Composite Forensic Scores
Beneish M-Score & Altman Z-Score

Earnings-manipulation screening and bankruptcy-risk assessment computed directly from the general ledger, trailing twelve months ended August 31, 2026, with prior-TTM comparatives.

Prepared 2026-08-24 · Source: NetSuite GL (transactionaccountingline) · All figures USD · Consolidated, Subsidiaries 1–3

1Revenue Trajectory — 24 Months

Monthly posted revenue (account type Income, GL sign-corrected). The prior TTM window is charted in gray, the measurement TTM in navy; the strongest month of the current year is highlighted in red. Growth is steady rather than spiky — there is no quarter-end hockey stick, which is one qualitative point against deliberate revenue stuffing.

2Beneish M-Score — Earnings Manipulation Screen

The Beneish model (Beneish, 1999; refreshed coefficients per Beneish, Lee & Nichols, 2013) combines eight financial-statement indices into a probit-style score. Scores above −1.78 indicate financial-statement characteristics statistically associated with earnings manipulators. The eight-variable form used here:

* AQI and DEPI are set to the neutral value 1.0 because their inputs are degenerate in this ledger (no soft assets in either year; no depreciable PP&E base in the prior year). This is the standard treatment and is conservative in neither direction — see Assumptions A4/A5. Benchmark means: Beneish (1999), Table 2.

What is actually moving the score

TATA decomposition — where the accruals came from

The accruals are almost entirely receivables (+$1.12M) and inventory (+$1.18M). Cash conversion is lagging paper earnings: of the $1.64M TTM net income, roughly $1.79M is tied up in non-cash working-capital growth.

3Altman Z-Score — Bankruptcy Risk

Because this is a private company, the correct variant is Altman's Z′ (private-firm) model, which substitutes book equity for market capitalization. Given the mixed retail/distribution profile, the Z″ (non-manufacturer) variant is also shown — it drops the asset-turnover term that penalizes asset-light service businesses. Both are reported; both agree.

Why the score is so strong: the company carries no long-term debt and no bank credit; total liabilities of $1.49M are entirely trade payables, accrued expenses, and sales-tax collections against $7.03M of assets. Working capital of $5.52M is 79% of the balance sheet. Even the pessimistic ironies of the M-score cut the other way here — the AR/inventory buildup that inflates TATA also inflates X₁.

4Analyst Interpretation — Reconciling the Two Verdicts

The M-score flag is real arithmetic, not an artifact of the model — receivables and inventory genuinely doubled while revenue grew 17%. The forensic question is why, and the ledger offers context that the composite score cannot see:

5Underlying Financial Statements (As Computed from GL)

Condensed balance sheet

Condensed income statement (TTM)

6Methodology, Assumptions & Limitations

Assumptions register

Limitations

Composite scores are screens, not verdicts. The Beneish model was estimated on 1980s–90s US public-company COMPUSTAT data; applying it to a small private ledger — particularly one whose comparative year is its first year of operation — stretches the estimation population, and the 33% "implied probability" should be read as rank ordering, not a literal likelihood. The Altman model likewise predicts formal bankruptcy filings, a rare event for unlevered private companies regardless of health. Neither model detects fraud types that leave ratios undisturbed (e.g., proportional revenue/receivable/cash fabrication). Data reflects posted GL only; unposted or memorized transactions are excluded.

BAppendix — Source Queries

All extraction was performed via SuiteQL against the posted general ledger. Queries follow account conventions (posting flag on transaction, natural GL signs on transactionaccountingline.amount). Reproducible verbatim.

Q1 — Balance-sheet balances at both measurement dates

SELECT
    a.accttype   AS acct_type,
    a.id         AS account_id,
    a.acctnumber AS acct_number,
    a.fullname   AS account_name,
    ROUND(SUM(CASE WHEN t.trandate <= TO_DATE('2025-08-31','YYYY-MM-DD')
                   THEN tal.amount ELSE 0 END), 2) AS bal_t1,
    ROUND(SUM(CASE WHEN t.trandate <= TO_DATE('2026-08-31','YYYY-MM-DD')
                   THEN tal.amount ELSE 0 END), 2) AS bal_t
FROM transactionaccountingline tal
JOIN transaction t ON tal.transaction = t.id
JOIN account a     ON tal.account = a.id
WHERE t.posting = 'T'
  AND a.accttype IN ('Bank','AcctRec','OthCurrAsset','FixedAsset','OthAsset','DeferExpense',
                     'AcctPay','CredCard','OthCurrLiab','LongTermLiab','DeferRevenue','Equity')
GROUP BY a.accttype, a.id, a.acctnumber, a.fullname
HAVING SUM(CASE WHEN t.trandate <= TO_DATE('2026-08-31','YYYY-MM-DD') THEN tal.amount ELSE 0 END) <> 0
    OR SUM(CASE WHEN t.trandate <= TO_DATE('2025-08-31','YYYY-MM-DD') THEN tal.amount ELSE 0 END) <> 0
ORDER BY a.accttype, a.acctnumber

Q2 — P&L flows for both TTM windows

SELECT
    a.accttype   AS acct_type,
    a.id         AS account_id,
    a.acctnumber AS acct_number,
    a.fullname   AS account_name,
    ROUND(SUM(CASE WHEN t.trandate >= TO_DATE('2024-09-01','YYYY-MM-DD')
                    AND t.trandate <= TO_DATE('2025-08-31','YYYY-MM-DD')
                   THEN tal.amount ELSE 0 END), 2) AS flow_t1,
    ROUND(SUM(CASE WHEN t.trandate >= TO_DATE('2025-09-01','YYYY-MM-DD')
                    AND t.trandate <= TO_DATE('2026-08-31','YYYY-MM-DD')
                   THEN tal.amount ELSE 0 END), 2) AS flow_t
FROM transactionaccountingline tal
JOIN transaction t ON tal.transaction = t.id
JOIN account a     ON tal.account = a.id
WHERE t.posting = 'T'
  AND a.accttype IN ('Income','OthIncome','COGS','Expense','OthExpense')
GROUP BY a.accttype, a.id, a.acctnumber, a.fullname
HAVING SUM(CASE WHEN t.trandate >= TO_DATE('2024-09-01','YYYY-MM-DD')
               AND t.trandate <= TO_DATE('2025-08-31','YYYY-MM-DD') THEN tal.amount ELSE 0 END) <> 0
    OR SUM(CASE WHEN t.trandate >= TO_DATE('2025-09-01','YYYY-MM-DD')
               AND t.trandate <= TO_DATE('2026-08-31','YYYY-MM-DD') THEN tal.amount ELSE 0 END) <> 0
ORDER BY a.accttype, a.acctnumber

Q3 — Monthly revenue trend (Fig. 1)

SELECT
    TO_CHAR(t.trandate, 'YYYY-MM') AS yr_month,
    ROUND(SUM(-tal.amount), 2)     AS revenue
FROM transactionaccountingline tal
JOIN transaction t ON tal.transaction = t.id
JOIN account a     ON tal.account = a.id
WHERE t.posting = 'T'
  AND a.accttype = 'Income'
  AND t.trandate >= TO_DATE('2024-09-01','YYYY-MM-DD')
  AND t.trandate <= TO_DATE('2026-08-31','YYYY-MM-DD')
GROUP BY TO_CHAR(t.trandate, 'YYYY-MM')
ORDER BY TO_CHAR(t.trandate, 'YYYY-MM')

Verification queries (run, not shown in full)

• Cumulative net income at both dates (ties retained earnings to $0.00).  • Monthly GL flow 2024–2026 to establish the data-completeness window.  • GL activity by subsidiary confirming subsidiary 4 (xElim) posts nothing.  • Chart-of-accounts type census. All computation performed in an auditable sandboxed calculation with every intermediate value retained; sign conventions verified by the A = L + E and ΔRE = NI tie-outs.