Ten years after the technology bubble, some unsubstantiated beliefs remain. The so-called Fed model, which is the idea that high stock prices are reasonable when nominal interest rates are low, is still very common. My own research and others’ have shown this proposition to be a form of money illusion with no power to predict (even noisily) long-term stock returns. But the Fed model still yields a far more bullish forecast than focusing just on equity prices (unadjusted for nominal interest rates), as it has for a long time. Its bullishness probably accounts for its continued popularity, particularly among strategists on Wall Street.
The Shiller P/E (the current price of the S&P 500 Index divided by the previous 10-year average real earnings) has become the lingua franca of those that discuss the ERP and how it relates to current equity prices. This choice is not because the Shiller P/E is perfect—no measure is—but simply because it is reasonable and historically consistent. It also helps to have a common standard.
Recently, the Shiller P/E has been back in the news because some broker research has called it into question. The attacks are mostly ridiculous; they are based on bullish researchers using Wall Street’s long-term preferred “operating” earnings, which are earnings before negative events are deducted, or throwing out historical periods that the researchers do not want in the data. If the price of the S&P 500 is compared only with other times when the price was high, then of course it will look lower.