August 14, 2000 The increase in North American stock prices in 1999 and early 2000 has generated interest in the valuation assumptions that would make these price levels sustainable. Here, commonly used valuation techniques are applied to stock markets in Canada and the United States. For the comparative yield approach, real interest rates (rather than nominal rates) are preferred as the comparator of choice to yields on stock market indexes. The spreads between real interest rates and stock market yields have generally increased over the last two years. The dividend-discount model (DDM) approach provides an analytic linkage between the equity-risk premium and the expected growth of dividends. It suggests that market values (measured at the end of February 2000) could be sustained only by rapid growth of dividends in the future or by the continued assumption of an uncharacteristically low risk premium on equity. The spectacular rise in the value of technology stocks in 1999 is noted (Chart 4), and then the valuation measures for the Canadian stock market excluding the technology sector are examined. When this is done with the comparative yield approach, yield spreads are slightly lower, and for the DDM approach, one does not need to assume as high a growth of dividends or as low a risk premium to validate market valuations. Two effects of the "new economy" on the stock market are noted. One is the lowering of dividend yields, as new-economy technology companies tend to have a high reinvestment rate and a low dividend payout rate. Another relates to the potential for a higher track for the economy's productivity growth, which would mean that higher-than-historical assumptions about future earnings growth would be more plausible. Several explanations for the decline in risk premiums on equity are considered. While short-term volatility in the stock market has, if anything, increased in recent years, low inflation and improved economic performance, along with demographics and investor preferences, may have contributed to a decline in the risk premium demanded by investors. A scenario of rapid growth of dividends in the near term slowing to historical norms in the longer term is examined. While this approach can go partway towards explaining high stock market valuations, it requires assumptions that are outside historical experience.