The S&P 500 Index of US stocks is nearing a new record close to eclipse the peak of 1,565.15 set in October 2007. Simple interest-rate based valuations, such as the so-called Fed model, suggest the benchmark could more than double again - but on long-term assumptions, it would then crash-land.
The Fed model equates the earnings yield on stocks with the prevailing 10-year Treasury bond yield. This yardstick proved a reasonable predictor for shares between 1982 and 2007, with stock markets rising as interest rates fell. The relationship has broken down since then.
The model named for the Federal Reserve is flawed, accounting neither for the inflation protection built into stocks nor, on the downside, for the extra risk of stock over US government bonds. Nevertheless, ultra-low interest rates tend to push stock prices higher. Were the S&P 500 to align with the Fed model so the earnings yield equaled today's two per cent return on 10-year Treasuries, it would top 4,000. Matching the previous high in 2000 for the index adjusted for the nominal size of the US economy would bring a new peak just shy of 2,500.
Even that would represent a significant bubble. In fact, the current level of the S&P 500 is a stretch. Based on current as-reported earnings and a 1950 to 2012 average price-to-earnings ratio just under 19 times, according to data compiled by economist Robert Shiller, a figure of around 1,600 is on the money. Today's US corporate earnings, however, are flattered by low interest rates. They totaled 12.5 per cent of national income in the most recent government statistics. Since 1950, the average is 9.5 per cent. Overlay that adjustment, and the right level for the S&P 500 would be nearer 1,220. Alternatively, apply the Fed model using a long-term average Treasury yield. Take current S&P 500 earnings and apply a 50-year average yield of 6.6 per cent, and the result would put the index at about 1,290. Assume earnings will decline as a proportion of national income, as above, and that drops to about 980. Dividends reveal a similar picture. The average dividend yield on the index between 1950 and 2012 was 3.3 per cent. The current yield is 2 per cent. Assume a reversion to the long-term mean, and the S&P would drop by nearly 40 per cent.
Or look back to February 1995 when then Fed Chairman Alan Greenspan embarked on policies that expanded the money supply much faster than the pace of economic growth. The S&P 500 stood at about 487. Apply the growth in nominal GDP since, and it implies an index level of 1,060. Once interest rates and other factors normalize, the index could slump by a third from current levels - and far more should investors first inflate it further.
The Fed model equates the earnings yield on stocks with the prevailing 10-year Treasury bond yield. This yardstick proved a reasonable predictor for shares between 1982 and 2007, with stock markets rising as interest rates fell. The relationship has broken down since then.
The model named for the Federal Reserve is flawed, accounting neither for the inflation protection built into stocks nor, on the downside, for the extra risk of stock over US government bonds. Nevertheless, ultra-low interest rates tend to push stock prices higher. Were the S&P 500 to align with the Fed model so the earnings yield equaled today's two per cent return on 10-year Treasuries, it would top 4,000. Matching the previous high in 2000 for the index adjusted for the nominal size of the US economy would bring a new peak just shy of 2,500.
More From This Section
Or look back to February 1995 when then Fed Chairman Alan Greenspan embarked on policies that expanded the money supply much faster than the pace of economic growth. The S&P 500 stood at about 487. Apply the growth in nominal GDP since, and it implies an index level of 1,060. Once interest rates and other factors normalize, the index could slump by a third from current levels - and far more should investors first inflate it further.