Aggregate Market Valuation Index

Premium membership includes access to the CMV Aggregate Market Valuation Index (AMVI) model, our aggregate score of the current valuation of the US stock market. This is presented as a combined score showing approximately how many standard deviations the market is from historic norm, per our models.

The AMVI is pictured below, only shown through 2017 (members get data through the most recent week). This model is a composite of several of the independent valuation/economic models tracked on the site, and provides a one-stop summary of the aggregate US stock market price relative to historical norms, giving you a good idea of how overheated (or oversold) the market is at any given moment.

The Index correlates well with long-term future stock market performance. Below is a scatterplot showing the AMVI value on the x-axis, and subsequent 5-year S&P500 returns on the y-axis.

This scatterplot shows the index's monthly values plotted against the S&P500's subsequent 5-year returns. The left half of the chart shows all instances where the index ranked market valuation as undervalued. In almost every single instance, subsequent 5-year market returns were positive. In general, you can see that the further left/undervalued the AMVI score, the higher the subsequent market performance over the next 5 years.

The opposite is also true. Points on the right half of the chart correspond with AMVI values that were historically overvalued (index score > 0), and while very often subsequent 5-year returns were still positive, returns tend to be lower overall than after periods when the market was undervalued.

This relationship is summarized by the light blue regression line. As the index score goes from undervalued to overvalued (left to right on the chart), the subsequent 5-year S&P500 returns tend to fall substantially.

Premium and Data members are able to view the index score, updated weekly, for an up-to-date understanding of the market's current valuation.

Isolating the Extremes

Looking only during periods of model over or under valuation, the trend is even more clear:

This is the same data but with all of the AMVI "Fairly Valued" datapoints removed, retaining only datapoints where the AMVI was more than 1 standard deviation from its historical trend value. There are many fewer datapoints here, but it's visually much easier to see how all the instances of undervalued market ratings (left side) correspond with relatively high returns after 5 years, and all the instances of overvalued market ratings (right side) have much smaller, and often negative, returns after 5 years.

Predictive Value of the Model

Like everything else on this site, this model is not intended to be used for, or capable of, producing significantly above-average investment returns. This model provides a high-level view of overall market valuation. In live portfolios, market peaks and troughs happen much more quickly than these models can predict or update. Also, future market performance may differ very materially from past trends - past performance does not guarantee future results.