The Meb Faber Show Sep 18, 2026 43m 26m saved
With Roger Ibbotson, Chairman of Zebra Capital and Founder of Ibbotson Associates, who taught finance at Yale for four decades
A dollar put into large-cap American stocks a century ago is worth $14,751 today, with every dividend reinvested and no tax, fee or transaction cost taken out.
The obvious question is why that has not made everyone rich, and Roger Ibbotson's answer has nothing to do with stock picking. It is that almost nobody holds the position the arithmetic requires.
"First of all, most people don't reinvest. They actually consume."
Ibbotson built the long-run return series the industry still quotes, first published in 1974, and he has now gone back and marked his own 1976 forecast against 50 years of data that did not exist when he made it. His new book, out with the CFA Institute, carries a forecast to 2050 that deliberately docks the US record.
The full interview is covered here so you can skip it. 43 minutes of audio, 17 minutes of reading.
Here are the 14 lessons that matter.
Key Takeaways
The $14,751 figure assumes no taxes, no fees and full reinvestment, which is why it is not a real-world result
There is no alpha in the number — it is the whole market, not a selection from it
Small caps beat large caps by less than 1% a year, and most of that came between 1974 and the early 1980s
Ibbotson still tells young investors to hold 100% equities, on the grounds that their wealth is mostly future earnings
Bond yields ran from about 2% in 1940 to 14% in 1980, which made 40 years of Treasury returns poor
A hundred years of data is not long enough to contain the outcomes that matter most
His 1976 forecast was almost exactly right in nominal terms to 2000, over- in nominal and under- in real terms since
The forecast to 2050 is 7% nominal and 5.6% real, about 1.5 points below what US history alone would give
Dividends plus buybacks come to about 4% of market value, roughly constant for two centuries
Only 5% of SpaceX stock was newly issued, which he reads as a supply problem waiting to happen
1. Why Isn't Everyone Rich
The book's title is a question, and a host put it straight to Ibbotson: if a dollar becomes $14,751, why has almost nobody managed it? Ibbotson started with what the number assumes.
The number assumes conditions nobody lives under
It's $1 without any taxes or fees or anything, and completely reinvested.
Roger Ibbotson
The first leak is consumption, not fees
First of all, most people don't reinvest. They actually consume.
Roger Ibbotson
Taxes, fees and transaction costs come on top of that.
So the result is hard to reach in practice
So this is actually quite difficult to obtain.
Roger Ibbotson
The one thing the figure does not require is skill.
There is no stock selection in it
But I will say there's no alpha in here. We're just talking about buying the whole market.
Roger Ibbotson
2. The Chart Nobody Had Seen
The series was built in total-return terms at a time when the industry was not reporting total returns at all. Ibbotson dated the practice to the Fisher and Lorie studies at the University of Chicago, which set up the Center for Research in Security Prices.
Total returns were a new idea when he started
We did this in total returns. And by the way, the notion of total returns is pretty new because people were not reporting total returns even when I first did this in the 1970s
Roger Ibbotson
And the full return only accrues to someone who reinvests all of it
So, we're not going to get the full total return unless we reinvest all of it and keep our costs very, very low.
Roger Ibbotson
The chart first appeared in 1974
Well, we first came out with this in this actual chart in 1974.
Roger Ibbotson
The first version had large caps only. Small caps were added a few years later, with help from David Booth and his co-author Rex Sinquefield, who founded Dimensional Fund Advisors at around the same time. The timing of the publication is what made it news: 1973 and 1974 had produced the worst equity returns since the 1930s.
The finding was that stocks beat everything else
And then we come out with this chart which shows that stocks did way better than bonds and way better than treasury bills and way better than inflation.
Roger Ibbotson
Which arrived while investors were asking whether to quit stocks
This is the first time people saw this chart and suddenly they say well maybe stocks are all right to buy because they just had this big crash in the stock market.
Roger Ibbotson
On small caps, the edge over the full century is under a percentage point a year, and Ibbotson placed most of it in one stretch.
Most of the small-cap premium came in one run
A lot of the return from small caps came from 1974 to the early 1980s.
Roger Ibbotson
A host noted that the chart has to be read on a log scale, because the slope is the return; on a linear scale a century of compounding looks like a flat line that jumps at the end.
3. Timing Crashes And Booms
A host raised the pattern that money arrives after markets have done well and leaves when they are cheap, and asked whether a crash or a boom tells an investor anything. Ibbotson said it does not.
A crash or a boom is not a forecast
So just knowing there's a crash or a boom isn't really enough to predict the future very well.
Roger Ibbotson
And the attempt itself is the problem
I mean, people have great difficulty in timing the market.
Roger Ibbotson
His example was 1987, which barely registers on the century chart despite containing the largest single-day fall.
The worst day in the data was a positive year
It was a 20% drop in one day. But actually for the year the total return for 1987 was a slightly positive number.
Roger Ibbotson
4. Human Capital At 100%
A host described a listener in their late thirties who has compounded at 15% a year since starting at 20, wants to hold nothing but equities and might use leverage, and asked how such an investor should prepare for a drawdown that lasts a decade rather than a quarter. Ibbotson's first move was to point at the decade before the current run: the technology collapse of 2000 to 2002 and the 2008 financial crisis sat inside the same ten years.
A bad hit is the base case, not the tail
So you definitely have to expect you're going to have some risk here. Most people are going to in their lifetime get hit badly at some point in these markets.
Roger Ibbotson
Forgetting risk is what a long run does to people
And of course you forget there's risk after you have a long run without the risk. They forgot about the risk after the 1920s and then they got hit by the 1930s.
Roger Ibbotson
He would not endorse an all-equity portfolio in general.
He will not recommend 100% equities as a rule
So there definitely is risk in the market and I wouldn't recommend you go 100% in stocks.
Roger Ibbotson
For the young, the answer changes, and the reason is that a young person's largest asset is their future earnings rather than their portfolio.
What Roger Ibbotson thinks about age and equity weight
I definitely recommend young people go 100% in stocks in their investment portfolio, but as you get older, they should be more risk-averse actually because you have less time to recover and your human capital gets diminished.
Roger Ibbotson
He applied it to himself: still working, so some earning power left, plus social security, but with financial capital now doing almost all of the work.
5. The Bond Mountain
Asked about the chart of bond yields that rises from the 1940s to 1980 and falls back to near zero, Ibbotson walked both sides of it.
Yields bottomed around 1940 and peaked in 1980
So you can see that the yield pretty much bottomed about around 1940 but then climbed all the way up to double digits in 1980.
Roger Ibbotson
The cause was inflation, and the resolution was Paul Volcker cutting the money supply to bring both inflation and rates down. What the rise did to bondholders is mechanical.
The arithmetic of a rising yield
So if you buy a long bond like with a 10-year duration say, and when the yield goes up a percent you lose 10%.
Roger Ibbotson
And the move was 12 points
And you can see the yields going from two to 14 here.
Roger Ibbotson
So the four decades to 1980 paid a coupon and handed back capital losses. From 1980 to roughly 2022, when yields fell under 2%, holders collected a high starting yield and capital gains on top. The last couple of years have run the other way, with yields now above 5% and bond returns poor again.
The driver underneath all of it is expected inflation
The most important driver of all this is the expected inflation
Roger Ibbotson
6. Fat Tails And Ruin
On the probability of ruin, Ibbotson's point was that the usual distribution is not wide enough. Equity returns are asymmetric, since losses stop at 100% and gains do not, and they carry a second problem.
The distribution has fat tails on both sides
But the other thing they have is kurtosis, it's called, and kurtosis is fat tails.
Roger Ibbotson
That admits outcomes the historical record does not contain, including bankruptcy at the level of a whole economy.
A century is not a long sample
Despite how long a hundred years feels like in terms of statistics and mathematics, it's not that long.
Roger Ibbotson
The line he uses for it
So, we often joke that to be a good investor, you got to be part historian, but also part comedian because you realize things will happen in the future that have never happened before.
Roger Ibbotson
7. Correlation Is An Average
A host raised the trap in the correlation table. Large, mid and small caps run at 0.9 to 0.97 with each other, which surprises nobody, but small caps against long bonds is close to zero, and readers take that as a guarantee of diversification.
The host's objection: zero is the average, not the behavior
But what I don't feel like they realize is that's the average.
A host
In 2022 both fell together. Ibbotson agreed with the framing.
He does not call them uncorrelated
Yeah, they're not so non-correlated.
Roger Ibbotson
The relationship persists in whichever state it is in
Sometimes they're very negatively correlated and when they're running negatively correlated that'll go run for a while so you can predict that they're negatively correlated
Roger Ibbotson
The same holds when the correlation is positive, which makes it somewhat predictable over short runs even though the century average is about zero.
8. The Popularity Premium
Ibbotson's framework is supply and demand, and his claim is that the demand side explains more of the return differences between assets than risk alone does.
Popularity is the demand side
Well, popularity is about demand. People pay different amounts for the same for similar things at times depending on how popular it is.
Roger Ibbotson
The original version of this is the capital asset pricing model: risk is unpopular, an unpopular asset is priced lower, and a lower price is a higher expected return.
Which is why equities pay more than bonds
So that's why the capital asset pricing model says that stocks have higher returns than bonds because stocks are unpopular.
Roger Ibbotson
The second unpopular characteristic is illiquidity
The most obvious second one is liquidity. People like liquidity. If you give them something less liquid, you have to offer them a higher return, a higher expected return.
Roger Ibbotson
The third is reputation, and here the effect is behavioral rather than compensatory.
A good brand is paid for out of future returns
Companies that have good reputations, even though two companies might have the same cash flow expectations, the one that has a good brand with a good reputation will go for a higher price than the one that has a poor one, even if you have the same expected cash flows.
Roger Ibbotson
Companies people do not want to own end up cheaper and, on his account, higher-returning. The model also runs in reverse, which is what makes it a description of inefficiency rather than only of equilibrium.
Popularity can overshoot
Things get too popular and not sustainably popular.
Roger Ibbotson
9. Grading A 1976 Forecast
The building-block forecast published in 1976 now has 50 years of out-of-sample data against it, and Ibbotson went back to check it. A host noted that the new version reproduces the original in the same font.
He grades it as accurate
Actually, it's pretty accurate. It was almost perfectly accurate out in nominal terms out to 2000.
Roger Ibbotson
And he took the credit at the time
And so I went on a lot of TV shows in 2000 and took credit for perfect forecasts
Roger Ibbotson
The errors since then point in opposite directions
In nominal terms, because inflation was high, we overestimated. In real terms, we underestimated, I guess, but basically we're still pretty close across the board there.
Roger Ibbotson
He was blunt that the honest product of the exercise is a distribution rather than a number, and that a distribution is not what anyone wants.
The accurate answer is the unsatisfying one
We have no idea really what's going to happen. We have a wide distribution of possibilities here.
Roger Ibbotson
His analogy was a blackjack hand: sitting on 20 against a dealer's six is a good position and not a guaranteed one.
10. 7% Nominal Out To 2050
The new forecast runs to 2050 and does not simply extend the historical US equity risk premium. The objection Ibbotson accepts is that his original work measured the one market that happened to win.
He docks the US record for survivorship
So we actually take the Dimson Marsh Staunton differential, the fact that the average return is less than the US return, to make our forecast because we don't want to just assume that the US will be the winning country over the next 25 years.
Roger Ibbotson
Two chapters of the book are written by Elroy Dimson and his co-authors on 125 years of global data, in which the United States is almost the best-performing market. Markets like Argentina looked strong in 1900 and did badly; losing combatants in the world wars saw their markets broken up.
The haircut is about a point and a half
So for that forecast it turns out to be I think about one and a half percent less than the US forecast when you look at the average of all these countries.
Roger Ibbotson
Asked whether that means cutting a 15% expectation to 13.5%, Ibbotson gave the actual figure.
The forecast is 7% nominal, 5.6% real
We're forecasting 7% nominal and 5.6% real. That's the median, the mean forecast, not the whole distribution.
Roger Ibbotson
He made the same point about the historical record. Inflation multiplied 18 times over the century, so the headline figure shrinks considerably in purchasing-power terms.
$14,751 nominal is 820 times in real terms
So after inflation you made 820 times your money.
Roger Ibbotson
Bond forecasts start from the yield, since the market's inflation expectation is already inside it. The book's numbers are struck at the end of 2025; yields have risen since, which has made this year's bond return negative.
11. Why CAPE Calls It Late
A host raised the cyclically adjusted price-earnings ratio, the work of a Yale colleague, and the fact that it is close to the levels printed in the late 1990s. Ibbotson's answer was about timing rather than validity.
The signal works, the timing does not
The CAPE data has been pretty good historically. It basically calls things really late.
Roger Ibbotson
Robert Shiller's Irrational Exuberance appeared shortly before the 2000 collapse, but Shiller had left the market several years earlier and missed the late 1990s. Ibbotson credited the call and blamed the entry point.
Overvalued today, and overvalued in the mid-1990s too
So getting the timing right is really difficult basically, and yeah the CAPE ratios would basically say the markets are way overvalued today, but they also said that in the mid-'90s.
Roger Ibbotson
12. Buybacks Make It 4%
Asked what he finds strange about the present, Ibbotson took the S&P 500's roughly 1% dividend yield, and said the yield is measuring the wrong thing. He has written a paper on buybacks with a co-author at Morningstar.
Dividends plus buybacks come to about 4%
Basically, if you consider buybacks and the yield together, it's about a 4% cash out.
Roger Ibbotson
And 4% has held for two centuries
That 4% is pretty much a constant number over the last couple centuries of the kind of payouts that you had on the stock market.
Roger Ibbotson
The switch began in the 1980s, and his argument for it is tax and timing: a buyback is taxed as a capital gain and lets the holder choose when to take cash out.
He reads the change as a financial innovation
But basically, buybacks are much more tax efficient than getting your money through dividends.
Roger Ibbotson
So the 1% yield does not alarm him
And so I'm not alarmed by the 1% dividend yield. I think we're getting our cash out in different ways.
Roger Ibbotson
He added that the practice is spreading, naming the United Kingdom, Japan and China. The exchange closed with the observation that politicians and journalists treat buybacks as the villain while managements understand them as more flexible, more tax-efficient dividends.
13. SpaceX And The New IPOs
Asked what is on his mind now, Ibbotson named the largest technology stocks and the newly listed SpaceX.
He does not think the market knows how to price it
But basically, people don't seem to know how to value things like that at all.
Roger Ibbotson
A host disclosed a long record of declining the private deal.
The host passed on SpaceX 24 times
I passed on investing in SpaceX 24 times as a private investment over the years.
A host
The bear case put on the table was about share supply rather than the business.
Only 5% of the stock is new
My view on SpaceX is there are all these people who bought in as part of venture capital or private equity in different forms and they made a bundle and basically they only issued less than 5% of the new stock.
A host
And the old holders are sitting on enormous gains
I think there's a lot of supply on the market now of people who want to get out of SpaceX
A host
One beneficiary named was the University of North Carolina's endowment, which a host said he had recently seen reported at something like 10% SpaceX after a position went vertical. On the wider question of a listing wave, Ibbotson's own dissertation was on initial public offerings.
His dissertation found IPOs deliberately underpriced
I've actually done a lot of work on IPOs. That was my original dissertation and I showed how the IPO market basically was underpriced, purposely underpriced basically.
Roger Ibbotson
A 15% pop, then nothing
I did my original work in the 1970s but the average stock at that time went up 15% on the issue and then didn't do so well afterwards basically
Roger Ibbotson
He also wrote a paper on hot issue markets, and the pattern it describes is that cold markets are not cheap markets.
Cold markets do not reward the few issuers either
There are periods when a lot of IPOs are issued and they all go up and then there are periods, or cold markets, where hardly any issues are issued but they don't go up either.
Roger Ibbotson
A host raised prediction markets now quoting odds on individual listings, and asked whether supply is the thing that ends the bull run. Ibbotson expects the companies to come, and said the composition has changed.
Fewer listings, far larger ones
Now these companies are already big before they go public. So we're having not a lot of companies going public, but the ones that are going public are going out at big capitalizations.
Roger Ibbotson
14. Comsat, 1964
Asked for his most memorable investment, Ibbotson went back to a teenage trade.
The IPO that started his career
I bought Communications Satellite Company as an IPO back in 1964, and I bought it, I knew it was way oversubscribed, and so I bought it at 20 and it popped immediately to 24 and over the summer it went up to into the 60s before I sold it.
Roger Ibbotson
The profits helped pay for college and, by his account, convinced him he was good at picking stocks. Then he arrived at the University of Chicago.
Where he was taught the other explanation
By the time I got to the University of Chicago, of course, they taught me about efficient markets and buying the whole market.
Roger Ibbotson
His own verdict on the record
Well, fortunately, I was in a lucky period.
Roger Ibbotson
He dated it to the Nifty Fifty years of the 1950s and early 1960s.
Bonus Insights
The tails are what a data set cannot hold
Things happen that weren't in the data set.
Roger Ibbotson
Two smaller exchanges are worth keeping. Ibbotson said the popularity model applies to efficient and inefficient markets alike, which is unusual for an asset-pricing model and is the reason he treats overpricing as a demand phenomenon rather than a mistake. And on the book itself, he said the published articles behind it forecast every line item, while the book covers a selection; readers are pointed at the CFA Institute edition and at his Yale listings.
Ibbotson's bottom line is that the century of returns is real, the arithmetic is not available to anyone who consumes, pays tax or pays fees along the way, and the right forecast for the next 25 years is about a point and a half below what the American record alone would suggest.
Products, Companies & Tools Mentioned
Zebra Capital and Ibbotson Associates (The firm Ibbotson chairs and the firm he founded, which built the long-run return series the industry quotes)
Dimensional Fund Advisors (Founded by David Booth and Rex Sinquefield, who helped add the small-cap series to the original 1974 chart)
SpaceX (Newly listed, and the case against it on the show was share supply: under 5% of the stock is newly issued)
Anthropic and OpenAI (The listings the hosts expect next, and the reason the supply question came up at all)
CFA Institute (Publisher of Ibbotson's new book, where the full forecasts to 2050 sit)
Books & Resources Mentioned
Exponential Wealth: A Century and More of Stock and Bond Returns – Roger Ibbotson (The book the interview is built on, with chapters by Elroy Dimson and his co-authors on 125 years of global returns)
Irrational Exuberance – Robert Shiller (Published shortly before the 2000 collapse, which Ibbotson used to argue the CAPE ratio calls things late)
If this was worth your time, send it to someone who has to have a view on this.
Get the latest market chatter as it happens: