On the quest for the holy grail of diversification

For over a century, academics and market participants have studied two fundamental aspects of finance: asset valuation and portfolio construction. Despite the rich body of literature, studies tend to focus on only one of these concepts and use relative return volatility over a short time period (i.e. days, weeks, or months) as the investment risk metric. In contrast, certain investors utilize fundamental analysis and rely on mean-reversion, which occurs over a much longer timescale (three to five years). Here we argue that the proper risk metric for longer time horizon investments is the variance of the cash flows the asset generates instead of short timescale market price volatility. Accordingly, the values obtained for return versus risk take on a fundamentally different form. When integrated with Modern Portfolio Theory, this leads to a different efficient frontier that does not require as much diversification to obtain optimal portfolios – i.e. only six to eight assets. Although there is substantial empirical evidence for small portfolios in value investors, private-equity firms, and hedge funds, this represents (to the authors’ knowledge) the first quantitative formalism to reach such a conclusion. The authors demonstrate this concept in practice with a discounted cash flow model in conjunction with a Monte-Carlo simulation to determine the probable variance in cash flows. These results suggest that the integration of portfolio construction with fundamental analysis may minimize the risk of large losses, while still creating the opportunity for profits without dampening the effect through over-diversification. These results call into question the over-diversification of fund portfolios and suggest a general strategy for long-term value investing.