This post is a rejoinder to a remark by a commenter to the effect that “the stock market doesn’t always go up”.
How to read the graph above:
- The x-axis spans nearly four decades.
- The y-axis barely moves.
- Long flat plateaus dominate meaning there are many consecutive positive years.
- Increases occur only in crisis clusters:
- 2000–02 (dot-com / accounting scandal regime)
- 2008 (global financial crisis)
- 2018 (policy tightening scare)
- 2022 (inflation + rates shock)
Over the last 38 years the S&P 500 has only shown year-on-year decreases seven times.
That’s certainly not random behavior and it differs markedly from the behavior over preceding decades.
If the stock market “does not always go up” in any meaningful probabilistic sense, we would expect negative years to occur with some regularity. Instead, since 1987, negative S&P 500 years have been rare and highly clustered. The market goes up almost every year and only declines during systemic crises. In the previous period the stock market posted year-on-year declines a third of the time. The present pattern is incompatible with randomness and inconsistent with a market primarily driven by earnings.
However, it is completely consistent with a stock market that is underwritten by policy, both monetary and fiscal, and structural changes. The vertical annotations in the graph illustrate Federal Reserve interventions.
That has implications including decrease in “real investment”, e.g. building factories, greater income inequality, and increasing dependence of indices on a handful of stocks.
There is a strong correlation—over the long run—between stock market valuation and GDP.
GDP explains the economy. Liquidity explains the market.
The market went down after the dot com bubble. After the housing crisis. COVID. etc Over shorter periods its event driven, not calendar driven. No need to be a sophist, Dave. Over the long haul its primarily GDP driven (in real terms) (as Zach observed) , hence the so-called Buffet ratio.
The effect of the Fed and liquidity is to change the relative risk adjusted returns of equities and debt. Perhaps since Greenspan its been too loose IMHO. But corporations are not stuffing their balance sheets with equities. They have done stock buybacks, but that’s a return of capital, a totally separate issue.
As an aside, I find the more interesting questions right now to be: 1) how accurate are the indices right now, driven as they are by the Magnificants? All the indexes have issues, but for example the Russell 2000 is much more tamed, although more small cap oriented. 2) but probably, by any index, valuations are high. So what to do with rates? Especially having tanked the housing market?
And todays report, although positive, does not a trend make.
If the government is going to keep spending like drunken sailors, and the Fed is going to monetize it, then we better start making more goods and services here.
I completely agree with that, indeed, I’ve been saying it for some time.
The key problem is that “making more” requires capital investment.
The only service subsector that’s really growing is healthcare and that’s a problem. The healthcare sector grows by adding more practitioners or increasing prices. There is a considerable lag to the former and the latter just draws more money out of the rest of the economy without actually producing more.
The absolute best stock market chart I have ever seen put together is by Crestmont Research. The matrix shows stock returns since 1900 on a real basis. What it shows is that there are repeated 20 year periods where the 20 year cumulative average return goes to zero or near zero on an after inflation basis on a regular cycle. We have been in an unusually long period without a 20 year zero return cycle (although 2000-2020 did do it briefly). Here is the real return (dividend reinvested, tax exempt matrix).
https://www.crestmontresearch.com/docs/Stock-Matrix-Tax-Exempt-Real3-11×17.pdf
There are a lot more than just annual returns on the matrix. It shows whether the P/E is increasing or decreasing, real GDP, inflation, events. etc. That data uses annual index averages instead of a single point year end number (dampens the noise in the data).
The data is also shown on a cumulative average annual return (CAAR) instead of a simple annual average return. This is critical. If your stock goes up 100% one year, then declines 50% the next, you made no money. The simple average annual return in this case is 25%. The cumulative average annual return is 0%. The CAAR is the correct number for judging returns.