Fasanara: a NAV facility, and who measures the loss

The Lender and the Fund It Lent To

Fasanara lends to businesses for short stretches, buying invoices and other short term receivables at a small discount and collecting them when they come due, which earns it roughly 3.5 percent a year. The handful that go unpaid are meant to be absorbed by the spread on the rest before any of it reaches what investors see. A lender writing a NAV facility against the fund sits one layer further out, holding a claim on the fund with a cushion of fund value underneath that claim.

Before writing the facility a lender needs two answers: how thick the cushion under the loan really is, and what a genuine loss on the loan would look like. The fund trades at no public price and reports figures that cannot be taken at face value, so neither answer can come from it. The cushion gets set from the asset class and the cohort of peers running the same strategy, and the loss gets read from the lender's audit rather than from anything the fund submits.

The pages that follow set the cushion from the cohort, lay the fund's reported losses against the verified ones to show why its figures cannot govern, and close on a single realized loss read from the lender's audit. For the wider standard this study fits into, see document 00, The Settlement Standard, and document 01, The Oracle.

Set the Cushion From the Asset Class and the Cohort

Every fund commits in writing to what it may buy before it takes a dollar, in its offering document, the PPM. That document is enough to place the fund among peers that do the same thing. Fasanara's says it buys short term business receivables through online lending platforms, which lands it in short term business lending, a corner of the wider private lending asset class. Sorting the fund this way takes nothing but the document and no call to the manager.

What the band answers is how far a lender can advance against the fund and stay covered, and the line a cohort sits at inside it becomes what the loss is later measured against. Specialty finance and receivables trades around a 10 to 18 cent band, and the band is market practice, not a number PARI derives. PARI publishes the realized loss, and that read positions this cohort inside the band: the verified loss runs near the calm end of the range, so the line sits toward the high end, near 17 cents on the dollar, rather than coming from anything on Fasanara's own books. A facility struck at or below where the read places the line conforms; one struck above it gets flagged as too thin to carry the strategy's ordinary losses.

Every record also carries a label for how much of the underlying data has been checked, so a reader can see how solid the figure is. With the cushion set, the next question is why the fund's own figures cannot be the ones the loss rests on. For how cohorts are governed, see document 02, How the Numbers Are Made.

Why You Cannot Trust the Fund's Own Numbers

Two documents describe the very same pool of loans and disagree on what it did. The fund's sales sheet reports almost no loss. ARC Ratings, which has worked through this fund's books since 2018, reports a larger one. The distance between them, laid out below, is the whole reason a lender cannot let the fund's own figures govern anything.

Neither side is lying, which is the uncomfortable part. The fund values its own holdings and chooses when to book earnings, and both levers run the same direction here. The sales sheet took a short, calm window and counted only the loans that had gone badly late, while ARC took a full year and counted every loan that slipped at every stage. So the verified rate of loans gone bad lands above 1.20 percent against the 0.16 percent on the sheet, and money actually lost moves from 0.00 percent to 0.40 percent on the same pool. A figure that swings that far depending on who reports it should not be the figure a facility is sized against.

One thing is easy to misread. Even ARC's 0.40 percent is the fund's loss, not the loss on a facility, and it earns its place as context for the cushion rather than as a settlement figure. A facility struck inside the band, near 17 cents here, carries enough cushion that a 0.40 percent fund loss reaches the loan as nothing at all. The loss on the facility is measured on its own, from the lender's side, and the swing on these very loans is the reason it has to be.

The cushion comes from the asset class and the cohort, the warning against trusting the manager comes from the gap above, and the realized loss comes from the lender's own audit. None of it comes from the fund.

What Counts as a Loss, and How Deep the Cohort Runs

It helps to be exact about what counts. A loss on the facility is a realized credit loss and nothing else: the fund genuinely cannot repay, the collateral is enforced and worked out, recovery comes in under the loan balance, and the audited shortfall after that recovery is the loss. A voluntary sale at a discount in the secondary, a soft writedown on paper, and any move that does not breach the cushion are all left out, since each is something a holder can manufacture. The bar is a real credit event, not a decision to sell low.

Held to that bar, the fund's 0.40 percent is again context. Funds in this cohort lose a little in calm years and roughly twice as much in a downturn, going by industry loss history from ICISA and Atradius, and Fasanara's verified 0.40 percent sits near the calm end of that range. A real fund loss, but on a facility struck below the cushion it reaches the loan as no realized credit loss at all.

What does move the cohort is public conditions rather than anything the fund hands out. When more people lose their jobs, more of the fund's borrowers fall behind, and across eight years the fund's verified bad loan rate rose and fell almost in step with the United States unemployment rate. Its reported returns track no public number, because the fund marks its own holdings and its spare earnings smooth over the bumps.

The 0.84 link earns its keep as an early warning, since the fund cannot smooth a public number it does not control, and the jobless rate turns before the borrowers do. It says nothing about what Fasanara will lose next year. The figure a facility settles on stays the realized credit loss read from the lender's audit once the collateral has been worked out.

Read the Loss From the Lender's Audit

Everything before this resolves into one record for the facility, carrying the cushion the loss is measured against, the realized credit loss read from the lender's audit, who confirmed it, and how thoroughly the inputs were vetted. The fund submits nothing to it.

What the lender gains is independence from the borrower's account of itself. It does not rely on the sales sheet's 0.00 percent and takes no number from the fund at all; the loss on its facility comes from its own audit, sits against a cushion drawn from the cohort, and registers only as a realized credit loss, so a 0.40 percent fund loss leaves a well cushioned facility untouched. Size, pricing, and what happens when things slip remain the lender's decisions. Ravariant publishes only the cohort standard, the PARI private credit index, and leaves the loan to whoever writes it.

One facility carried the whole argument. The cushion came from the cohort, the fund's own books showed roughly 0.16 percent bad loans and 0.00 percent lost while ARC found roughly 1.20 percent and 0.40 percent on the same pool, and a gap that wide is why the settlement loss is read from the lender's audit rather than the fund, counting only when a real credit event has run its course. For the wider standard, see document 00, The Settlement Standard, document 01, The Oracle, and document 02, How the Numbers Are Made.