As Figure 4 shows, when our return expectations have been less than 2%, realized returns have often been higher than expected. Although we were apparently overly bearish, our return forecasts were well within the bounds of best and worst realized returns. It is also worth mentioning that market valuation levels have been generally rising, and yields falling, since 1971, so it is possible that our forecasts were correct, net of the (very long) secular trend in valuation levels.
For forecasted returns higher than 2%, the median return for each bucket is in line with expectations, with the gap between the minimum and maximum returns becoming smaller as the expected return gets larger.
It’s important to recognize our expected returns are based on yield, a contrarian signal which echoes our investment belief that the largest and most persistent active investment opportunity is long-horizon mean reversion. Investing using a yield-based signal does not come without its challenges. One big challenge is that a yield signal is a valuation signal that does not come with a timing signal. Because the yield is signaling an asset is attractive today does not mean it will not continue to get more attractive. If the asset’s price falls further, increasing the long-term return outlook, unrealized losses in the portfolio can be uncomfortable. This discomfort is not due to dollars actually lost, but by the sickening feeling that accompanies downside volatility. As American investor and writer Howard Marks has said, “The possibility of permanent loss is the risk I worry about.” We agree. Volatility should not be confused with risk. The permanent loss of capital,8 which happens when investors succumb to fearful thoughts and thus sell at inopportune times, is the investor’s true risk.