Long-Term Investing and the Frequency of Investment Decisions

Abstract

This article analyzes the impact of the frequency of investment decisions on long-term investment results. Long-term investment returns are shown to be a nonlinear function of skill, transaction costs, volatility, and frequency. The author uses compound returns to show the minimum skill level that is required to result in a positive expected return after costs and shows the skill level required to outperform the buy-and-hold investor. These can be high for markets that have high transaction costs per unit of volatility. The optimum frequency of investment decisions is derived as that frequency that maximizes the expected terminal value of wealth. Empirical estimates from the US equity and fixed-income markets show that this optimum frequency is considerably different between these markets, even when the investor has an equal amount of skill in each. The author also shows how long-horizon value-at-risk varies with the frequency of investment decision-making.

TOPICS: Wealth management, equity portfolio management, fixed income portfolio management, risk management, portfolio construction, performance measurement

Key Findings

• ▪ The author derives the success rate that is required to result in a positive expected return after cost and applies it to directional strategies in the US equity and fixed-income markets. For fixed income, the investor requires a success rate of more than 51%, regardless of the frequency of those decisions. For directional strategies in the US equity market, this minimum success rate varies more with frequency, but it is higher than 53%.

• ▪ The optimum frequency of investment decisions is derived as that frequency that maximizes expected terminal wealth. The article shows that this frequency is equal to skill times volatility over costs, squared. Because neither of these parameters is likely to be constant through time, the optimum frequency of investment decision-making is also likely to change through time, which has important implications for how an investment process is to be structured. Empirical estimates show very different results for different markets.

• ▪ Even for highly skilled investment processes, downside risk is meaningful when frequency is low. Downside risks are meaningfully lower only for high-skill investment processes at high frequency. This is an important consideration for investors who place high importance on downside risks at long investment horizons.