Is the Buffet Indicator a reliable predictor of the stock market’s future?

Secular Analysis

Aaron Soderstrom


April 6, 2023

The Buffet Indicator is named after Warren Buffet. W. Buffet is famous for his value investing approach, which focuses on long-term buying and holding undervalued stocks. This ratio gauges whether the stock market is overvalued or undervalued relative to the overall economy.

The idea behind the Buffet Indicator is that when the stock market is overvalued relative to the overall economy, it is likely to experience a correction. In contrast, an undervalued stock market is expected to experience a rally.

I will not dive into much detail here about the theory behind valuation and the proper way of calculating it. There are a lot of different valuation methods, each with its own merits. Instead, I will focus on the buffet indicator or a simplistic variation of it. This is not a claim that the buffet indicator is the correct way to value the equity market. I am just throwing this out to the world, not trying to offer advice or make recommendations based on this valuation tool.

First, the indicator I am showing is different from the official buffet indicator; what I am showing you is more of the poor man’s version of the indicator. The Buffet Indicator is a ratio of the total market capitalization of all publicly traded companies in each country to that country’s GDP. While it is possible to sum the market cap of all companies in a country, the data will be fragmented with revisions and biases, not to mention a lack of historical data. You can use enterprise value (EV) from some well know publications, but for my purpose, I will cheat. My poor man’s version of the indicator is simply the Wilshire 5000k index representing the market capitalization of the U.S. divided by nominal GDP. Simple right?

Here is my poor man’s version:

The Ability to Predict Market Direction:

The Buffet Indicator has a cult-like following, swearing that the Buffet Indicator has a proven track record of predicting market direction. Some say the Buffet Indicator has accurately predicted market crashes and recessions. Their thesis usually sounds like this:

The Buffet Indicator was at an all-time high of 146.4% in 2000, just before the dot-com bubble burst, leading to a market crash. Similarly, the Buffet Indicator was high at 110.2% in 2007, just before the global financial crisis. In both cases, the Buffet Indicator correctly predicted the market’s direction, and investors who heeded its warning could avoid significant losses.

But I am a quant and a natural skeptic and want to dive deeper into this thesis that the buffet indicator has predictive abilities. Time to break out some data science!

I know, I know, OH @#@!!#! Not statistics.

But don’t worry; we will not make null hypothesis tests or t-tests, and I promise no p values. I will keep this simple, or at least simple, in my mind. I will just be giving some basic regression analysis with pretty charts. I know we all like stunning charts.

Hopefully, my target audience of macro dorks will appreciate my analysis.

Okay, so my approach is to take the indicator’s value and compare it to the future annualized returns. How far in the future? I am still determining, so we will compare it to multiple lots! We will test for R-squared (R²). R-squared is a statistical measure that represents the proportion of variance in the dependent variable (Y) that is explained by the independent variable(s) (X) in a linear regression model. It is also known as the coefficient of determination. Or, for simple-minded individuals like me, the higher the number, the more the model makes sense.

The test results offered hope for the indicator, but I don’t think it is what the buffet junkies expected. As you can see, the poor man’s buffet indicator has no prediction capability in the short term. But things start to get interesting the further out we go.

R² increases as time passes, peaking at 13 years, then loses predictive influence as we extend further. The indicator might be a good tool for a long-term secular approach 13 years later.

If you think about it, this makes sense since the market can remain overvalued for extended periods. But markets don’t tend to stay undervalued for extended periods. If this holds in our analysis, this indicator may be valuable (more on this shortly). But first, let’s dive into that magical 13-year model and explore the data. First, I hope you can see the visual pattern of the regression chart. There is a relationship between the indicator and future returns.

While it is not a perfect fit, and the model has much variance, some information can be taken away from it. Currently, the indicator shows a ratio of 155, and following the linear regression model, we can assume that the next 13 years will average -1.39% approximately. This means there should be turbulence ahead for the market and maybe another lost decade.

But there is a silver lining; looking at the distribution curve of the indicator, I see a tail risk on overvaluation but not undervaluation. I confirm that the market can remain overvalued for some time but not undervalued.

For the market to be the fair value, I would want the indicator to be at or below 85, with anything below 50 deep value. Fire sales of this level will likely need to associate with a recession.

Final Thoughts

So, my conclusion of the poor man’s buffet index is simple, just because the market is overvalued, getting defensive would be a mistake because the market can remain overvalued for a long period. However, an overvalued market like now should be part of a secular outlook. Furthermore, an undervalued market can be viewed in the shorter time horizon because the tail risk of it remaining undervalued remains low. This might be a good tool for deciding when to look for value stocks with healthy financials when “blood is in the street.” The current indicator level tells me that this market crash we experienced last year is not showing any discount yet. The opposite is true that the equity market is still overvalued.

compliance tracking: 5613227.1