Still Stocks Rather than Bonds for the Long Run

J. Bradford DeLong and Konstanin Magin (2008), "The U.S. Equity Premium: Past, Present, and Future" http://www.j-bradford-delong.net/2008_pdf/20080228_jep_submit.pdf

ABSTRACT: For more than a century, diversified long-horizon investors in America’s stock market have invariably received much higher returns than investors in bonds: a return gap averaging some six percent per year that Rajnish Mehra and Edward Prescott (1985) labeled the “equity premium puzzle.” The existence of this equity return premium has been known for generations: more than eighty years ago financial analyst Edgar L. Smith (1924) publicized the fact that long-horizon investors in diversified equities got a very good deal relative to investors in debt: consistently higher long-run average returns with less risk. As of this writing the annual earnings yield on the value-weighted S&P composite index is 5.53%. This is a wedge of 3.22% per year when compared to the annual yield on 10-year Treasury inflation-protected bonds of 2.31%. The existence of the equity return premium in the past offered long-horizon investors a chance to make very large returns in return for bearing little risk. It appears likely that the current configuration of market prices offers a similar opportunity to long-horizon investors today.

Comments

[Charles]

Why does this the phrase "Dow 36,000" come to mind?

The equity premium was lower in the 19th century and rose in the latter third of the 20th century. One suspects that social/political factors played a role in explaining these facts.

At this moment, the American empire is in decline. Many of the advantages that it enjoyed, notably the best technology startups in the world, are being lost, changing the investment environment. Could the basic factors be changing?

Similarly, emerging markets and small cap stocks may have much higher growth rates than the S&P, and perhaps greater equity premiums. Could the increasing concentration of corporate wealth cause the equity premium to decline?

These are some of the questions that would pop to the head of the line were I refereeing.

[VennData]

Sprint buys Nextel, loses $30B in a couple of years. Knowledgeable managers made - what turned out to be - an error.

Sometimes it's fraud (Enron, MCI.) Other times it's a foreign competitor coming out of nowhere (GM, Ford) or you've got an innovative idea but Microsoft copies it, adds it to their bundle and charges zero; killing you - a threat - off (the largest sink hole of "innovation" in the US economy.)

Not only is business risky and very hard, but there's a huge amount of luck involved, ergo, most of the time you get equities at a discount. That's the time to buy.

[maynardGkeynes]

1. Earnings yield is not the same as dividend yield. I won't buy AOL with my retained earnings, like Time Warner, nor blow billions on Yahoo. I will put my kids through college with the "earnings."

2. Try a 20 year TIPS and get approx 2% plus inflation. That will could be pretty close to your 5.53%, but with ZERO risk.

[The 5.53% is an ex-inflation number. The gap is about 3.2% per year. That looks to be a lot.]

[Bruce Ferguson]

I used to know a professor at Cal that had a license plat "T BONDS' they were like 15% at the time. Well I never knew him well.

[christofay]

No, buy a TIP and get about CPI + 2% which approximates the inflation rate.

Business Week had a cover story in the early 1980s "Are Equities Dead" which was the starting gun to buy equities. There were decades within the past hundred years when equities didn't perform, the 1930s through the point the Dow Jones passed its earlier high in the 1950s. Equities need dividends to outpace bonds, it's hope in man's endeavor plus income that does it. The last couple of decades have seen appreciation, perhaps just asset price appreciation, become the dominant aspect of equities, good-bye dividends. Unfortunately you picked an observer from the 1920s and here we are today standing on the edge of another recession.

Business itself sucks and makes you sweaty. I haven't heard of any meaningful start ups getting going. It's best to get a degree from one of the dozen good schools and do meta-business on Wall Street trading borrowed money back and forth

[a]

If you look at enough statistics, then you will find at least a few that seem to be very long-run and meaningful. But sorry, past performance is no guarantee of future performance.

[PSP]

In 1980, the youngest baby boomers turned 35 and first started thinking about retirement. They have been (or should have been) increasing their savings and investments ever since.

So what happens when the draw down is in full swing?

[David Lloyd-Jones]

PSP,

Net net there is no draw down. Marginal propensity to invest and net investment both increase monotonically by age of cohorts.

Sure there are some folks who sell their houses and eat the money, but they are the low end of old investors.

Overall the thing that gets drawn down might be Social Security, but remember that's just cashing in bonds which have already been issued and charged off in years past. The Big Picture of retirement is that on the whole people live off part of their income and part of their capital gains, reinvest the rest in amounts even larger than when they were working.

[Jack]

There is a spelling mistake in footnote 13.

In Mehra and Prescott and other Lucas model based analysis, isn't it also a puzzle that equities trade at all? M&P at least assumes away any exposure to the price of equities and therefore are underestimating the denominator in their risk/reward calculations.

Is it too hard to combine several of the approaches catalogued? There seems no reason not to have sticky preferences, liquidity constraints, transaction costs and scary downsides at the same time.

Do you dismiss the massive lightening of the tax burden on equity investments as noted in McGrattan & Prescott as a significant contribution to historical estimates of equity returns?

To me the equity premium puzzle looks like a situation where overlapping generations would be a more natural approach than the single representative agent that predominates. Minford's OLG model at least doesn't find a puzzle.

It would also be good to ask why equity doesn't make good bank capital.

[Jack]

DLJ, Are those trends you can extrapolate? Will there really be no drawdown of investments tracking the drawdown on the social security fund? Even a slowdown could have some effect.

[Jack]

DLJ, Are those trends you can extrapolate? Will there really be no drawdown of investments tracking the drawdown on the social security fund? Even a slowdown could have some effect.

[robertdfeinman]

Up until the 1990's the proportion of the population with investments in stocks was very small. It started to increase after this time because of the creation of new instruments that individuals could participate in. The most important have been 401K and IRA plans as well as mutual funds.

This has put a continual demand on stocks as people accumulate assets in these plans. Increased demand led to an expansion in the number of shares on offer as well as a rise in their prices. I distinctly remembering pundits touting mutual funds because they claimed that the demand from retirement funds would mean that demand would always continue. I think this was responsible for much of the rise over the past 20 years. Of course the constant demand and the abdication of all oversight by federal agencies has also led to a rise in Ponzi schemes and their modern variants. Nothing like greed to bring out the "best" in people.

Before this period the group of investors was quite small and could extract value from companies using the usual capitalist techniques of underpaying workers and monopoly pricing. Shifting wealth from the mass to a small elite is a time-tested technique. This made share investments' yields higher than otherwise.

We have now reached (at least in the US) a stage where the ownership pressure on shares has tapered off as the number of new people entering the investment process becomes more of a replacement business. This is why Wall Street is so keen to privatize Social Security, it is the only way to boost demand beyond normal demographic effects.

If there is to be a continuing yield benefit to stocks someone will have to explain where the extra value is going to come from.

[Bram Cohen]

This link has some analysis which claims a flat-out lower number for long-run inflation-adjusted market returns, using available statistics. Any idea what the cause of the numerical discrepancy is? Or is it the analysis which is different?

[Bram Cohen]

Sorry, forgot to include the link, here it is

http://homepage.mac.com/ttsmyf/

[Dave]

I have noticed that S&P index funds rarely do as well as the S&P. I am told this is because the S&P simply "gets its money back" if one of their companies is disqualified. Index funds do not get any such refund that they can simply re-invest in another stock.

Over the course of many years, the S&P assumptions almost always work to make the average return higher, rather than lower. I have heard that it is relatively hard to incorporate "assumed" profits into your actual portfolio.

So I carry a few AA munis. They are dogs, of course, but I have never yet opened the paper to find that they are "so last week" or that they are under investigation for gaming the oil market. If I had sold them when interest rates were up, I would have lost money, but I didn't sell them.

Dare to be dull.

[Sonia]

I haven't checked all the numbers in Bram Cohen's cite however one important distinction between return on equities and the price of an equity index is that the return on equities includes dividends. An important component of the return on equities through at least the 1980s, was dividends.

One of the arguements for a continuing high return on equites through the retirement of the baby boom generation is that foreign residents particularly in developing countries will want to invest in US equities,buying up the stocks sold by retirees. I am skeptical of this arguement. Another arguement advanced above is that retirees will not suddeningly change their equity allocation. I am more convinced of this. Remember that equities are disproportionately owned by high income wealth famlies who have no reason to change their allocation.

[Charles]

Robert Feinman, the internationalization of stock markets means that there is another mechanism to get new investors if the old ones...well, get old: shop overseas. I would imagine this is why Citi was so fixated on Nikkei Cordial.

But, so far, international investing is more-or-less limited to investing in big companies. This could produce some interesting and probably very bad effects, as capital gets skewed even more heavily away from startups.

[Bram Cohen]

Sonia, it's my understanding that the DJIA includes dividends (because, duh, it wouldn't tell you much of anything if it didn't) and that some of the more outrageous overclaims of stock market returns are due to people thinking that dividends weren't counted and adding them in again, thus double-counting them.

[phyron]

I don't quite understand...
I'm not quite sure where the comparison with "bonds" comes into play in Your paper..
Simple Bond math tells you that there is not much difference between the "return on a coupon bond"...(ex post) and the realized path of short term rates at which you Reinvest your ostensible coupon..
The real comparison is with the purchase of a similar horizon zero coupon bond..
I like to call them Bar Mitzvah bonds... in light of your introductory paragraphs..
If I can buy a zero coupon bond at around 5.5% not to far from where they have traded in the recent past ... then I am certain of doubling my Money in 13 years... certain of quadrupling it in 26 should buy a 25 year zero...
Even more compelling if now we have a vigilant Central bank that will not let inflation get out of check... and whose actions ex ante might be thought to flatten the yield curve and make longer term bonds even more attractive...
And of course I have an extra kicker... if "rates" decline not only do i have certain terminal value.. but i can capture extraordinary captial gains..
This year alone.. with risk in the equity market increasing apparently..
the happy purchaser of 2023 Zero Coupon bonds in June of 2007... would have saved himself a 15 % loss on his equity portfolio... and earned on a mark to market basis.. nearly a 25% return on his risk free Treasuries..
thats a net differential of 40% in less than a Year...
The gambler as Mr.Keynes pointed out is the man who ignores odds like these...
I can think of no greater historical episode where the Barro scenario...
everything but bonds go down is played out...
As we age.. missing this one instance.. may be enough to make or break your entire future path...
Ex post the decision as to whether to be in stocks or bonds has made a 40% difference in your wealth..

[Mats]

Good read as far I can tell, really interesting! Some random reflections:

The references to T-Bond returns are dropped (if I read it careful enough) when it comes to the 90's and onwards and replaced by T-Bills. Probably because leveraged bets on Gov't bonds, up to the risk-level of equities, outperformed. Doesn't change the story though, isn't it really about a *risk*-premium puzzle?

Didn't some Tsar bonds actually pay out after the Sovjet perestrojka making these *much* more profitable than pre-revolution interest in Russian companies?

Isn't it ironic: it would be too few rather than too many of the perma-bullish sell-side equity analysts!?