Exhibit 48: S&P 500 Price-to-Trend Earnings vs. Exhibit 49: S&P 500 Valuation Multiples by Subsequent Calendar-Year Price Return Inflation Environment Starting valuation multiples tell us little about equity returns — Periods of low and stable inflation have supported higher over the following year. equity multiples. S&P 500 Returns 1 Year Forward (%) Multiple (x) 50 - R?=5.2% 30 5 m@ Unconditional Average Over Entire Period 40 | rs | m Average During Periods in Which Inflation Is 1-3% and Stable ° " | > & + * * | 225 22.6 20 @ ¢ - ¢ ° 10 re sds * oe a 18.6 135 ; [a | 18.7 16.6 & %% ¢ —z 15 - feo % . | -20 - aa e 10 4 -30 | 2 | -40 | o 5 a | 0 5 10 15 20 25 30 35 40 0 Price-to-Trend Earnings Multiple Shiller CAPE: 1881-2016 Shiller CAPE: 1945-2016 Price-to-Trend: 1945-2016 Data as of December 31, 2016. Data as of December 31, 2016. Source: Investment Strategy Group, Bloomberg, Datastream, Robert Shiller. Source: Investment Strategy Group, Bloomberg, Datastream, Robert Shiller. This bull market is already quite old by historical underweighting equities based exclusively on high standards, second in length only to the almost 10- _—_—- valuations underperformed a strategy of remaining year period that preceded the technology bubble invested across every one of the 20 countries and in 2000. Moreover, valuations now stand in their three country aggregates they examined.'¢ In 10th decile, indicating they have been cheaper short, valuations alone are a poor tactical timing at least 90% of the time historically. In the past, signal. Indeed, the S&P 500 has returned more starting from such a high base has led to muted than 36% since first entering its 9th valuation equity returns over the subsequent five years, with — decile in November 2013, a time when many only a third of those episodes generating a profit were already suggesting that US equities were (see Exhibit 47). in a bubble. Even so, high valuations should not be Valuations must also be considered in confused with