Down goes the Dow

The difference between the average annual change (9.3%) and the expected value (< 8%) is one of the costs of volatility (or risk). Note, though, that while nearly 8% is "expected," it is not guaranteed, and you are fairly likely to end up with quite a bit more or less.

Now for retirees, or others thinking about drawing down their portfolios, the volatility becomes more important. If we imagine somebody taking $50k every year (adjusted for inflation, assumed to be 2% in this case) from that initial $1 million from the same 25 imaginary starting points, we get the following results:

ibb.co

Slide10

Image Slide10 hosted in ImgBB

Open link

The average outcome after 25 years of withdrawals is about $2 million remaining in the account. Of course you could do better and end up with around $6 million. But the problem is that 5 of the imaginary trials (20%) end up broke before 25 years are up. If I tweak inflation up to 4% you get 9 of 25 trials running out of money (36%).

Let's come back to the idea of using CAPE, or the like, to predict future returns. I dislike CAPE and any derivates because it tries to rely on factors that correlate with market activity, but which may not have constant relationships over time. Personally, I like to take the simplest, most naive view possible, and only look at the index itself. Let's look back at the index values over time with a logarithmic y-axis, including the best fit logarithmic trend which is a straight line in this plot:

ibb.co

Slide1

Image Slide1 hosted in ImgBB

Open link

If I divide the index value by the best fit, you get this:

ibb.co

Slide2

Image Slide2 hosted in ImgBB

Open link

In this plot of "ratio to best fit", if the curve is above 1, the index was doing better than the long term average trend. If it was below 1, the index lay below the long term trend. In other words, for values over 1, stock prices were high relative to the long term average (or P/E would have been higher than average), and for values lower than 1, stock prices (or P/E) were below the long term average.

If I look at the value of this ratio (i.e., the position of the index relative to the long term average) at any given time and then check the subsequent 20-year change, I get the following plot:

ibb.co

Slide3

Image Slide3 hosted in ImgBB

Open link

Low values of my ratio (index lies below the long term trend) give generally higher 20-year annualized returns, and values above 1 (i.e., the index is above the long term trend) give generally lower 20-year returns, with a pretty strong correlation (some scatter, but not too much).

If I plot the same correlation between Shiller's CAPE10 values and 20-year returns, I get the following, which has more scatter, suggesting weaker predictive power:

ibb.co

Slide4

Image Slide4 hosted in ImgBB

Open link

If you like, we can compare my ratio (of index value divided by long term trend) to CAPE10 over time:

ibb.co

Slide5

Image Slide5 hosted in ImgBB

Open link

You can see the character of the two curves is very similar, except that CAPE10 increases gradually over time. This is why I suggested that a CAPE10 value of 20 around 1990 is about the same as a value of 13 in 1950 and a value of about 27 today, which is why I'm much less alarmed about today's CAPE10 value than Shiller's disciples, since I think their metric is just too cute, trying too hard to make magic from things that correlate to the index, rather than from the index itself.

If you look back at this plot I shared earlier:

ibb.co

Slide3

Image Slide3 hosted in ImgBB

Open link

The current SP500 value is about 1.4 times that of the long term trend, suggesting 20-year annualized returns of something like 2.5 to 7.5%, with a best guess of ~ 4%.

4% annualized returns for the world's dominant stock index would be pretty lame, and I think I can currently do better with bonds, with better returns and far less volatility. I would love the markets to drop 40-60% and give a new buy-in opportunity, and I think that day will come, but don't know when... could be this year, could be in 10 years.

OK, end of long rant, let's get back to the regularly scheduled programming. :-)

Comfortably - are you forgetting to include dividends? Over the last 100 years 40% of the total return of the S & P has come from dividends.

The average annual return of the S&P 500 over the last 100 years, assuming dividends are reinvested, is 10.628%1. Adjusted for inflation, the average stock market return (including dividends) is 7.454%1. Since its inception in 1928, the average annualized return of the S&P 500 is 9.90%2.

You can always input your portfolio and draw-down plan to FiCalc and see how it would have performed historically. If it survives with a start date of 1965/66 you're safe.

Yep, sure.

However, dividend payout has varied substantially over time. Current dividend yield of SP500 is around 1.5%, which is much, much lower than the long term average. Mainly thanks to low or no dividend payout of the monster-cap tech stocks currently carrying the index.

Thanks for that. If I use $50k inflation-adjusted withdrawals from an initial $1 million it gives 79% chance of avoiding going broke, very close to the result I shared above.

I could be wrong but I think I remember reading that most FA's consider an 80% (or better) chance of success to be a good plan.

Thanks for your work CRFH. Interesting discussions as well from all posters.

Adding to discussion, Hussman this morning:

I’m not sure what FAs consider good, but I think each individual has their own comfort level. Personally I don’t want any chance of depleting savings before we reach the finish line, which makes me wired toward more conservatism than most. But to each their own.

This is irrelevant navel gazing, IMHO.

No funds, only stocks. I don't trade much; I try to find sure things at a good time and hang onto them.

Agip, I find this very interesting. And frankly, it's all part of my general impression of what to expect if he wins.

IF he wins: Markets will rise in the short term, but really only if Congress does not also go Republican. Markets prefer balance of powers. Markets will like his lower coprporate tax rates in particular.

Then markets will be very fidgety in that DJT represents uncertainty, which markets don't like, and his vague econimic platform (if you can call it that) is not taken seriously, so no one knows what to expect from all that bluster except that he will ensure (not insure) lower tax rates for business. The markets will dodge around with every hint of what is actually going to be implemented.

And to your concerns, the markets will most likely sink if even half of what he promises about tarifs and social security, for example, are actually implemented. But those are not a given, and his favorable handling of corporate taxes is.

My view is that it is too early to be positioning oneself for these possible outcomes.

I guess it's a two-part arbitrage:

I think Trump is serious.

I think he will at least make a sincere attempt to expel large numbers of workers. Enough to do some damage to migrant-dependent businesses.

and I think Trump will succeed in putting tariffs on a good amount of stuff.

and I think Trump's success will hurt stock prices.

Most people think either

a) he's just kidding about tariffs and mass deportations

and/or

b) even if he succeeds it won't do a lot of damage.

I guess the reason I'm not selling now is b). that he may succeed in doing these things but they're not enough to derail supertanker America.

yeah we've been in a period where the big companies everyone knows...apple, google, meta, microsoft...have been gold mines for those who buy and hold them.

It's made a lot of people a lot of money. The most popular trade has been the most winning trade. Remarkable and great for investors.

I love how Hussman cherry-picked the year 2000 right before the market collapsed. What a weasel.

I am not sure how MSFT has done since 2000, but since 2004 MSFT is up 2,256%.

Since 2004, MSFT has had an average annual return of over 17%.