Is There An Earnings Bubble?

“Matt, you don’t understand. It’s the earnings that are in a bubble. Once you live through one of these cycles, you’ll get it. You’re just too young to understand.”

That’s the PG version of a DM I got last week. If I copy-pasted, this blog would get rated R fast. I’m trying to keep it family-friendly.

I get it. This is a very contentious topic. I am not here to underplay it. All I want to do is weigh in and give my perspective.

But first, at the risk of speaking out of line, I want to tell you exactly who I am and who I am not.

1/ I DO NOT claim to be a “fundamental, bottoms-up, three statement financial model Wall Street analyst.” I have no foresight into how many dollars per share the S&P 500 companies are going to print over the next 12 months.

2/ I DO claim to approach markets and provide answers to big questions through the lens of history. That requires a spreadsheet and data. Never thought I’d be saying it, but I’m proficient with those tools.

I am not going to rely on my hunch, a viral Substack article, or a LinkedIn DM from a stranger as my primary sources to see if we’re in an earnings bubble.

Instead, I’m going to do my best to look at the data, make some baller charts, and share them with you all.

So that’s what I did.

Let’s begin.

Below you’re looking at two lines:

My goal here is to show you how reality (the dark blue line) has aligned with what analysts expected reality to look like (the light blue line) for S&P 500 earnings.

And here’s my takeaway: most of the time, the lines look pretty f’ing (staying PG here) similar.

The only time that the lines really diverge notably is during and after recessionary “shocks.” I’m articulating this point in the chart below which illustrates the % spread between how far actual earnings came in above or below what analysts had expected.

Theoretically, if analysts had a crystal ball, 0% would be plotted across the entire chart above. Unfortunately, we don’t live in that reality.

During recessionary periods, earnings are nowhere near estimates from a year prior. That’s because shocks are hard to forecast.

But here’s the important part: outside of recessionary periods, the spread between actual and estimated earnings shrinks to just 1.5% (median across all periods). That means outside of recessions, the median estimate has been essentially “in-line” with what actually happened.

On February 28th, 2012, analysts expected $106 of EPS for the S&P 500 over the next 12 months. On February 28th, 2013, the S&P 500 had actually printed $106 of EPS over the past 12 months. The analysts nailed it.

On July 31st, 2016, analysts expected $124 of EPS for the S&P 500 over the next 12 months. On July 31st, 2017, the S&P 500 had actually printed $124 of EPS over the past 12 months. Nailed it again.

I’m cherry-picking, yes, but there’s more examples of this level of precision. And if you don’t believe me, take a look at the chart below.

The most frequent spread between actual and estimated earnings is between -1% and 0% meaning the most likely case is that analysts are “spot on” in their forecasts. 67% of the time, actual earnings are within plus or minus 5% of estimates set by analysts from a year prior.

Here’s my point: the analysts actually have a strong historical track of forecasting earnings outside of recessions.

If you think we are heading into a recession over the next 12 months, then saying the earnings are overstated is perfectly fair.

But that’s different than saying all of the earnings are fake and the analysts don’t know what they are talking about for the next 12 months.

The historical data simply doesn’t support that claim.

That’s all for today. Thank you, as always, for reading. I’ll be back charting for you next week!