We make four contributions in this paper. First, we provide a core of macroeconomic time series usable for systematic research on China. Second, we document, through various empirical methods, the robust findings about striking patterns of trend and cycle. Third, we build a theoretical model that accounts for these facts. Fourth, the model's mechanism and assumptions are corroborated by institutional details, disaggregated data, and banking time series, all of which are distinctive Chinese characteristics. We argue that preferential credit policy for promoting heavy industries accounts for the unusual cyclical patterns as well as the post-1990s economic transition featured by the persistently rising investment rate, the declining labor income share, and a growing foreign surplus. The departure of our theoretical model from standard ones offers a constructive framework for studying China's modern macroeconomy.