{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,12]],"date-time":"2026-10-12T03:45:15Z","timestamp":1791776715282,"version":"4.3.4"},"posted":{"date-parts":[[2026]]},"group-title":"SSRN","reference-count":25,"publisher":"Elsevier BV","license":[{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"unspecified","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"abstract":"<jats:p>Industry allocation is an important component of asset allocation and a major source of excess returns in active investing. This study investigates the predictability of sector index returns in China using a multifactor framework. Using monthly data from January 2019 to June 2026 for eleven CSI 800 first-level sector indices, we construct 11 factors including indicator (CPI, PPI, PMI, M2), interest rates (Shibor), technical indicators (momentum and volatility), and a market factor. We model each industry separately and estimate by using two linear models: ordinary least squares (OLS) and Lasso regularized regression, with the period from January 2025 to June 2026 as an out-of-sample test window. A one-shot training scheme (fixed sample of January 2019 to December 2024) yields negative out-ofsample R\u00b2 for both models (-0.72 for OLS and-0.12 for Lasso), which was normal in monthly financial return prediction. To improve the forecasts, we adopt rolling window re-estimation: at each month's end, the models are refitted on the most recent 72, 60, and 48 months of data. Rolling-window estimation improves both numerical accuracy and strategy returns: the 72-month rolling Lasso raises the average out-of-sample R\u00b2 to-0.07 and delivers a Top-3 industry rotation strategy with a cumulative return of 65.45%, an excess return of 48.03 percentage over the equal-weight benchmark; the 72-month rolling OLS earns 37.46% (excess +20.04%). Since both models produce negative R\u00b2 for out-of-sample prediction, we further investigate their ranking power using Spearman rank information coefficient (Rank IC) calculated on a monthly cross-section basis: using the 72-month rolling window approach, the mean Rank IC equals +0.215 (t = 2.96) for Lasso, compared to just +0.057 (t = 0.60) for OLS. Thus, the ranking power of Lasso strategy provides support to the presence of excess returns, while the ranking power of OLS strategy is not statistically different from zero. These results suggest that rolling-window estimation can help models adapt to changing market conditions. Compared with OLS, Lasso shows stronger out-of-sample ranking performance, which is consistent with the excess returns generated by the corresponding industry rotation strategy.<\/jats:p>","DOI":"10.2139\/ssrn.7548878","type":"posted-content","created":{"date-parts":[[2026,10,2]],"date-time":"2026-10-02T13:34:42Z","timestamp":1790948082000},"source":"Crossref","is-referenced-by-count":0,"title":["Predicting CSI 800 Sector Index Returns in China\u2014\u2014&amp;nbsp;A Multifactor Statistical and Regularized Learning Approach"],"prefix":"10.2139","author":[{"ORCID":"https:\/\/orcid.org\/0009-0009-3622-2004","authenticated-orcid":true,"given":"Caiyu","family":"Zhao","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","reference":[{"key":"ref1","journal-title":"Model and window selection on the test window: the window length and headline model were chosen after comparing eight specifications on the same 18month test period; the reported 65.45% is the best across those versions and may overstate performance on new data"},{"key":"ref2","journal-title":"Extend the factor library to include fundamental and trading factors such as industry earnings growth, valuation levels (PE\/PB), turnover"},{"key":"ref3","journal-title":"Extend the sample period to cover multiple bull-bear cycles and examine the robustness of the strategy across different market regimes"},{"key":"ref4","journal-title":"Build transaction costs and holding constraints directly into portfolio optimization"},{"key":"ref5","author":"A Appendix","journal-title":"List of Industry Indices The 11 CSI 800 first-level industry indices examined in this study are"},{"key":"ref6"},{"key":"ref7"},{"key":"ref8","journal-title":"CSI Industrials Index (000930.CSI)"},{"key":"ref9"},{"key":"ref10"},{"key":"ref11"},{"key":"ref12","journal-title":"CSI 800 Financials Index (000974.CSI)"},{"key":"ref13"},{"key":"ref14"},{"key":"ref15"},{"key":"ref16"},{"key":"ref17","author":"B Appendix","journal-title":"Rank IC (Spearman Rank Information Coefficient) Analysis References"},{"issue":"1","key":"ref18","doi-asserted-by":"crossref","first-page":"3","DOI":"10.1016\/0304-405X(93)90023-5","article-title":"Common risk factors in the returns on stocks and bonds","volume":"33","author":"E F Fama","year":"1993","journal-title":"Journal of Financial Economics"},{"issue":"1","key":"ref19","doi-asserted-by":"crossref","first-page":"65","DOI":"10.1111\/j.1540-6261.1993.tb04702.x","article-title":"Returns to buying winners and selling losers: Implications for stock market efficiency","volume":"48","author":"N Jegadeesh","year":"1993","journal-title":"Journal of Finance"},{"issue":"1","key":"ref20","doi-asserted-by":"crossref","first-page":"267","DOI":"10.1111\/j.2517-6161.1996.tb02080.x","article-title":"Regression shrinkage and selection via the lasso","volume":"58","author":"R Tibshirani","year":"1996","journal-title":"Journal of the Royal Statistical Society: Series B"},{"issue":"4","key":"ref21","doi-asserted-by":"crossref","first-page":"1455","DOI":"10.1093\/rfs\/hhm014","article-title":"A comprehensive look at the empirical performance of equity premium prediction","volume":"21","author":"I Welch","year":"2008","journal-title":"Review of Financial Studies"},{"issue":"4","key":"ref22","doi-asserted-by":"crossref","first-page":"1509","DOI":"10.1093\/rfs\/hhm055","article-title":"Predicting excess stock returns out of sample: Can anything beat the historical average?","volume":"21","author":"J Y Campbell","year":"2008","journal-title":"Review of Financial Studies"},{"issue":"3","key":"ref23","first-page":"187","article-title":"A study of cross-industry momentum and reversal effects in the Chinese stock market","volume":"21","author":"S Wang","year":"2012","journal-title":"Operations Research and Management Science"},{"issue":"4","key":"ref24","first-page":"53","article-title":"Momentum and reversal effects in the Chinese A-share market: A re-examination based on a turnover-improved HS model","volume":"40","author":"X Deng","year":"2021","journal-title":"Forecasting"},{"issue":"2","key":"ref25","doi-asserted-by":"crossref","first-page":"194","DOI":"10.1111\/1468-0106.12379","article-title":"Return predictability between industries and the stock market in China","volume":"27","author":"Y Zhang","year":"2022","journal-title":"Pacific Economic Review"}],"container-title":[],"original-title":[],"deposited":{"date-parts":[[2026,10,12]],"date-time":"2026-10-12T03:00:06Z","timestamp":1791774006000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.ssrn.com\/abstract=7548878"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026]]},"references-count":25,"URL":"https:\/\/doi.org\/10.2139\/ssrn.7548878","relation":{},"subject":[],"published":{"date-parts":[[2026]]},"subtype":"preprint"}}