Wow. A huge amount of data here. I like the inclusion of dividends,
discussing the index (any index) is silly if dividends are ignored. Over
a long enough period the start/end for an index is meaningless. So
including inflation works for me as well, although I think inflation is
different for two different people. Scary to see how many 20 year
returns failed to keep up.
Thanks for posting that. Fascinating. (I do wish we could get the raw
data in an easier format to work with - I'd love to see, say, the median
30-year numbers just like they show the 20-year numbers there)
And it amazes me just how often the dividends are ignored. I'd say
more often than not. (And that's without even starting to talk about
Presumably, they used the CPI-U which, as we all know, is no individual
person's actual inflation experience, but rather an aggregate estimate.
I'd like to know how they figured "average taxes and fees" though.
If they are assuming average marginal tax rate on the dividends with
the remainder reinvested, when do we see the capital gains (at the end
point of each period assuming a full sell-out?) And "fees" - are they
also assuming an average mutual fund's expense ratio (which is vastly
higher than a well-run index fund's expense ratio). Those "fees" could
easily be eating up a full percent per year.
I take it back - poking around, I found some answers to these
questions - here:
These two assumptions are unfortunate:
They "realized" 90% of cap-gains each year as long-term gains.
And they charged "fees" of 2% (!) per year from 1900-1975, then
1.5% until 1999 and 1% since.
Still, it's a fascinating study. Thanks again for posting it, Rich.
These fees are not what a long term investor would experience. The
charts are close to what I'd like, but I'd rather see zero on fees than
this. My S&P index in my 401(k) has a .05% expense. 1% over 20 years is
nowhere near the 20%+ assumed here. No wonder the chart is so grey,
tough to overcome 2% plus inflation.
For those of us who are not sure what you meant by the above, let me
ask a question.
Are you saying that equity-indexed annuities do not index to the
dividends-reinvested index, but to the price-only index instead?
I've analyzed two EIAs, both did not include the dividend. When you go
back to the Crestmont Assumptions, you see a 100 year average of 4.3%.
Even if we are at a new plateau of lower dividend yields, it's still
money out of the client's pocket. In adition to the other EIA rules such
as capped or a percent participation to the index itself, as well as fees.
It only adds insult to injury that these products are regulated
differently (than stocks or mutual funds) so the information we seek (a
clear prospectus for analysis) is not readily available to the public.
(any seller of these products wishing to sent me a full prospectus, not
a marketing flier, I'd be happy to see it. This is a working email I
On the origins and meaning of a chart printed in the NY TImes in 2011:
Bread, well done.
I think the NY Times chart should be labeled as an advertisement for
Crestmont Research (founder, Ed Easterling) and a plug for
Easterling's book, _Unexpected Returns_.
Newspapers everywhere need to have a sub-section of the Business
section called, "Business Gossip and Fanciful Thinking" with the
disclaimer: "Someone stands to profit from the product this article
I looked at the chart for general validity. At first glance, I thought
it was undoing all the arguments for long-term stock investing. Then I
read the line towards the bottom that says, "After accounting for
dividends, inflation, taxes and fees, $10,000 invested at the end of
1961 would have shrunk to $6,600 by 1981." This seemed a cause for
60 to 133 for an average annual return of about 4%. The average yearly
dividend yield was over 3%. The yearly annual return before inflation,
taxes and fees is therefore 7%. Then Ed Easterly (not the Times and
not some generally agreed upon expert) has the gall to throw in
assumptions about fees, taxes and inflation. The fees assumption is
not applicable to today and in my opinion should not be used for
future decisions regarding investing. As for inflation and taxes, why
Easterly fails to do the same study for bonds is interesting. I
suspect Mr. Easterly (formerly a used car salesperson?) knows that
bonds did similar or after taxes, worse.
Stuffing $10,000 dollars in a mattress from 1961 to 1981 and using the
CPI over this period would appear to result in a reduction to about
So what should a person have done from 1961 to 1981? Mr. Easterling
elsewhere on his Crestmont "research" site will tell you: Time or buy
into a hedge fund. The following excerpt from his book is surely one
of the most eloquent pitches for timing I have ever seen:
. Whereas studies time and again show the hazard of timing and hedgefunds.
Mr. Easterling and his company sell a hedge fund. His chart argues
against buy-and- hold, to me via a not very clever manipulation of
Well, if you had felt like potentially learning something instead of
going off on yet another one of your tiresome ad hominem rants against
anyone who dares to attempt to make money as a manager or adviser, you
might have noticed that there's a whole set of those charts at his
website, including ones that show nominal returns with no adjustments
for taxes. You would then have been free to make your own assumptions
about what taxes and inflation might be and how that could affect
things instead of complaining about assumptions you disagree with.
But doing that would have deprived you of the chance to get up on your
self-constructed high horse.
Rich Carreiro firstname.lastname@example.org
Easterling's chart's assumptions are off-the-wall; investing for the
long term in a mix of stocks and bonds has been studied and proven;
timing and hedge funds have been established to be a gambler's game.
The chart is another interesting example of how the latest investing
fashion sells in the media if only because it seems new. Easterling's
writings are a flash in the pan compared to those of Siegel, Shiller
and Ben Graham. My post is directed at the interested, thoughtful
reader who should hear all opinions and then decide for him/herself.
Different (and more) colors, but I've removed fees and taxes. If
you'd like more permutations, I can cook some up.
Your friendly, neighborhood applied mathematician,
Rich/Elle - I appreciate the time everyone spends sharing here. I'm
certainly glad Rich brought this article/chart to our attention.
If we can get away from the personal attacks, I'd just like to discuss
the effect of the Times publishing such charts without the data to
explain it. People tend to see things the same way we hear, in sound
bites. The takeaway from the Time's chart is that stocks have brief
periods of excelling, but far more frequently barely stay ahead of
inflation. How would the typical investor react to this?
After seeing the rest of the assumptions, I'd only ask why they (the
Times and Author) chose to ignore.
Over the years, I've become cynical as to how any data, especially those
I can't easily reproduce, are constructed.
Dalbar has a great report (QAIB 2009- Extract of Quantitative Analysis
of Investor Behavior) claiming the 20 year return of the S&P (ending
12/31/08) was 8.35%, but the average equity investor saw a return of
1.87%. The number was not inflation adjusted.
The 2009 report (2010 public available report did not show the numbers
as clearly as '09) is still available at
In article ,
Peter Lynch made a similar observation (can't remember where I saw it)
that most of the investors in Fidelity Magellan never made the returns
that the fund did because they tended to buy high (or near high) and
sell somewhere near low. .
And here's a similar source document from the original author that
the NYT got their chart from.
This one (right from the source) includes dividends but omits
both inflation and taxes (I agree with "Bread" that the
assumption of realizing 90% of gains every year is unrealistic):
a few points-
first, thanks for hanging in, and staying with this thread
second, this chart is more to my liking, as I can pull inflation out if
I wish, or at least adjust on my own.
last - I've never been more thankful that I sit in front of dual 24"
EIA's are popular, largely due to the appeal of the "you match the
market on the way up, and are guaranteed not to lose money when the
market goes down" idea. At least this (or something very similar) is
how EIA owners have explained their buy decision to me.
Yet when I review the difference between tracking a price-only SP500
index and a total return SP500 index, what I usually get is a blank
stare. That indicates to me that either they were not advised of this
when they bought the annuity, or they suffer from a uniform lack of
Here's a link that suggests that over the past 80 years, dividends
have accounted for 44% of the total return - a significant statistic.
my question to EIA promoters is this: Do you cover the differencebetween indexing a price-only index and a total return index in yourpresentations? If so, how do you portray a price-only as a benefit tothe buyer?
(Apologies in advance to those who actually index to a total return
index. I know of none, but if anyone has evidence to the contrary,
please correct me.)
Ironic. The fees that Crestmont showed are below the underlying fees
plus dividends which are the extra cost of EIAs. So, at best, that 20
year line on the original chart was more indicative of an EIA investor's
I am just ignorant enough of insurance regulations to really not
understand why the information doesn't flow. We can openly discuss
mutual fund expenses, returns over time, correlation to any data set we
wish. These pseudo-investments (not pejorative, but they are an odd
beast) are not regulated like stocks, but sold as stock replacements.
Every last salesman of this product in particular will explain how I
don't understand them, yet won't spell out the equations used to produce
the results. As I mentioned, the two I analyzed were from people who
were about to buy, the had all the paperwork ready for signature.
The last one mention here certainly had a missing piece as it was
described as giving up only 3%/yr to clip all down years. Too good to be