Redefining And Assessing 'Risk' Over Long Horizons: A Practical Approach For Long-Term Investors

“Risk, then, comes in two flavors: ‘shallow’, a loss of real capital that recovers relatively quickly, and ‘deep’, a permanent loss of capital. Capital managed for near-term liabilities should be guided by shallow risk, while capital managed for for long-term liabilities should be guided by deep risk.”

WILLIAM BERNSTEIN ^I

 A Sequel Letter

Our February Letter was titled “Why Long-Term Investors Need To Rethink ‘Risk’”. It argued that while return volatility was a key risk metric for investors with short horizons, long-term investors should focus on what William Bernstein above calls ‘deep’ risks that lead to permanent capital losses. The Letter showed that at a macro level, such ‘deep risk’ losses with stocks are caused by two kinds of events: 1 Buying stocks at unsustainable ‘bubble’ prices, and 2. Buying stocks ahead of extended periods of poor profitability performance by the corporate sector.

The Gordon Model for long-term investment returns (R=Y+G) captures these two deep risks nicely: 1. Buying at ‘bubble’ prices is captured by buying at a too low Y (i.e., current income yield), 2. Poor profitability performance is captured by buying when G prospects (i.e., real growth in investment income) are poor.

The Letter pointed to three implications of these realities for long-term investors:

1. ‘Shallow’ risk captured by return volatility does not offer a logical basis (e.g., through stochastic multi-decade computer simulations) for assessing the ‘deep’ risks facing long-term investors.
2. The plausibility and probability of ‘deep’ risks are best assessed through constructing future multi-decade capital markets scenarios.
3. Long-term investors require feedback loops that provide useful information on Y and G outcomes and prospects over time.

This sequel Letter explores the practical consequences of these three implications.