BILL Holdings [BILL] — Trade Construction & Risk

As of 2026-07-29 · spot $46.36 (close 2026-07-28); $47.70 last print 2026-07-29

This document constructs; it does not recommend. The memo issues no position verdict. Sizing, admission and slot competition belong to portfolio-book.

1. What the analysis actually supports

The shape of this is a multiple-reversion trade with a buyback tailwind on a decelerating but cash-generative business, not a growth compounder. It should be constructed as such: modest, defined, and with a hard exit tied to the one number that would invalidate it.

The Quality Criteria FAIL is BINDING for the long-only absolute-return strategy. On that strategy's own rules this name does not get a slot. That is the book's decision, not the memo's — but it should not be discovered late.

2. Liquidity Criteria — PASS

$98m of daily liquidity supports any size this book would take. Exit is not a constraint.

3. Vehicle A — common stock

The default. At 61.5% realised volatility, inverse-vol sizing does the work the framework intends: BILL sizes down automatically relative to a 25-vol name by roughly 2.5x.

4. Vehicle B — Jan-2027 call spread (chain pulled, not assumed)

Per Liquidity Criteria, the actual chain was pulled before proposing anything. Alpaca options snapshots, 2026-07-29, expiry 2027-01-15 (169 days):

The chain is genuinely investable — 13 strikes quoted, open interest 67 to 5,645, bid sizes 150–535 contracts. This is not an HCA-style 18-contract chain.

Structure: buy Jan-2027 $50 call / sell Jan-2027 $65 call.

Honest caveats. 1. The bid/ask spreads are wide — $1.42 on the $50 and $1.90 on the $65. The difference between a mid fill and a market fill is ~$1.66, or 34% of the debit. This must be worked; a market order destroys the trade's edge. 2. There is no volatility edge. Jan-2027 IV of 61–63% sits on top of 252-day realised of 61.5% and above 63-day realised of 56.3%. Options here are neither cheap nor expensive; the structure is a leverage and defined-risk choice, not a vol trade. 3. Expiry timing. 2027-01-15 captures the Q4/FY26 print (late Aug 2026), the Q1 FY27 print (early Nov 2026) and the Q2 FY27 print (early Feb 2027 — marginally after expiry, based on prior years' 6 Feb 2025 / 5 Feb 2026 release dates). Two of the three catalysts land inside; the third does not.

Vehicle preference. Given the Quality Criteria FAIL and the fact that ~15pp of the 37pp target is multiple mean-reversion rather than earnings, the defined-risk structure is the more coherent expression: it caps the loss at a known number on a thesis whose central leg is a re-rating that may simply not happen. But the wide spreads mean the common-stock version is cleaner at small size.

5. Invalidation — the specific numbers

These are the observations that end the thesis, not sentiment triggers.

6. Downside Criteria (MEASURED — logged, blocks nothing)

Base bear: mechanism exhaustion. Probability 30%. −19% to ~$37.50.

Named cause. Take-rate expansion is BILL's only growth engine. Customers grew +1.1% year-over-year and fell 0.9% sequentially in the March-2026 quarter (498,500 → 493,800). Net dollar-based retention is 94% — the installed base shrinks in revenue terms. So transaction-fee growth = TPV growth × take-rate growth, and take rate has already run 28.2bp → 33.3bp in eight quarters against a ceiling set by supplier card acceptance. When it plateaus, transaction growth converges to TPV growth (~12%), then toward customer growth (~1%). Add float declining ~10% a year into a cutting cycle and total revenue growth reaches 7–8% within two years.

At 7–8% growth the exit multiple defensibly compresses toward payments-processor levels, and BILL's own since-2024 floor of 1.78x EV/Sales becomes the reference. On FY2027E revenue of $1,831m: EV $3,259m + net cash $338m = $3,597m ÷ 96.0m = $37.47.

Deep bear: SMB credit cycle in the Divvy card book. Probability 10%. −36% to ~$29.90.

Named cause. Acquired card receivables are $819.4m against an allowance of $18.3m (2.2%). Provision for expected credit losses was $18.8m in Q3 FY26, +26% year-over-year, running 4.6% of revenue. The book is financed by two revolving facilities ($180.0m at 6.60%, $150.0m at 6.02%) that carry liquidity covenants. An SMB default cycle hits interchange revenue (volume), the provision line and covenant headroom at once. Implies ~1.5x EV/Sales: EV $2,747m + $338m = $3,085m ÷ 96.0m = $32.13; allowing for revenue impairment, ~$29.90.

Going concern: NOT argued

Corporate liquid assets $2,173.3m against total debt $1,834.9m at coupons of 0%, 0.32% and 0.48% on the converts, maturing April 2027 ($123.5m) and April 2030 ($1,400.0m). Nine-month operating cash flow $304.8m against $46.7m of capex and capitalised software. Solvency is not the risk here. The risk is that a decelerating business gets re-rated to what a decelerating business is worth.

Ledger entry — bear case −19%, p = 0.30; deep bear −36%, p = 0.10; base +37.2%, p = 0.45; flat-multiple +17.6%, p = 0.15. Cause codes: mechanism_exhaustion, smb_credit_cycle.

7. Momentum Criteria (MEASURED — timing only, never admission)

Absolute 12-month momentum is deeply negative; the short-term reversal is violent and already 49% old. The timing read is do not chase — the base target is 37% away and 15pp of the last 5 sessions has already been taken. If a position is taken, scale it, and accept that the entry is materially worse than it was six weeks ago.