XBRL Analytics

Failure Case: ROA Calculation

Analysis of Coca-Cola’s Return on Assets (ROA) in FY2023 using XBRL filings.

File Sources

• 10-K filings: https://www.sec.gov/edgar/browse/?CIK=21344&owner=exclude

• XBRL instance file: https://www.sec.gov/Archives/edgar/data/21344/000002134424000009/ko-20231231_htm.xml

Fig. 9 Parsing Failure:

• Incorrect contextRef format used (expected “c-n” pattern)

• Failed to extract values from XML structure

Fig. 10 Context Errors:

• Wrong contextRef “c-31” instead of correct “c-1”

• Noise-induced confusion in financial contexts

Fig. 11 Asset Value Errors:

• FY2023: contextRef “c-23” ($97,703M) vs model output

• FY2022: contextRef “c-26” ($92,763M) vs model output

Validation Testing

1. Closed-book Testing:

   • Model failed to provide ROA without data

   • Correct formula recall: Net Income / Average Assets

2. Ticker Recognition:

   • Successfully mapped “KO” to Coca-Cola

   • Still lacked financial data access

3. Online Search:

   • Retrieved 10.97% from Stock Data Online

   • Formula discrepancy: website vs standard calculation

4. Formula-Guided Search:

   • Correct calculation: 11.25% using Macrotrends data

   • Validated formula: (NI_2023)/((Assets_2022+Assets_2023)/2)

Failure Case: Revenue Forecasting

3-year revenue prediction using 2019-2023 growth rates.

Fig. 12 Calculation Errors:

• Growth rate rounding: -11.41% → -11.4%

• Average growth miscalculation: 5.84% → 5.34%

• Compound errors in 2025-2026 projections

Financial Statements Reference

Fig. 13 Source financial data for all analyses:

1. Income Statement 2023

2. Balance Sheet 2023

3. Income Statement 2022

4. Income Statement 2021