturing, and expect to be asked about it.
The kWh tranche. A facility advanced in currency and repaid out of verified kilowatt-hours converted to cash at a contracted index.
Worked, at a scale that fits one signature. A site with a 120.0 GWh annual load. A £4,000,000 retrofit verified to save 14.0 GWh a year — 11.67 percent of load.
annual value of saving 14,000 MWh x £95.00/MWh = £1,330,000
assigned to repayment x 75% = £997,500
simple payback £4,000,000 / £997,500 = 4.01 years
simple annual return £1,330,000 / £4,000,000 = 33.25%
WACC 9.0%
The number that decides it, and it belongs on the front page of the paper:
breakeven = £4,000,000 / (5 x 14,000 x 0.75) = £76.19 / MWh
Above £76.19 the facility repays. Against a £95.00 index the headroom is £18.81, or 24.7 percent. At a £70.00 forward price it recovers 91.9 percent of capital over the term and fails.
What that calculation has actually produced is a strike price. An efficiency proposal has become a position with a floor, which is a thing your treasury already knows how to hold, hedge and report. That is the entire commercial argument for the instrument, and it is the sentence to put in front of the CFO.
Debt service and covenant. At a 6.0 percent coupon over five years the annual service is £949,586. At a 75 percent share, DSCR is 1.050 — too thin for external finance. At 80 percent, repayment is £1,064,000 and DSCR is 1.120. Negotiate the floor price, not the coverage ratio, because the ratio is a function of it.
Balance sheet. The facility is a liability at amortised cost; the retrofit capitalises to the improved asset and depreciates over its improved life. The energy-indexed repayment is an embedded derivative — raise it with the auditors at term-sheet stage, not at year end. Where the index is your own procurement price the hedge relationship is usually straightforward; where it is a market index you do not buy at, it is not.
Counterparty. Treasury to business unit first. Two completed internal facilities give you a track record; only then take it to the ESCO market or the sustainability-linked desks, both of which will ask for the M&V protocol by name.
Get it into the pack. Anything reviewed monthly persists. One line: energy quantity, exergy-weighted quantity, price paid, and the share of cost in the top 1 percent of hours. Four numbers, one row, every month.
Put the conversion table under governance. It needs an owner with no position in the trades — the same function that owns transfer pricing. Version it, date it, and restate history when it changes. Emergy's literature paid 1.271 on every published figure for learning this late; you can learn it in a meeting.
Name a second owner. One executive is a preference; two is a policy.
The delight, stated commercially. The pleasure in this is real and it is worth naming to your team: it is the pleasure of finding that a fixed cost was a timing decision nobody had made. Somebody looks at a load profile, finds the plant running through the four most expensive hours of the year, moves it, and the saving arrives with no capital, no vendor and no business case. That happens more often than it should, and it is the cheapest money in the building.
The reversion clause, and why it is worth more than it costs. Until the facility repays, £997,500 of the £1,330,000 annual saving services capital and £332,500 reverts to the operating unit. After repayment the whole £1,330,000 reverts. That reversion costs the balance sheet nothing — the capital is already recovered — and it is what converts the operating unit from a reluctant host into a genuine beneficiary. Write it into the term sheet at the start. It is free then, and it is contested in year five.
What to say when someone asks whether this is a sustainability initiative. It is not, and saying so is the stronger position. It is a floor-price position on capital you were going to deploy anyway, measured in a unit that does not need a vintage. The energy column is a measurement improvement of the kind your controllers make every year; the tranche is a financing structure your treasury already recognises from energy performance contracting. Nothing here requires a change in standard, mandate or belief — which is precisely why it survives a change of sponsor.
Someone proposes settling in energy. The tell is a proposal that quietly drops transferability or the ability to hold the unit across a period. That is Technocracy's design, which scored 1 of 3 on the jobs of a numeraire — 33.3 percent — and the answer is the arithmetic: 409.1 times, one market, one year.
The conversion table moves silently. Three years of comparison become meaningless and nobody notices for two more. Version, date, restate.
The boundary creeps. Widen the boundary annually and the account improves annually with nothing physical changing. Treat the boundary as a disclosure and a change to it as a restatement.
The metering is too coarse. Below interval metering the account degrades to an annual aggregate, which cannot see the 409 and will report stability through a year in which the exposure moved by three orders of magnitude. The right answer is to fix the metering and say plainly that the account is not yet available — not to publish a smoother number.
The baseline is agreed after deployment. This is the one that produces disputes. Sign it first.
| Day | Action | Artifact |
|---|---|---|
| 1–15 | Disaggregate the energy line; find the unhedged share | Four-way table |
| 16–30 | List where you already settle in energy units | The precedent list |
| 31–45 | Exergy-weight consumption; publish the conversion table | Table v1.0, dated |
| 46–55 | Join interval data to prices; compute the top-1% cost share | Load-and-price study |
| 56–65 | Agree and sign the energy baseline with finance | The signed baseline |
| 66–75 | Compute the breakeven price; test against the forward curve | One-page term sheet |
| 76–85 | Settle the embedded-derivative treatment with audit | Accounting memo |
| 86–90 | Board paper; one number on the front page | £/MWh breakeven |
Title. A floor price for £4.0m of efficiency capital.
Recommendation. Approve a £4,000,000 facility repaid from 75 percent of verified savings over five years, at a breakeven energy price of £76.19/MWh against a current index of £95.00.
One paragraph of context. The unit consumes 120.0 GWh a year. The retrofit is verified to save 14.0 GWh, 11.67 percent of load, worth £1,330,000 a year at the contracted index — a 33.25 percent simple return against a 9.0 percent WACC.
The risk, stated as a number. The facility fails below £76.19/MWh. At £70.00 it recovers 91.9 percent of capital over the term. Headroom against the current index is £18.81/MWh, 24.7 percent. Recommend a floor-price covenant rather than a DSCR covenant; at a 75 percent share DSCR is 1.050 and at 80 percent it is 1.120.
What is new in the reporting. An energy column, exergy-weighted, versioned and dated, published beside the money column in the monthly pack. It does not settle anything. It measures something the money column cannot: the same quantity across years, currencies and price regimes.
Decision requested. One signature, inside existing delegated authority.