Phase Space AI

Valuation

BILL Holdings [BILL]

BILL Holdings [BILL] — Valuation

As of 2026-07-29 · spot $46.36 (close 2026-07-28) · framework v1.7.0

Two outputs, per references/valuation.md. Neither replaces the other.

---

1. Verified inputs

Funds held for customers ($3,999.136m) and customer fund deposits ($3,999.136m) are equal and offsetting and are excluded from both sides, as the screen also did correctly.

---

2. The implied-path test — the Valuation Criteria

reverse_dcf.py, solving for revenue CAGR.

Parameters held fixed, named explicitly: - Terminal EBIT margin 11.2% (the screen's industry median (pre-profit) anchor) - Exit multiple 24.1x EBIT (the screen's GROWTH_MATCHED, peer n = 254) - WACC 10.0%, horizon 5 years - Revenue base $1,600.3m, net cash +$338.5m, shares 99.596727m

Result

The market requires a 5-year total-revenue CAGR of 9.8%.

The margin — demonstrated − required

The choice of "demonstrated" is the whole argument, so all four candidates are shown:

* Not directly comparable: the reverse DCF runs on total revenue, which includes a float line that is shrinking 10% a year. Using core growth against a total-revenue EV would double-count. The correct blended forward rate is core growth × 90.6% weight + float growth × 9.4% weight ≈ 15.4% × 0.906 − 10.0% × 0.094 ≈ +13.0%, which is where the 12.5–13.5% observed range comes from.

The margin is +2.7pp, not +16.3pp. The screen's figure was inflated by both errors simultaneously: a stale, float-padded demonstrated rate and an EV overstated by $1.18bn.

Implied multiple compression — stated as a number

Today's EV/EBIT is undefined (TTM GAAP EBIT is −$61.4m). The comparison has to be run on sales:

• Exit multiple implied EV/Sales = 24.1 × 0.112 = 2.70x
• Today's EV/Sales = 2.67x
• Implied compression: +0.03x, i.e. essentially zero.

This is the cleanest statement of what the price asks for: hold the sales multiple flat for five years, deliver an 11.2% GAAP EBIT margin, grow 9.8% a year, and the buyer earns 10%. No re-rating is required and none is assumed.

Verdict: PASS, thin, and conditional

Required 9.8% sits below demonstrated 12.5% and below the company's own FY2026 guidance. The price does not require acceleration. That is a PASS on the stated test — but see §3, because the PASS lives or dies on one input I did not verify.

---

3. Sensitivity over the exit multiple — the highest-variance parameter

Per criteria.md, sensitivity is run on the exit multiple, never on scenario probabilities.

The break-even exit multiple is 21.3x EBIT. Above it, BILL passes. Below it, BILL fails. Everything else in this valuation is noise by comparison.

The exit multiple is partially UNIDENTIFIED — declared, not defaulted

valuation.md requires an exit multiple drawn from a comparator set whose growth brackets the subject's growth at the exit year, or a declaration that it is unidentified.

The 24.1x came from the screen's GROWTH_MATCHED procedure over a 254-name peer set that I did not rebuild inside the time box. Two reasons to distrust it here, both structural rather than speculative:

1. Classification. BILL is SIC 7372, "prepackaged software." A growth-matched anchor drawn from that universe is a software multiple. 82.1% of BILL's revenue is payments and float (transaction fees 72.7%, float 9.4%). The comparator set is almost certainly the wrong industry, and the error runs in the expensive direction — payments processors trade at materially lower EBIT multiples than software.
2. Growth at the exit year. At a 9.8–12.5% CAGR, BILL's FY2031 growth rate is plausibly high-single-digit. A growth-matched software set at that growth rate is a different, and lower-multiple, set than the one that produced 24.1x.

I did not pull Corpay, WEX, Marqeta or Toast multiples, so no replacement number is asserted. The multiple is declared partially unidentified, the break-even (21.3x) is stated, and the Valuation Criteria PASS is recorded as conditional on a ≥21.3x terminal EBIT multiple being defensible for a low-teens-growth payments business. That condition is the one thing a reviewer should attack first.

---

4. Sensitivity over the terminal margin

Inverted — solving for the margin at a given growth rate:

Is 11.2% GAAP EBIT reachable? TTM GAAP operating margin is −3.8%. FY2026 guided non-GAAP operating margin is 18.6%. The gap is stock-based compensation at ~$236m annualised (14.7% of revenue) plus D&A. Holding SBC dollars flat at $236m while revenue compounds 9.8% for five years to $2,556m puts SBC at 9.2% of revenue — which gets GAAP EBIT to roughly the high single digits, not 11.2%, unless non-GAAP margin also expands. So 11.2% requires both flat SBC dollars and continued non-GAAP margin expansion. It is achievable and it is not free. The 9.9% needed at the demonstrated 12.5% growth rate is the more comfortable version of the same requirement.

---

5. The 12-month target

Built per valuation.md: near-term estimates × the name's own multiple history, with the percentile stated. Not a DCF. Not a peer median.

Step 1 — FY2027E revenue (fiscal year ending 30 June 2027)

FY2026 guidance (reaffirmed 2026-05-26): total revenue $1,642–1,652m (midpoint $1,647m), core revenue $1,496.3–1,506.3m (midpoint $1,501.3m). Guide-implied FY2026 float: $145.7m.

Base rationale: core decelerates from the FY26 guided 15–16% by ~2pp on a customer base that shrank sequentially in Q3 FY26 and NDR of 94%, partly offset by continued take-rate expansion. Float declines with Fed policy — BILL's realised yield on ~$4.0bn of customer funds is ~3.7% TTM; each 100bp of cuts is worth roughly $40m of revenue, 2.5% of the total.

Step 2 — the multiple anchor: BILL's own EV/Sales history

Daily EV/Sales, using TTM revenue as known at each date (quarterly XBRL lagged to filing date, so nothing is forward-looking) and the current verified share count and net cash — so the series measures multiple movement.

(percentiles computed on the 2026-07-29 print of $47.70; at the $46.36 close used for the target the current multiple is 2.67x and the percentiles are marginally lower)

Regime declaration. The full post-IPO history spans an unambiguous regime change — BILL traded above 100x EV/Sales in 2021 and is 86.1% below its 2021-11-09 high of $342.26. A full-history median of 15.5x is not an anchor, it is an artifact. The full-history anchor is declared UNIDENTIFIED.

The since-2025-01-01 window is used. It is 393 sessions (~19 months) and it spans the entire current regime: low-teens total growth, mid-teens core growth, declining float, and GAAP operating margin approaching zero. It is the only window in which BILL has been the business it is now.

Step 3 — the target

FY2027E total revenue $1,831m; net cash +$338.5m; 96.0m shares (99.6m today, less partial execution of the $1.0bn authorisation announced 2026-05-07 — BILL repurchased $272.7m in the first nine months of FY2026, ~1.0m shares in Q3 FY26 alone).

12-month target: $63.60, +37.2% above spot.

Of the +37.2%, roughly +18pp is revenue growth carried at today's multiple, +15pp is multiple mean-reversion to BILL's own 19-month median, and +4pp is share-count reduction from the buyback. The multiple leg is the contestable one and it is stated separately for that reason.

Named events that move it inside 12 months

Each appears in BILL_Catalyst_Calendar.md with what is and is not known about its date.

1. Q4/FY2026 results and the first FY2027 guide — the single largest. It will be the first guide that sets a float assumption into a cutting cycle and the first that guides core growth off a customer base that shrank sequentially.
2. $1.0bn repurchase execution — 21.7% of the current market capitalisation, 24-month authorisation, funded from existing cash.
3. Fed policy — 9.4% of revenue is float; ~$40m per 100bp.
4. The proxy/activist process — a campaign is evidenced (DFAN14A 2025-09-08; "activism" in the last three earnings releases). Its terms are not.

Sanity band against the external target

Press-reported Street mean target: ~$53.91 (secondary source, unverified — used only as a check, never as a calibration input, per valuation.md).

Our base sits +18% above Street. The gap is entirely the multiple: at today's 2.67x held flat our target is $54.51, i.e. Street's target implies essentially no re-rating — it is our "flat multiple" case almost exactly. So the divergence is not a disagreement about the business; it is a disagreement about whether the 30th percentile of BILL's own recent range is where it should still be trading in a year. We say partial mean-reversion; Street says none. That difference is explicit and testable.

Framework note (item B16). This is a target above spot, as valuation.md says should be common and as the 16-of-16-below-spot record says was previously not.

---

6. Reconciliation to the screen

The screen reached the right letter by two offsetting errors of similar magnitude. Had only the net-cash error been present, BILL would have looked cheap for the wrong reason; had only the stale-CAGR error been present, it would have looked like a 31% grower at 3.4x. Both together produced a +16.3pp margin that is really +2.7pp — a six-fold overstatement of the ranking variable the strategy sorts on.