COVID-19疫情下的金融經濟現象與資產配置策略

Abstract

觀察COVID-19疫情引發的金融經濟特殊現象，本研究想知道後疫情經濟的資產配置與投資策略為何，運用Python機器學習以K Means Clustering將雙因子在過去週期循環特徵分群 (債券利差10-2YR、產出缺口因子)，嘗試標示後疫情的景氣循環可能位置，來進行可能的配置，最終從製造周期、庫存與相關指標方向確認臺灣的股市小周期位置，本研究限制在於，需要人為再進一步判讀，主要優點為清楚辨識週期相對位置，讓主動型資金配置對金融資產有相對位階概念。 此次是少見的受迫性與事件型經濟衰退，因疫情導致快速度全世界同步大撒幣與大擴張，也將出現少見可能的週期快速復甦大成長與衰退或大幅度趨緩增長停滯的週期，目前已經進入景氣循環末週期的股債表現向下修正同向週期，關鍵結構因子為通膨變化，需要進一步觀察其水位變化，決定後續的公債配置以及股市內涵配置，大抵而言，股債的資產配置週期因景氣位置、貨幣政策循環不已。

Observing the special financial (inverted yield curve before rate hike cycle) and economic phenomena caused by the COVID-19 pandemic, this study wants to know what the asset allocation and investment strategies of the post- pandemic economy is. Using Python machine learning to K Means Clustering, trying to mark the possible stage of the post-pandemic boom cycle, through the two-factor (bond spread 10-2YR, output gap factor) labeling large cycle, to carry out possible allocation, and finally from the manufacturing cycle, Inventory and related indicators to confirm the small cycle stage of Taiwan's stock market. The limitation of this study is that it needs to be further interpreted artificially, the main advantage is to clearly identify the stage of the cycle. This is a rare forced and event-type recession, and may will also be possible of hotter and shorter cycle. The current has entered the end of the boom cycle. The performance of the stock and bond may downward correction of the cycle.