Forecasting Yesterday

Does the stock market affect your mood? Do you feel happy when the stock market is up and upset when it goes down? This is natural, and it affects many of us. A problem occurs when people act on these feelings. This article analyzes the drivers of change in stock prices, with some practical consequences.

Stock prices change every day. Some of the changes are due to company-specific information, and some are related to the whole economy. Changes related to the economy may occur in the following sequence:

  1. Economic data is released, resulting in change of expectations for the future of the economy. The change is assessed relative to the prior expectations.
  2. This new economic data gets reflected quickly in the stock market. For example, an expectation for a slower economy than previously expected will cause the stock market to decline soon after this information becomes public.
  3. When the stock market declines, some individuals and professionals become negative about the future of the stock market, and expect it to keep going down.

This last step is where the logical sequence breaks:

  1. You can expect the future of the economy to affect today’s stock prices.
  2. The future of the economy should not affect tomorrow’s stock prices. Once the data is released, it gets reflected in stock prices quickly (typically within seconds to minutes). There is no economic reason for stock prices to repeat their adjustment to the same data the next day.
  3. You cannot expect today’s stock prices to affect tomorrow’s stock prices, since prices are related to expected economic value, not prior prices.

Furthermore, stock prices do not change according to the expectation for the economy, but according to the changed expectation for the economy. The results may be counterintuitive, for example:

  • If investors expect the economy to boom in upcoming quarters, data indicating a slower expected expansion would be a reason for stocks to decline.
  • The same works in the other direction: if investors expect the economy to slow down in upcoming quarters, new data indicating a more moderated slowdown would be a reason for gains in stocks.

Once you digest the ideas above, there is another factor to consider: Stock valuations ultimately gravitate towards neutral values:

  • The higher the valuation (e.g. Price/Book Value) of stocks, the more the price will sink given a less positive prediction for the economy.
  • Similarly, the lower the valuation of stocks, the more the price will jump given a less negative prediction for the economy.
  • For example, in March 2009 stock valuations hit extremely low values. When bad news came, but just slightly less bad than before, stock prices went up. Given the extremely low valuations, instead of a moderate increase, stocks shot up very high very fast.

One more phenomenon plays into the mix: Given that investors (including some of the biggest institutional ones) are not perfectly rational, they sometimes tend to keep buying stocks that are already expensive or sell stocks that are already cheap. This is what I referred to as “the herd effect” in prior articles, and what is called “momentum” by the investment profession. The results:

  • You can make money by buying expensive stocks (or selling short cheap stocks) for some stretches of time. Given that valuations do not change indefinitely in one direction in the long run, these upward and downward runs get broken by huge upward and downward corrections. As a result, stock valuations cannot be used to predict the near-term changes in stock prices in a systematic way.
  • In general, high valuations cannot be used to predict declines even for long time horizons, since book values of companies tend to grow over time. By the time a bubble gets popped, book values can keep up with prices, resulting in the next bottom being higher than current prices. For example, Extended-Term Component by QAM was more expensive than usual in the end of 2004. Despite that, it had gains all the way to the end of 2008 (around the recent bottom of the worst recession in nearly a century), and +17.3% annually for the entire period ending in the end of 2009, well before a full recovery.
  • Low valuations can be used to predict gains in the long-term. Global stock markets always grow in the long run (at least as evidenced in the past few hundreds of years, and as logically expected). The chances of gains, and specifically abnormally high gains, go up the lower the current valuations are.

One way to benefit from this last point is to buy stocks with low valuations (Price/Book Value), and hold them for the long run.


Not much can help us predict the future of the stock market. The best you can do is hold stocks with low valuations for the long run, to enjoy high average returns.

New economic data affects stock prices today, but neither this data or current stock prices can do a lot to help you predict future stock prices. Understanding this can help individuals as well as professionals avoid ‘forecasting yesterday’.

“Valuation” is a term for how expensive a stock is. Two common measures are Price/Earnings and Price/Book Value.

Disclosures Including Backtested Performance Data

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.