Why the market has outperformed for 10 years

December 15, 2025

---

"There always were two ways to live in a world that is often dark and full of tears. We can curse the darkness or we can light a light… a little light drives out much darkness." — Rabbi Jonathan Sacks

---

Today is an important edition of Daily Interest. We’ll look at the market performance over the past 10 years.  And maybe what can it tell us about the next 10 years. It gets a little technical, but if you’re a numbers nerd, you might enjoy it.

Over the past ten years, the S&P 500 returned +14.5% per year.  That’s pretty good!  In the fifteen years prior, it was not nearly as good. 

2016-2025: +14.5% per year

2006-2015: +9% per year

2000-2015: +6% per year

Over the last 50 or 100 years, the annual return (assuming dividends are reinvested) is typically 10-11%ish.

So, this begs the question: why have returns been so good lately?  

Let’s pull it apart.

The Mega Caps

We’ll start with the largest components.  The largest 15 members of the S&P 500 account for 46% of the index.  And those 15 stocks returned +27.7% per year over the past 10 yrs.  But the 485 remaining averaged only 10.4% per year.  So, we have our first clue:   27.7% for the top 15, versus 10.4% for the bottom 485.

Apparently, the massive outperformance of the mega-caps pulled the whole index higher.  Without them, returns would have been meaningfully lower.

But now we have a new question.  Returns of 28% per year over a decade is an amazing streak.  How did they do that?  

Why were the mega caps so strong?

Stock returns are driven by three variables:

(1) earnings growth,

(2) valuation, and

(3) dividend yield.  

Let’s do “valuation” first.  

In the list below, some multiples expanded dramatically (e.g., Apple and Walmart) while some shrank a lot (e.g., Amazon and Meta).  But if we take the group as a whole, they were roughly in the same range.

Tesla’s multiple is massive (as we’ve talked about in the past), so let’s exclude that for a second.  The average Price/Earnings multiple (excluding Tesla) went from 23.4x in Dec. 2015 to 27.8x today.   That’s a 19% increase over 10 years – or about 1.7% increase per year.

So, think about it like this: only about 1.7% of the 27% annual return was due to the P/E multiple expanding.  With a few exceptions, the basket does not scream “dot com” bubble to me.

Next let’s do dividend yield.  The current average dividend yield for these 15 stocks is 0.6%.  So that didn’t move the needle very much.  (To be fair, these are growth stocks.  Growth stocks typically have lower dividend yields, because they reinvest most of their earnings back into the business or share buybacks.)

Lastly, let’s take a look at the elephant in the room:  earnings growth.

Over the past five years, these mega 15 stocks had average annual earnings growth of about 27% per year.   The other 485 stocks in the index?  The average of those was 10%.  US corporate profits overall typically grow 6-7%.  

Get the picture?  Earnings growth made all the difference. Why has the market outperformed over the last 10 years? Multiple expansion helped, but the primary driver was high earnings growth.

Below I’ve copied the annual earnings growth over 3, 5, and 10 year periods.

Nvidia was the clear champion – with annual EPS growth of 88% over the past three years!   Nvidia’s dominance in AI semiconductors gave them a near monopoly on the AI chip market.  It allowed Nvidia to birth a whole new industry nearly all on its own.  But even if you exclude Nvidia for a second, the remaining 14 stocks still averaged 22% per year of EPS growth over the past five years.  This was a broad-based earnings growth wave.

What’s the punchline?

So what does this tell us:

(1) Multiples are not cheap.  But is it a bubble?

Multiples are not cheap.  We know that for sure.  The cap-weighted average of the entire S&P 500 is currently 33.7x. The median is 20x.  Right after the Great Financial Crisis in 2009, that figure was closer to 12x.   Now, that was cheap.  

But today’s valuations are only a little higher than 10 years ago.  Going from 23x to 28x over 10 years?  Yes, that’s higher, and multiples are currently above most measures of long-term averages.  But is that a “bubble”?  I’m going to be bold and say that it’s not.  There are lots of inputs to calculate an earnings multiple.  (Most of it is technical and beyond the scope of this little newsletter.) For our purposes, let’s think of multiples like a water balloon.   My view is that the multiples are “full” but not “stretched”.

(2) Earnings growth drives the bus.  

The point of this story is to keep your eye on earnings growth and earnings estimates.  When growth rates start to slow down, and when earnings estimates start to fall, that’s when the party will end.

For now at least, the party is still rocking.  The top 15 companies have expected earnings growth of 20% over the next two years.  (Based on current analyst estimates for these companies.)  Only three companies (Meta, Tesla, and Oracle) have seen estimates fall recently.  

As someone who used to publish earnings estimates on Wall Street, I can guarantee you that these will almost surely be wrong to some degree.  But they’re our best estimates at the moment.  If they pan out, we could likely see a few more years of above 10% annual market returns or more.

Sustainable, above market EPS growth is the holy grail of stock selection.  It’s why the Mag 7 and their friends have had such a terrific run the past few years.  

One eye on the multiple, and one eye on the growth.