A Conceptual Framework for Computing U.S. Non‐manufacturing PMI Indexes
研究提出一个两步法概念框架，通过相关分析和主成分分析计算美国非制造业加权综合指数，发现保留全部六个扩散指数时新订单权重最高、供应商交付权重最低，而仅用两者等权简化指数信息损失很小，且新指数与多个关键经济指标高度相关。
SUMMARY This research develops a conceptual framework for computing new weighted composite indexes for the U.S. non‐manufacturing sector using a two‐step sequential approach — a correlation analysis, followed by a principal components analysis. The results suggest that different weights (i.e., the highest weight to New orders and the lowest weight to Supply deliveries) be assigned if all diffusion indexes in the initial set of six are retained. It also turns out that a simpler index based on two (New orders and Supply deliveries) of the six diffusion indexes, with equal weights, can be computed with little information loss. The new indexes are shown to correlate highly with many key business/economic indicators.