Original, self-directed research — questions I wanted answered that nobody assigned me. Each paper runs the full arc: sourcing the data, building the pipeline, testing the result, and writing up what the numbers actually support, including where the method falls short.
A Large-Scale Analysis of Management Guidance Accuracy in FY2016 Earnings Call Transcripts
Executives issue forward-looking statements every quarter, and almost nobody checks whether those statements come true. This study audits 32,290 guidance items pulled from 3,166 FY2016 earnings calls and grades each one against what the company actually reported. Management delivered on 70.3% of verifiable claims — but the average hides the real story: revenue guidance is dependable (82.5%), EPS guidance is a coin flip (51.8%), vague commentary beats precise targets, and credibility turns out to be a persistent firm-level trait rather than a company-by-company accident.
These are independent research exercises, not published academic work and not investment advice. Every paper documents its own limitations — I'd rather show where a method breaks than pretend it doesn't.