Methodology

A layered approach to financial data.

AtlasEQ standardizes raw filings so companies can be compared on a consistent basis, then makes a focused set of adjustments available where accounting conventions can obscure the economics. These are adjustments, not a rebuild of the accounting. The approach draws on decades of equity-valuation research.

01
The comparability layer

Standardization

Two companies can report the same economic fact under different labels, scales, currencies, and statement structures. Raw filings use thousands of company-specific XBRL tags. Side by side, they are not comparable.

The data starts at the raw filing. From there, each one is mapped onto a single canonical line-item model, resolving the noise that makes raw filings impossible to compare directly.

  • Inconsistent tagging across filers and across time, since tags change year to year.
  • Units and scale, whether reported in thousands, millions, or as-reported figures.
  • One consistent reporting currency per period.
  • Restatements and presentation changes.
Archetype-aware templates

A bank’s balance sheet is not an industrial company’s balance sheet. The canonical model recognizes distinct company types and maps each onto the right structure.

Operating / industrial

Working capital, inventories, PP&E, and the operating versus non-operating split.

Banks & financials

Deposit funding, federal funds purchased and repos, trading and investment securities, provision for credit losses, FHLB advances.

IFRS filers (ESEF / 20-F)

Reconciled into the same canonical model as US-GAAP filers, so cross-jurisdiction comparison holds.

Special structures

Multi-class share structures, REITs, MLPs, and similar cases are handled explicitly.

02
The adjustment layer

Adjustments

Standardized statements are still written under accounting rules that can obscure economic reality. R&D is expensed instead of treated as investment. Leases sit outside the operating picture. One-time items muddy what real earnings look like. On top of standardization, a focused set of well-established adjustments is available, each one grounded in financial-statement-analysis research and applied only when you choose to.

Examples of the adjustments available include:

  • Operating versus financing reformulation.

    Operating performance can be separated from how the business is financed, so you see what the company actually earns from operations.

    Framework: Penman & Nissim, Ratio Analysis and Equity Valuation (2001); Penman, Financial Statement Analysis and Security Valuation.

  • R&D capitalization.

    R&D is an investment in future value, but accounting expenses it immediately. Qualifying R&D can instead be carried as the asset it economically is.

    Framework: Lev & Sougiannis (1996); Damodaran on R&D and value.

  • Operating-lease interest reclassification.

    For lease-heavy businesses, the interest cost buried inside rent overstates operating expense. That implicit interest can be moved out of operating results and into financing costs, so operating profit reflects the business rather than its lease financing. It is a reclassification: it relocates the cost without changing pretax or net income.

    Framework: Damodaran, Dealing with Pension and Lease Liabilities, applying the same economic logic IFRS 16 already uses.

  • Core (sustainable) earnings.

    One-time and non-recurring items can be stripped out, so you see the earnings power that is likely to repeat, not last year’s lawsuit or asset sale.

    Framework: Penman’s core earnings; the earnings-quality literature.

  • Clean-surplus repair.

    Gains and losses that bypass the income statement still change shareholder value. The accounting identity can be restored, so nothing leaks out of book value.

    Framework: Ohlson (1995) clean-surplus relation; the Feltham-Ohlson residual-income framework.

03
Why it holds up

Built to be inspected

Ledgered

Every adjustment is recorded.

A filed number is never silently overwritten. Each adjustment is a traceable delta against the standardized base, so the original figure and the adjusted figure are both inspectable.

Auditable

Reversible by design.

Because adjustments are deltas in a ledger, they can be inspected, attributed, and reversed. Nothing is a black box.

Sourced

Regulatory inputs only.

Primary inputs are official filings. No scraped estimates stand in for a source figure.

References

Research lineage

The methodology is grounded in the frameworks below. Citation reflects influence on the approach, not endorsement of AtlasEQ or peer review of its specific implementation.

  • Penman, S. & Nissim, D. (2001). Ratio Analysis and Equity Valuation. Review of Accounting Studies.
  • Penman, S. Financial Statement Analysis and Security Valuation.
  • Lev, B. & Sougiannis, T. (1996). The Capitalization, Amortization, and Value-Relevance of R&D. Journal of Accounting and Economics.
  • Ohlson, J. (1995). Earnings, Book Values, and Dividends in Equity Valuation. Contemporary Accounting Research.
  • Feltham, G. & Ohlson, J. (1995). Valuation and Clean Surplus Accounting.
  • Damodaran, A. Writings on R&D, leases, and financial-statement adjustments.

This page is a public extract of the methodology and does not disclose the full pipeline.