Retrieving specific data not covered in the standard statements

[Wessel-Jan]

Hello Dwight,

First of all, thanks for all your work on Edgartools — it genuinely saves me hours of manual data collection and makes this sort of research possible for non-coders like me.
I'm currently trying to replicate some of the long-term identification and valuation analysis described in Phil Oakley’s How to Pick Quality Shares. To do this I need to extract certain raw accounting inputs consistently over at least 10 years. Some of these fields are straightforward, while others are inconsistently tagged across filings. In the long run I’d like to build a screening tool based on these criteria and a dashboard that displays the resulting metrics and ratios for any selected company.
Below I’ve summarized the key metrics I want to calculate and the raw inputs required. My goal is to find a stable way to pull these items across different companies and years.

---

1. Revenue
   Revenue generally works well using the multi-year financial data API (historical trend analysis), though there are some odd cases (e.g., LULU 2018).
   Docs reference:
   https://edgartools.readthedocs.io/en/latest/guides/multi-year-financial-data-api/?h=asset#historical-trend-analysis-endpoints

---

2. Return on Capital Employed (ROCE)
   Operating income is consistently available over time. Normalized operating income (useful for comparing to reported operating income to detect aggressive accounting or heavy exceptional costs) is not widely reported.
   ROCE = Operating income / Capital employed
   Capital employed = Total assets − Current liabilities
   Phil Oakley recommends adding back short-term borrowing, but due to data inconsistency I’ve simplified the formula. This may understate capital employed for firms with high levels of short-term interest-bearing debt.

---

3. Lease-Adjusted Capital Employed
   I want to adjust for off-balance-sheet operating leases.
   Hypothetical purchase price (PP)
   = Operating lease payments × factor (6–8)
   The main challenge: operating lease payments are not consistently available.
   Tags attempted:
   • us-gaap:OperatingLeasePayments
   • us-gaap:PaymentsForOperatingLeases
   Lease-adjusted ROCE =
   (Operating income + PP × interest-rate estimate)
   / (Total assets − Current liabilities + PP)
   To compute this I need reliable access to:
   • Operating lease payments
   • A consistent way to filter annual 10-K data
   I used the “Track Revenue Growth Over Time” examples, combined with .by_period_length(12) and by_form_type('10-K'). Results were still mixed, so I wrote a helper function that attempts both and uses the earliest match. Unfortunately, this is still inconsistent.
   Docs reference:
   https://edgartools.readthedocs.io/en/latest/guides/company-facts/?h=facts+api

---

4. Dupont Analysis
   This breaks ROCE into profit margin and capital turnover.
   ROCE = (Operating Income / Revenue) × (Revenue / Capital Employed)
   All inputs are already mentioned above. I mainly include this to show where the research is going:
   • Avoid firms with highly volatile profit margins (often cyclical or high fixed-cost structures).
   • Capital turnover can rise either by reducing capital employed (less sustainable) or by increasing sales through new products (more sustainable).

---

5. Free Cashflow to Firm (FCFF)
   FCFF = CFO + Dividends from Joint ventures - Taxes paid − CAPEX
   Inputs needed reliably:
   • Cash flow from operations (from cash flow statement; tried: “Net Cash Provided by Operating Activities”)
   • Taxes paid (from cashflow statement; e.g., “Taxes Paid”)
   • Capital expenditures (mixed success across facts API and historical trends)
   • Dividends from Joint ventures Hard to find consistently.
   I also want Free Cashflow to Equity (FCFE) to compare against FCFF, since similarity between them indicates low leverage (a desirable trait in Oakley’s framework).
   FCFE = CFO − CAPEX + Net debt issued
   Additional inputs needed consistently:
   • Depreciation & amortization (for D&A / CFO)
   • Working-capital movements year over year (to spot firms boosting current profitability via inventory reductions, extending receivables, etc.)
   • Inventory-to-turnover ratio

---

My question
Do you have recommendations for the most reliable tags or Edgartools functions to extract the accounting figures mentioned above across different companies and years?
Or more generally:
Is there a “preferred” set of tags or an Edgartools method that normalizes these items when companies use different taxonomy extensions?
Thanks again — any guidance would be greatly appreciated, and I'm happy to test any code or approaches you recommend.

Best Regards,
Wessel