Phase Space AI

Financial Model Notes

SharonAI Holdings [SHAZ]

SharonAI Holdings Inc. [SHAZ] — Financial Model Notes

As of: 2026-07-29 · Framework: v1.5.1 / criteria.md 2026-07-29 · All figures traced to primary EDGAR filings.

Per the brief, the screen's inputs were not trusted. Share count, TTM revenue and net cash were independently verified against primary filings, and net income ÷ shares ≈ filed EPS was cross-checked for scale. Every discrepancy is reported; none is silently adopted.

---

Verified inputs — the financial statements exist

Recorded first, because the brief's premise was otherwise: SHAZ filed a 10-K for FY2025 on 2026-03-31 (0001493152-26-014068) and a 10-Q for Q1 2026 on 2026-05-15 (0001493152-26-023769). Both are XBRL-tagged and both were used here.

EPS cross-check (net income ÷ shares ≈ filed EPS)

($19,915,764) ÷ 13,959,245 = ($1.427). Filed EarningsPerShareBasic for Q1 2026 renders as −1 at integer precision in the raw XBRL dump. FY2025: ($39,623,909) ÷ 9,804,075 = ($4.041), and the filed figure renders as −4. Both reconcile in sign and magnitude. No scale error detected — and the FY2025 tie at −$4.04 against a filed −4 is a clean two-way confirmation that the share count is right.

Net cash reconstruction — and the screen's error

Screen error: $204.3m, and a sign error. SHAZ was reported as holding $172.8m of net cash when it holds $31.5m of net debt — and the omitted item was $199.4m of convertible notes classified as CURRENT, i.e. the most urgent liability on the balance sheet. Mechanical cause identical to HIVE, HUT, IREN and WYFI: the resolver looks for LongTermDebtNoncurrent, found nothing, and still set net_cash_complete: true. The corroborating signal it already held and ignored: LiabilitiesCurrent of $221,469,455 against a computed debt of zero.

A second, distinct screen defect specific to SHAZ

The record assigns exit_multiple: 4.5, exit_multiple_basis: "GROWTH_MATCHED", exit_multiple_peer_n: 9. The growth input was the 257.4% CAGR, which measures the discontinued Filecoin storage business. The correct input is the latest-quarter YoY figure of −9.6%, for which no comparator set exists. GROWTH_MATCHED is asserted over an anchor matched to an artifact — the same class of unfounded completeness flag as net_cash_complete: true. The scan computes and flags CAGR/latest-quarter divergence on other names (IREN, WYFI, CIFR, RIOT, WULF, APLD) but does not feed that divergence into the anchoring routine.

Modelling notes and hazards

1. The revenue base is discontinued. FY2024's $438,292 and part of FY2025's $1,566,631 are Filecoin decentralised-storage revenue, and "the Filecoin operations ceased during the second quarter of 2025." Do not build any growth rate across Q2 2025. The screen's 257.4% CAGR does exactly this.
2. The convertible notes are CURRENT, not long-term. $199,358,226 in ConvertibleNotesPayableCurrent. Any net debt calculation that looks only at non-current debt tags returns zero — which is what happened.
3. Post-balance-sheet financings are material and are NOT in the 2026-03-31 accounts: the $350m of 6.00% notes due 2031 (April 2026, conversion ~$48.24) and the $50.0m New Era note collection (April 2026). Pro forma liquid assets ≈ $573m against debt ≈ $554m.
4. A disclosure inconsistency in the source document. The Q1 2026 10-Q states the $350m offering "closed on April 30, 2026" in subsequent events and "has not yet closed" in MD&A. Unresolved; flagged.
5. Share history is not comparable across periods. Weighted-average basic went 1,067,213 → 13,959,245 in twelve months through a SPAC combination (Roth CH Acquisition) and a February 2026 IPO. Do not build a per-share series across it.
6. No digital assets, no bitcoin, never mined. The legacy business was Filecoin storage provision, not proof-of-work. No ASU 2023-08 marks to strip, nothing pledged — genuinely clean on this dimension.
7. Gross margin must be taken from the MD&A, not from a tagged GrossProfit fact. Q1 2026 cost of revenue of $526,000 against revenue of $294,014 is disclosed narratively. The screen's gross_margin_pct: 6.4 is the FY2025 figure; the current quarter is −78.9%.
8. Do not build a forward DCF. There is no contracted book, no named counterparty, no disclosed contract value and a negative gross margin. Any revenue ramp would be invented, and inventing one is the failure this memo is explicitly avoiding. The screen's own note — "reverse DCF unsolvable in range" — is correct and should be respected rather than worked around.

---

Reproduction

Commit: e67ab6405055a342b54a5ef03613a3650d59ab05