Haute Lumière
Commerce · I.06 · MMXXVI · daylight
Three instruments: a ten-point quiz, eight reflection questions, five essay prompts. The quiz checks comprehension rather than recall. The reflections are private and first-person. The essays are arguable from more than one side.
Four on recall.
1. Name the four layers of a blended stack, in the order they absorb losses, with the return each expects.
First loss (absorbs first, expects nothing), subordinated or patient capital (absorbs next, wants one to three percent and duration), senior commercial debt (absorbs last, wants market rate), and the optional guarantee wrap across the senior tranche, priced in basis points. One mark for the order, one for recognising that the guarantee is not a layer of cash — the guarantor posts nothing until something goes wrong.
2. What is an attachment point, and what is it for a senior lender sitting above ten percent first loss and twenty percent subordination?
The share of the structure that must be wiped out before a lender takes its first loss. Here, thirty percent. The stronger answer notes that this is the sentence to say in the first meeting: a credit officer who hears "attachment point" hears someone who has done this before.
3. State the residue clause and say what it changes on the provider's balance sheet.
Whatever the reserve does not absorb returns to its provider at maturity. It converts the placement from an expense at full face into an asset carried at face less an impairment allowance for expected loss — roughly two cents a year of recognised cost instead of one hundred cents at placement.
4. What does the USDA OneRD guarantee cover, and at what published fees?
Up to eighty percent of the loan, for a three percent upfront fee on the guaranteed portion and a 0.55 percent annual renewal fee on the guaranteed balance outstanding.
Four on application.
5. A foundation programme officer says: "We can't do this — we'd be giving away a million dollars and we only have four million to grant this year." Reframe it in one paragraph, and name the condition under which they are right.
The million is posted, not spent. At a 0.58 percent net charge-off and a three-year average life, the expected cumulative loss on a $10 million book is $174,000 — about 1.74 cents per dollar of capital mobilised, against 100 cents for the same dollar granted, with roughly 83 percent of the reserve returning at maturity. They are right if the foundation cannot survive a total loss of the tranche, or needs the money liquid inside the term: the provider owns the variance, not the mean, and the money is locked. Full marks require naming both conditions, not just the first.
6. Your treasurer computes a blended cost of funds of 4.60 percent against a net book return of 4.92 percent and calls it approved. What do you insist on before signing?
Recompute with the senior tranche 150 basis points higher. At 7.5 percent the blended cost becomes 5.65 percent and the margin becomes −73 basis points. The stronger answer states the corollary: if it fails the second computation the fix is a thicker first-loss layer, not a higher lending rate — raising the rate selects the borrowers who produce the losses that break the structure.
7. A mid-sized firm is offered a sustainability-linked revolving credit facility with a 5 basis point ratchet. What three things do you need to know before it is worth signing, and in what order?
The facility size, the expected drawn percentage, and the annual cost of compliance — data collection, assurance and the legal work of the reporting schedule. Break-even is cost divided by the product of ratchet and utilisation. Credit any answer that names utilisation; it is the term most models omit and it is frequently decisive.
8. Why does the chapter insist you approach the senior lender last?
Because the senior lender is underwriting the structure rather than the project, and without the layers beneath it there is no structure to underwrite — the answer is a version of "come back when you have the rest." Building from the back also means every later conversation is conducted from a signed term sheet rather than an intention.
Two that require the arithmetic to be done.
9. A firm is offered a $250 million sustainability-linked revolving facility with a 5 basis point ratchet. It expects to draw 40 percent on average. Compliance will cost about $90,000 a year. Does it pay? Show your working, and state the break-even facility size.
Benefit = 250,000,000 × 0.0005 × 0.40 = $50,000 a year. Against $90,000 of cost the label loses $40,000 a year. Break-even = 90,000 / (0.0005 × 0.40) = $450 million of facility at that utilisation. Credit any method reaching a loss of roughly forty thousand. The point of the question is that the answer is a subtraction, not a judgement about sustainability.
10. A proposed stack is 5 percent first loss at zero, 15 percent subordinated at 3.0 percent, and 80 percent senior at 7.0 percent. The book yields 7.5 percent, loses 1.0 percent and costs 3.0 percent to run. Does it clear? If not, what is the fix?
Blended cost = (0.05 × 0) + (0.15 × 0.03) + (0.80 × 0.07) = 0 + 0.45% + 5.60% = 6.05%. Net book return = 7.5 − 1.0 − 3.0 = 3.50%. Margin = −255 basis points; it does not clear, and not marginally. The fix is at the back: thicken the first-loss layer so the senior tranche prices lower and weighs less, or reduce operating cost. The wrong fix — and the one most often proposed — is to lend at nine percent, which closes the arithmetic gap by importing the credit losses that will reopen it.
These are not for a room. Write the answers by hand if you can; the slowness is the point.
Each is arguable from more than one side. Each requires at least one source the chapter cites and at least one it does not.
1. Is first-loss capital subsidy or investment? The chapter argues a first-loss dollar does roughly fifty-seven times the work of a granted dollar because most of it returns. Argue either that this is the single most under-used instrument in philanthropy, or that expected-value framing systematically understates the risk a foundation actually takes on — variance, illiquidity, and correlation between losses and the conditions that created the need. Use Convergence's State of Blended Finance, and one source on programme-related investment practice or foundation risk appetite that the chapter does not cite.
2. Did the sustainability-linked loan fail, or did it work exactly as designed? Deal counts fell 42 percent in 2025 and labelled loan volumes halved. Argue either that this is a market correcting an instrument whose pricing benefit never covered its reporting cost, or that the instrument succeeded at something else entirely — forcing KPI disclosure into thousands of credit agreements that would otherwise contain none. Engage Aleszczyk, Loumioti and Serafeim directly, and at least one published critique or defence the chapter does not cite.
3. Who should bear first loss — and who actually does? Blended finance mobilises roughly four private dollars per concessional dollar. Argue whether concessional capital deployed this way is a public good efficiently levered, or a transfer of downside from private investors to public and philanthropic balance sheets under a technical vocabulary. Use the Convergence data and the Ecuador transaction, and one critical source on blended finance or development finance additionality that the chapter does not cite.
4. The ninety-times case. Fifty million dollars of Ford Foundation credit enhancement stood behind $4.5 billion of mortgages for 50,000 families. Write the case that this is the most efficient philanthropic intervention in American housing — then write the strongest rebuttal, which must engage seriously with what the same capital could have done elsewhere and with what happened to comparable borrowers in the foreclosure crisis. The UNC Center for Community Capital research series is the starting point; find at least one account not produced by a programme participant.
5. Tenor, and what we call failure. The chapter claims that many findings of regenerative underperformance are findings about financing term rather than about the practice. Argue the counter-case: that long tenor entrenches unproductive assets, removes the discipline of renewal, and defers the moment at which a failing project is stopped. Use the chapter's CDFI Bond Guarantee Program material or the Clean Water State Revolving Fund, and one source on capital discipline, soft budget constraints or evergreen funding that the chapter does not cite.