자본시장연구원 Korea Capital Market Institute

We present methods for decomposing stock return autocorrelation into spurious components? the nonsynchronous trading effect (NT) and bid-ask bounce (BAB)?and genuine components?partial price adjustment (PPA) and time-varying risk premia (TVRP). Our methods are applicable to any return period (e.g., daily, weekly, monthly, annual returns, etc.), to both individual stock and portfolioreturn autocorrelations, and to any time horizon (over which the autocorrelations are calculated). The tests are direct in the sense that they are not dependent on any particular market microstructure model of PPA. The tests are constructed using the following four key ideas: theoretically signing or bounding the various components in different situations; computing returns over disjoint subperiods (return periods) separated by a trade to eliminate NT and greatly reduce or eliminate BAB; dividing the data period into disjoint subperiods (time horizons) to obtain independent measures of autocorrelation; and computing the portion of the autocorrelation that can be unambiguously attributed to PPA. We apply our methods to daily individual and portfolio return autocorrelations on the New York Stock Exchange (NYSE) over a data period of ten years, divided into five two-year time-horizon subperiods; our analysis indicates that TVRP is a negligible source of autocorrelation in this setting. We find that PPA is an important source, and in some cases the main source, of the daily return autocorrelation of individualstocks and of portfolios, especially among small- and medium-size firms. We find that a very substantial fraction of the total autocorrelation arises from PPA.