Reliable ETF Buy and Sell Limits

Abstract

A split-half research design is used to test the reliability of ETF trading limits calibrated to maximize returns. Use of prices relative to optimized timing of simple moving averages to detect turns in stock market trends is found reliable for only fifty-four percent of older ETFs. Exponential moving averages are less reliable. Calibration of weeks to compute moving averages of weekly initial unemployment claims in the United States, and trading limits based on them, produces superior returns that are reliably higher than buying and holding among most of a globally diverse sample of older exchange traded funds.