How debt drove market multiples

Investment Analysis Clubs Macro Economic Trends and Risks
1

To see how massive amounts of debt have propelled stock market values far above real GDP growth, I charted data taken from the Federal Reserve.

I chose 1980 as the index year because that was the beginning of a multi-decade drop in interest rates (rise in bond values) and was also the beginning of the great bull market in stocks.

image
image1846Ă—580 60.4 KB

I chose to plot this on a linear chart (instead of a logarithmic scale) to show the exponential growth in stock prices relative to GDP growth. I showed nominal GDP as well as real (inflation-adjusted) GDP to show that even the nominal GDP which shows the impact of inflation is far below the growth rate of debt and far below the growth rate of the corporate equity market. (FRED can’t show more than 10 years of SP500 due to a restrictive deal with S&P so I used a all-market series from the Fed.)

The massive amount of fiat money conjured out of thin air by the Fed goosed the asset markets starting in 2008.

Fed Chair Warsh has said that the Fed should stop meddling in the markets. Among other things, this probably means Quantitative Tightening (QT) or at least no more QE. (Other than the fed funds rate which is pumped into banks every day on their reserves held at the fed, currently 3.5%-3.75%.)

The 10 year Treasury yield, which is most important for business, has reversed the trend since 1980.

The era of cheap money is over. The stock market is inflated by over $1.5 Trillion in margin (almost 5% of GDP).

Can the stock market multiples continue to inflate exponentially without the vast ocean of fiat money – so much faster than the growth of real GDP?

image
image890Ă—884 55.9 KB

Wendy

9 Likes
2

In the dot-com and in the housing bubble burst the stock market appears to have dropped back to the solid blue line. If that happens today… wow…

5 Likes
3

It is worth noting that there are some issues with Shiller PE Ratio:

In particular:

Issue 1: Static Stock Composition Myth

The CAPE ratio assumes a constant mix of stocks over time, which fails spectacularly in today’s dynamic U.S. market. High-growth tech companies such as Apple, Microsoft, Nvidia, Google, Meta, and Amazon now dominate the index, with dramatically increased earnings and market weights over the past decade.

…

Take NVIDIA as an example: The price component accounts for its current ~7% weight in the index that reflects its 6350% earnings growth that has taken place over the past decade. Yet the earnings component includes its tiny 0.06% weight and much smaller earnings from a decade ago.

This mismatch creates a distorted picture of what an investor is actually buying. It would only be logical if one expected these companies’ earnings to plummet by over 90%—an extremely unlikely scenario for established market leaders.

Another issue that it does not account for share buybacks which increase the EPS without a change to earnings. More details on others at the link.

9 Likes
4

Exceptionally unlikely. A better floor to fear would be the P/E ratio during the first and second quarter of Covid.

In other words, we would likely have to have economic activity worse than April 2020 - when few were working and before government handouts, for the P/E ratio to drop even further. A bubble popping in an industry or two is not likely to cause that. More likely another, even more severe, black swan event*.

*'Cause we are unlikely to shelter in our homes and shut down businesses simply due to another pandemic that kills 1 million+ citizens.

3 Likes
5

If a million is insufficient, how many people would you need to kill before the vaccine and mask ignorance crowd got worried? 20 million? 50 million?

I bet AI’s biological weapon capabilities test this limit sooner than later.

https://www.rand.org/pubs/perspectives/PEA3853-1.html

intercst

4 Likes
6

The SPX has never been static. (Nor has the DJIA.) It’s constantly changing and the S&P committee may even change their own rules to admit Space X.

And today’s high P/E stocks have great reasons to be high P/E, just like Saul’s SaaS stocks before them…that is, before AI smacked them down…and of course we aren’t taking “distilled” Small Language Model competition to the “frontier” Large Language Models into account because they haven’t shown up yet to compete…yet…

As for the rest…this time it’s different. Of course. :wink:

Wendy

1 Like
7

That’s how markets work. If you’re not prepared to ride out a 50% plus stock market drop, you probably shouldn’t be an investor.

intercst

4 Likes
8

You clearly don’t know me. Nothing further need stated.

9

I agree that CAPE needs some sort of adjustment. With start dates from 1998 to 2016, the average 10-year real subsequent CAGR of the S&P 500 was 6.1%. CAPE’s average prediction (based on a 1988 regression) was -1.7%, and so the average CAPE prediction error (1998 to 2016) was 7.8%. This is a very large error.

Many different adjustments have been suggested. Applying a simple adjustment to CAPE reduces the average error (1998 to 2016) to about -2%, and so I use that as a baseline for what is possible. Looking at the CAPE data, there is a structural break around 1991. A CAPE adjustment should mention why it is more forceful after 1991.

This analysis suffers from a shortage of data. Using 5-year subsequent returns instead of 10 years would double the number of independent out-of-sample data points.

2 Likes
10

Jus to provide a counter-argument… the best time of the 4 year presidential cycle is from 4Q of mid-term to 2Q of the election year.

11

AI capex is now a significant part of the GDP. We are talking about over $1 T investment by hyperscalers alone and you can count the multiplier effect. If this bubble pops, it will not be just the tech, financial, but will impact even mundane sectors like utility. The finances of many city, states will be devastated.

It is not just Fed, but even states and City have bloated budgets that are not sustainable. There is a reason no politician today talks about deficit. If US government balances its budget we are looking at depression not recession.

4 Likes
12

Interestingly the earnings so far this year are so strong, the multiple have actually come down. Now analysts are predicting an $8 raise to the 27 estimates…

1 Like
13

A bubble popping in stock valuation does not necessarily mean that that main street is impacted - or even impacted at a point worse than it was during Covid.

14

There is a reason stock market gains are part of the inflation metrics. Wealth effect is real and it drives folks to spend.

We can create strawman like COVID or GFC etc, but the fact is it impacts main street.

15

What is your argument? You appear to have objected to my claim that if things pop, the Shiller P/E ratio won’t be worse than it was during Covid (see the chart). Even if it impacts mainstreet, my hypthesis is that it would need to be worse than Covid for the Shiller P/E ratio to fall further than it did during Covid.

Just what are you objecting to and how is my comparison to Covid a strawman?

16

At this time not sure. :confounded_face:

1 Like
17

A simple 20% bear market primarily hits sentiment, discretionary spending, and corporate hiring plans. However, unless the crash spills into the banking system or triggers widespread credit freezing, Main Street often absorbs a valuation drop far better than headlines suggest.*

Today’s AI spending boom and SPX concentration in AI related companies is similar to 2001 but different in others.

A 2001-style AI stock crash would not necessarily cause a deep 2008-style housing or banking crisis. Instead, Main Street would experience a selective white-collar recession: sharp tech layoffs, a slowdown in high-end consumer spending, a hit to passive retirement accounts, and localized slowdowns in data center construction and energy markets. However, the broader economy would likely be cushioned by the underlying financial strength of today’s mega-cap corporations.

An AI-driven market crash would likely cause a localized credit crunch centered in private markets and non-bank lenders rather than a repeat of the 2008 banking crisis. However, because private credit and banking lines have become intertwined, a slowdown in spending would ripple through non-bank financial institutions, making corporate credit harder and more expensive to obtain across the broader economy. [end Gemini quote]

Due to the Dodd-Frank law, the Fed will not rescue individual shadow banking lenders or private credit funds, but yes, it would act to backstop systemic liquidity across the broader financial system if a credit crisis threatened to freeze the real economy.

Wendy

  • Details from Gemini

1. The Main Street Impact of a 2001-Style Tech Bust

Historically, the 2001 recession was unusual: it was one of the mildest macro recessions in U.S. history for GDP, but brutal for white-collar employment and equity investors.

If an AI valuation bubble burst today, the ripple effects on Main Street would likely follow a similar pattern:

A. Concentrated White-Collar & Tech Layoffs

  • 2001 Playbook: When telecom and internet companies stopped buying networking gear, tech giants cut hundreds of thousands of jobs, spilling into consulting, marketing, and investment banking.

  • Main Street Today: A sharp pullback in AI spending would trigger immediate hiring freezes and layoffs across software engineering, data science, specialized AI startups, and enterprise consulting firms.

B. The CapEx Freeze & Supply Chain Contagion

  • 2001 Playbook: Companies like Cisco faced massive inventory write-downs when telecommunications firms suddenly canceled equipment orders.

  • Main Street Today: Big Tech companies (hyperscalers) spend hundreds of billions annually on data centers, chips, power generation, and specialized cooling systems. An AI spending freeze wouldn’t just hit chipmakers—it would hit Main Street suppliers: electrical grid contractors, real estate developers, data center construction workers, and local power utilities.

C. Widespread 401(k) / Retirement Portfolio Hits

  • Market Concentration: Today, the top 10 stocks in the S&P 500 account for nearly 40% of the index’s total market capitalization—even higher than at the peak of the dot-com bubble.

  • Main Street Impact: Because passive index investing (S&P 500 target-date funds, 401(k)s) is far more ubiquitous on Main Street today than in 2001, an AI equity crash hits the balance sheets of everyday broad-market index investors directly, even if they don’t hold individual tech stocks.

2. Parallels vs. Key Differences: 2001 vs. Today

While the hype cycles feel nearly identical, the underlying mechanics of the companies driving them are fundamentally different.

Attribute Dot-Com Boom (1999–2001) Today’s AI Expansion (2020s)
Market Concentration High (~25–30% in top 10 stocks) Unprecedented (~40% in top 10 stocks)
Balance Sheet Health Unprofitable startups funded by speculative IPOs & debt Mega-cap giants funded by massive existing free cash flows
Revenue Source “Eyeballs” and speculative promises Existing cash cows (Cloud, Search, Digital Ads, Enterprise Software)
Valuations (P/E) Extreme (S&P tech sector P/E > 60x; Cisco peak P/E > 130x) Elevated, but largely backed by actual earnings growth
Main Street Physical Reach Telecom fiber optics laid underground Massive power grid demands, data center real estate, energy infrastructure
4 Likes