Abstract
This study examines whether accounting fundamentals and valuation multiples predict quarterly stock returns within a hand-picked sample of ten US mega-cap firms observed over 83 quarters (2005Q2–2025Q4). The sample—Apple, Microsoft, NVIDIA, Amazon, Alphabet, Berkshire Hathaway, Walmart, Eli Lilly, Exxon Mobil, and Johnson & Johnson—collectively represents roughly one-third of US equity market capitalization during the sample period and spans four distinct economic regimes: the pre-GFC expansion, the 2008–09 financial crisis, the 2010–2019 low-rate expansion, and the 2020–2025 pandemic/inflation era. Because traditional cross-sectional factor tests are ill-suited to such a narrow cross-section, I adopt a panel case-study design using firm and quarter fixed effects with firm-clustered standard errors. The main specification relates one-quarter-ahead returns to lagged log price-to-earnings, log price-to-book, return on equity, leverage, liquidity, and firm size. I find three robust results. First, within-firm over-time variation in the earnings multiple negatively predicts next-quarter returns (β = −0.023, t = −2.88), consistent with a value-premium mechanism operating even within the mega-cap universe. Second, lagged return on equity negatively predicts returns (β = −0.0007, t = −2.31), indicating a profitability reversal that is inconsistent with a naïve quality-factor interpretation.
References
This work is licensed under a Creative Commons Attribution 4.0 International License.