Follow LF0 researchXLinkedInSubstackFacebook

Research Methodology

LF0 publishes independent equity research, SEC filing comparisons and quantitative strategy notes. This page explains how to read that work and its public scorecard.

Equity research

Reports combine company filings with financial and market data. They examine the business, growth, valuation, cash conversion, balance-sheet risk and peers. The source and period shown beside a table matter: a historical report is not a live valuation. Data-provider measures can differ from company-defined non-GAAP measures and should not be treated as interchangeable.

How ratings are tracked

The public performance dashboard links ratings to their original reports. Stock returns use adjusted daily closing prices from the entry date shown on each call. Buy signal returns follow the stock return; Sell signal returns reverse its sign. Strong Buy and Strong Sell signal returns use a 1.5 conviction weighting. Hold is tracked separately and excluded from directional hit-rate calculations. Calls are tracked for 182 days and then frozen.

Hit rate counts directional calls that have been correct so far. Open calls remain provisional. Average signal return is a research statistic, not a live portfolio return; it does not by itself include trading costs, short borrow fees or a common investment start date. Read the assumptions attached to any model portfolio comparison separately.

SEC filing comparisons and backtests

Filing-change notes compare new annual-report language with the prior filing in areas such as Business, Risk Factors and Management Discussion and Analysis. A wording change is a research lead, not proof of financial impact. Backtests should be read with their reported sample period, trading rules, transaction-cost assumptions and limitations. Historical simulated results do not establish future returns.

Dates, sources and corrections

An article’s publication date identifies the original thesis. A refreshed performance card does not update its financial analysis. Corrections should identify the affected claim, the correction date and its source; the original rating record remains available for accountability. See the research desk and post-publication reviews for more context.

LF0 publishes research and educational commentary. Nothing here is personalized investment advice or an offer to buy or sell securities. Past performance does not guarantee future results.

Automation and AI-assisted work

Parts of LF0’s data collection, filing comparison and drafting workflow use automation and AI-assisted tools. These tools can introduce errors. Read the stated data periods, inspect the sources, and check any dated corrections when assessing a report. An automated performance update does not represent a fresh editorial review of the underlying thesis.