Note from dshort: I've updated this commentary to include last week's release of the Advance Estimate for Q3 GDP.
Market Cap to GDP is a long-term valuation indicator that has become popular in recent years, thanks to Warren Buffett. Back in 2001 he remarked in a Fortune Magazine interview that "it is probably the best single measure of where valuations stand at any given moment."
The four valuation indicators I track in my monthly valuation overview offer a long-term perspective of well over a century. The raw data for the "Buffett indicator" only goes back as far as the middle of the 20th century. Quarterly GDP dates from 1947, and the Fed's B.102 Balance sheet has quarterly updates beginning in Q4 1951. With an acknowledgement of this abbreviated timeframe, let's take a look at the plain vanilla quarterly ratio with no effort to interpolate monthly data.
The strange numerator in the chart title, MVEONWMVBSNNCB, is the FRED designation for Line 39 in the B.102 balance sheet (Market Value of Equities Outstanding), available on the Federal Reserve website. Here is a link to a FRED version of the chart. Incidentally, the numerator is the same series used for a simple calculation of the Q Ratio valuation indicator.
I've updated the GDP denominator with the Q3 Advance Estimate. The numerator is from the Fed's Q2 Z.1 Financial Accounts released on September 18th with a linear extrapolation for a Q3 estimate.
Here is a more transparent alternate snapshot over a shorter timeframe using the Wilshire 5000 Full Cap Price Index divided by GDP. I've used the FRED data for the stock index numerator (WILL5000PRFC).
A quick technical note: To match the quarterly intervals of GDP, for the Wilshire data I've used the quarterly average of daily closes rather than quarterly closes (slightly smoothing the volatility).
How Well do the Two Views Match?
The first of the two charts above appears to show a significantly greater overvaluation. Here are the two versions side-by-side. The one on the left shows the latest valuation over two standard deviations (SD) above the mean. The other one is noticeably lower. Why does one look so much more expensive than the other?
One uses Fed data back to the middle of the last century for the numerator, the other uses the Wilshire 5000, the data for which only goes back to 1971. The Wilshire is the more familiar numerator, but the Fed data gives us a longer timeframe. And those early decades, when the ratio was substantially lower, have definitely impacted the mean and SDs.
To illustrate my point, here is an overlay of the two versions over the same timeframe. The one with the Fed numerator has a tad more upside volatility, but they're singing pretty much in harmony.
Incidentally, the Fed's estimate for Nonfinancial Corporate Business; Corporate Equities; Liability is the broader of the two numerators. The Wilshire 5000 currently consists of fewer than 4000 companies.
What Do These Charts Tell Us?
In a CNBC interview earlier this spring CNBC interview (April 23rd), Warren Buffett expressed his view that stocks aren't "too frothy". However, both the "Buffett Index" and the Wilshire 5000 variant suggest that today's market is indeed at lofty valuations, now well above the housing-bubble peak in 2007. In fact, we can see in the first chart above only four quarters (during the dot.com bubble) with higher valuations. The latest estimate is about half-way between two and three standard deviations above the mean valuation.
Wouldn't GNP Give a More Accurate Picture?
Note from dshort: GNP estimates lag GDP by a month, so this section still reflects data through the second quarter.
That is a question I've been asked multiple times. Here is the same calculation with Gross National Product as the denominator; the two versions differ very little from their Gross Domestic Product counterparts.
Here is an overlay of the two GNP versions -- again, very similar.
Another question I've been repeatedly asked is why I don't include the "Buffett Indicator" in the overlay of the four valuation indicators I update monthly. I've not included it for various reasons: The timeframe is so much shorter, the overlapping timeframe tells the same story, and the four-version overlay is about as visually "busy" as I'm comfortable graphing.
One final comment: While I see this indicator as a general gauge of market valuation, it it's not useful for short-term market timing, as this overlay with the S&P 500 makes clear.