Haute Lumière
Commerce · I.06 · MMXXVI · daylight
TEN CONCEPT BRIEFS · Chapter I.06 — Bringing the Money With You
One page each. A reader who reads only these ten pages has the chapter.
BRIEF 1 — The Stack
The idea. Capital is not one thing you get or fail to get. It is three or four kinds of money placed in an order, and the order decides whether any of it arrives.
first loss / grant 10% absorbs losses first, expects 0%
subordinated, patient 20% absorbs next, wants 1-3% and duration
senior, commercial 70% absorbs last, wants market rateThe senior lender is not underwriting your project. It is underwriting its own attachment point — the share of the structure that has to be wiped out before it loses a penny. With ten percent of first loss and twenty percent of subordination beneath it, a senior lender attaches at thirty percent. A project it would decline standing alone becomes an ordinary credit inside a stack.
Worked example. A $10 million regenerative agriculture book. Standing alone, the bank declines: no comparable loss history, unfamiliar collateral. With a $1 million reserve and $2 million of subordinated notes underneath, the bank is lending $7 million against $10 million of assets with $3 million of cushion. Same book, same farmers, same week. Different answer.
Why it matters. Almost every "we could not raise the money" conversation is actually a structure problem being reported as a demand problem.
You already know this because you have watched a house purchase fall through for want of a deposit, and understood immediately that the deposit was never about the money — it was about who takes the first loss if the price falls.
BRIEF 2 — First Loss, and What It Actually Costs
The idea. A first-loss dollar is not spent. It is posted, and most of it comes back.
Everybody prices a first-loss tranche at its face. Price it at its expected loss instead and the arithmetic inverts.
pool $10,000,000
first-loss tranche, 10% $1,000,000
sector net charge-off (OFN) 0.58% a year
average life 3 years
expected cumulative loss $174,000
returned at maturity $826,000
---------------------------------------------------------
expected cost per dollar mobilised 1.74 cents
the same dollar as a grant 100.00 centsFifty-seven times. That is the ratio, and it is a division, not a claim.
The boundary, which is what makes it usable. The tranche is exhausted at 5.7 times the sector's historic loss experience. Below that the senior lender is untouched; above it, the structure fails in the year it was most needed. And two costs never appear in the 1.74 cents: the provider owns the variance, not the mean, and the money is illiquid for the term. A funder who could not survive losing the whole tranche should not write one.
Why it matters. It converts the ask. You stop asking a foundation to spend and start asking it to place — and those two requests go to different committees with different mandates and different answers.
You already know this because you have guaranteed a friend's lease or co-signed a loan. You did not hand over the rent. You stood behind it, and in almost every case you were never called.
BRIEF 3 — The Residue Clause
The idea. One sentence turns a gift into a placement: whatever the reserve does not absorb returns to its provider at maturity.
Why it is load-bearing. A grant leaves a foundation's balance sheet and is expensed. A reserve with a residue clause stays on the balance sheet as an asset carried at face less an impairment allowance for expected loss. The foundation recognises roughly two cents a year of cost instead of one hundred cents at placement. That is the entire difference between a request the programme committee can approve and one the investment committee can.
What the clause must say.
- What the reserve absorbs, and in what order.
- The maximum — reserves are never uncapped.
- When the residue is determined, and on what date it is returned.
- Who computes the losses, and by what method.
- What happens if the facility is extended (the most-missed clause of the five).
Worked example. A community foundation posts $1 million. Over five years the book loses $290,000. At maturity $710,000 returns. The foundation's file records $10 million mobilised for $290,000 consumed — a page its trustees will approve again without being asked twice.
Why it matters. Structures are approved by people who must defend them internally. The residue clause is what they defend with.
You already know this because a rental deposit works exactly this way, and nobody thinks of a deposit as money they have given the landlord.
BRIEF 4 — Revolving
The idea. Money lent, repaid, and lent again is a stock with a regeneration rate. Manage r and the corpus grows while it works.
The public case. The Clean Water State Revolving Fund: federal capitalisation of over $42 billion by 2017 and cumulative assistance over $153 billion by 2021 — 3.6 times the federal contribution. Each state matches twenty percent; about a third of what goes out is leveraged bond proceeds and the rest is repayment, recycled.
The arithmetic of one dollar. Lend at four percent on seven-year terms, against annual losses of 0.58 percent and administration of one percent:
cycle 1 $1.00m lent, corpus now $1.16m
cycle 3 $3.51m lent, corpus now $1.56m
cycle 5 $6.89m lent, corpus now $2.11mFive cycles, $6.89 million of lending from $1 million, and the corpus is larger at the end than at the start.
The condition. This only holds while losses plus administration stay below the lending rate. A fund lending at three percent with two percent administration and two percent losses shrinks — quietly, for years, before anyone looks.
Why it matters. A revolving facility does not need re-approval, and a thing that does not need re-approval survives the departure of the person who built it.
You already know this because you understand why a library is not a subscription. The book comes back.
BRIEF 5 — Guarantees, Priced
The idea. A guarantee is a counterparty renting you their balance sheet, and the rent is quoted in basis points.
A real programme. The USDA OneRD Guaranteed Loan Program guarantees up to eighty percent of a rural business loan for a 3 percent upfront fee on the guaranteed portion and a 0.55 percent annual renewal fee on the guaranteed balance outstanding.
$4,000,000 loan, 10 years, 8%, 80% guaranteed
upfront, 3% of $3.2m $96,000
renewal fees over ten years $107,865
------------------------------------------------
total $203,865
average outstanding balance $2,451,474
annualised cost 83 bpsThe trap in reading that number. Eighty-three basis points looks expensive next to a cheaper loan. It is not being compared to a cheaper loan. It is being compared to a decline letter, and against nothing at all any finite number is cheap.
Where guarantees dominate. In 2024 they were 46 percent of all concessional instruments used in blended finance vehicles, and concessional guarantee volume rose 42 percent year on year. They are the cheapest form of credit enhancement because the guarantor posts no cash at all until something goes wrong.
Why it matters. A guarantee is the only instrument here that costs nothing while it is working.
You already know this because you have paid an insurance premium and felt mildly cheated when nothing happened. That feeling is the product functioning.
BRIEF 6 — Patient Capital, and the Size of the Concession
The idea. Patient capital is not cheap money. It is priced money with the price deliberately set below the market, and the gap is measurable to the basis point.
A real instrument. Iroquois Valley Farmland REIT issues Soil Restoration Notes under Regulation A: coupons between zero and 2.50 percent, terms of three, five and seven years, sold to ordinary investors. More than $115 million has been placed into organic farmland this way — about 32,000 acres, 70 farms, 20 states — and half a percent of the raise funds a pool that reimburses tenant farmers for soil amendments and agronomy during transition.
The concession, computed.
note coupon 2.0%
five-year Treasury, stated 4.0%
concession 200 bps
on $10m of notes $200,000 a year
over five years $1,000,000A million dollars of subsidy that appears on nobody's grant budget, in a document filed with the SEC.
What patient investors actually want, in their own order: capital preservation, a defined term, and only then a yield. Ask for subordination and duration; let price be the last item discussed, not the first.
Why it matters. It tells you what to ask for. The ask is not generosity — it is a position in the waterfall and a date.
You already know this because you have lent money to someone you trust and cared far more about getting it back, and when, than about what it earned.
BRIEF 7 — The Blended Cost of Funds
The idea. One line of arithmetic tells you whether a stack holds.
layer weight cost contributes
---------------------------------------------------
first loss 10% 0.00% 0.000%
subordinated 20% 2.00% 0.400%
senior 70% 6.00% 4.200%
---------------------------------------------------
blended cost of funds 4.60%
book: 8.00% yield − 0.58% losses − 2.50% opex = 4.92% net
margin +32 bpsThirty-two basis points, and that thinness is the honest reading. Move the senior tranche from 6.0 to 7.5 percent — one rate cycle — and the same structure returns −73 basis points and stops.
What follows. Always compute it twice: once on quoted rates, once with the senior tranche 150 basis points higher. If it fails the second computation, the first-loss layer is too thin. The fix is at the back of the stack, never at the front — raising the lending rate to close the gap selects precisely the borrowers who produce the losses that break it.
Why it matters. This is the single number a treasurer will ask for, and having it computed both ways is the difference between a meeting and a decision.
You already know this because you have refinanced something, and understood without being told that the average of your borrowings matters more than any one of them.
BRIEF 8 — The Sustainability-Linked Loan, and Its Break-Even
The idea. Your margin falls a few basis points if you hit agreed sustainability targets and rises if you miss. Observed ratchets run 2.5 to 25 basis points; a much-cited academic sample illustrates 175 basis points over Euribor with 5 basis points off on target.
The one division that decides it. Annual compliance cost ÷ ratchet.
annual cost 2.5 bp 5 bp 10 bp 25 bp
------------------------------------------------------
$40,000 $160m $80m $40m $16m
$75,000 $300m $150m $75m $30m
$150,000 $600m $300m $150m $60m(The cost figure is a stated assumption — data collection, limited assurance, the legal work of the reporting schedule. Place your own and read across.)
The sharper term. The ratchet is earned on drawn margin, and revolving credit facilities are frequently undrawn. At $75,000 of cost and 5 basis points: 100 percent drawn breaks even at $150m, 30 percent drawn at $500m, 15 percent drawn at $1 billion.
The market already knows. SLL deal count fell 42 percent in 2025; labelled loan volumes in the first half of 2025 were 52 percent below the same period in 2024; borrowers dropping the label say the benefit did not cover the burden.
Why it matters. Below roughly $150 million of genuinely drawn debt, buy a guarantee, a first-loss layer or a longer tenor instead. All three pay at any size.
You already know this because you have held a loyalty card whose paperwork cost more attention than the discount was worth.
BRIEF 9 — Matching Tenor to Biology
The idea. Finance the asset on the clock the asset actually runs on.
Soil organic matter rebuilds over a decade or two. A coppice rotation is seven to twenty years. A skilled workforce takes five years to form and thirty to mature. A three-year revolving facility financing any of these guarantees a result that looks like failure at the exact moment the asset begins to work.
Long money exists. The CDFI Bond Guarantee Program issues bonds with a 100 percent guarantee from the U.S. Treasury, maturities up to 29.5 years, a minimum bond issue of $100 million and a minimum bond loan of $10 million, purchased by the Federal Financing Bank; roughly $3 billion has been guaranteed since 2013.
What tenor is worth. $10 million at 4.5 percent for 29.5 years against 7 percent five-year bank debt is 250 basis points, or $250,000 a year. But the saving is the smaller half. The larger half is not a price at all — it is that the loan does not come up for renewal in year five, in a bad year, in front of a credit officer who has never seen this asset class.
Why it matters. Most "regenerative projects underperform" findings are findings about the financing term, not about the project.
You already know this because nobody sensible finances a house on a two-year loan and then concludes that houses are a poor investment.
BRIEF 10 — Credit Enhancement at Scale
The idea. At sovereign size the same four layers appear, with different names and eight more zeros.
The case. In May 2023 Ecuador repurchased $1.6 billion of its own bonds for about $656 million — 41 cents on the dollar — refinancing into a blue bond wrapped in $656 million of political risk insurance from the U.S. International Development Finance Corporation and an $85 million guarantee from the Inter-American Development Bank.
debt-service savings $1.1bn over 17 years = $64.7m a year
conservation funding $450m over 18 years = $25.0m a year
endowment, in perpetuity ~$12m a year
savings per $ of IDB guarantee alone 12.9x
savings per $ of total enhancement 1.48xBoth ratios are printed here deliberately. The 12.9× is the number press releases use and it flatters, because it counts only the smaller of the two enhancements. The 1.48× is the honest one. A structure that only looks good on the flattering ratio is a structure to walk away from, and the discipline of computing both is the whole of due diligence in this field.
Why it matters. It shows the pattern is scale-free. The same move — reposition existing money rather than find new money — works at $10 million and at $1.6 billion.
You already know this because you have refinanced a debt at a better rate because someone with a stronger balance sheet was willing to stand behind you, and you understood that their willingness, not your merit, was the thing that moved the price.
All figures in these briefs are computed in lib/verify/I_06.py, printed by python3 lib/verify.py I.06, and sourced in the chapter's Works Cited. Where a figure is an assumption rather than a published number — the annual sustainability-linked-loan compliance cost — it is labelled as one and its sensitivity is printed beside it.