« Equity Returns in the Future | Main | Why Oh Why Can't We Have a Better Press Corps? (Yet Another Washington Post Edition) »

January 04, 2005

Equity Returns in the Future II

The Economist writes a confusing article because it fails to consistently measure asset returns in real magnitudes. It writes, as I have straightened things out:

Economist.com | Investing money: ...over the past 100 years American shares have outperformed bonds, property, art and gold, with an annual average total return of 9.7%, or 6.3% after inflation. Government bonds returned less than 5% [or 1.6% per year in real terms]. This performance underlies the common view that as long-term investments shares are hard to beat.... But a lot depends on how long the long term is. In a book in 2002, Triumph of the Optimists... Dimson... Marsh and... Staunton... found... in 13 of the 16 countries... shares did worse than cash in the bank in at least one 20-year period in the 20th century. Over ten-year periods, negative real returns on equities were not that uncommon. An investor buying American shares in 1964 and selling in 1974 would have made a real loss of 35%.

Another reason for caution is that periods of exceptionally high returns--such as the 1980s and 1990s--are usually followed by phases of exceptionally poor performance. The 20-year bull markets in shares (which lasted until 2000) and in bonds (which continued into 2004) were fuelled by an almost continuous fall in inflation and hence interest rates. But now that inflation is low, neither shares nor bonds are likely to deliver double-digit returns.... [Nominal] bond returns in most countries will be broadly in line with their current yield of less than 5% [and, with 2% inflation, annual real bond returns of 3% or so].

What about equities?... price-earnings (p/e) ratios still look a bit high, notably on American shares, and share valuations are unlikely to benefit from falling interest rates in future. Meanwhile, lower inflation means that the pace of profits growth will slow. Assume that America's nominal GDP grows by 5% a year (3% in real terms, plus 2% for inflation).... Suppose... that profits do rise in line with GDP and that p/e ratios stay the same. Then... total [real annual] return[s] on American shares over the next decade will average [4].8% [in real terms] ([3]% [real] profits growth, plus dividends)....

Investing in emerging stockmarkets could also pay off handsomely over the next decade. The average p/e ratio in emerging markets... is around ten.... [I]f governments can maintain their current, sounder economic policies, growth should be more stable in the years to come--and returns should be higher and less volatile. European shares could also outperform Wall Street. The old continent's economies are widely derided for their rigid markets, high taxes and lack of entrepreneurial vim. But financial markets have discounted all this, and European shares look cheap next to American ones....

The most important lesson for investors is that when [real] GDP is growing by only [3]% [per year], asset prices cannot on average be expected to rise much faster than this...

Could be. But I think that the Economist is too pessimistic. Let me sketch out what I think of expected American equity returns:

I tend to start from the equation:

r = E/P + (rs-r)(E-D)/E + rs(E*-E)/E

where r is the expected real rate of return, E/P is the earnings yield, rs is the rate of return the average company earns on its reinvested earnings (which Glenn Hubbard and company argue is significantly in excess of the market return becuase of various information and signaling problems), and E*-E is the gap between accounting earnings and the business's real Haig-Simons earnings (which Eric Brynnjolffson and company argue is significant because of all the trial-and-error investments in organizational form that are inaccurately counted as operating expenditures). This makes me think of the current earnings yield of 5% as a lower bound to expected equity returns, which I tend to see as 5.5-6%. And when I compare this to a real long Treasury bond return of 2-2.5%, and when I meditate on long-run inflation risk to which nominal bonds are uniquely vulnerable...

Well, the upshot is that the 3-4% annual expected equity premium return that I see still seems very large to me: to correspond to the preferences of a 62-year-old male expecting to spend his wealth in the next fifteen years or to the preferences of a money manager for whom reporting a big loss in the next four years is a career-limiting move, rather than to the risk preferences of the economy considered as a frictionless social welfare maximizing machine. And this means that there is still a powerful, powerful case for stocks for patient investors with a long horizon of a quarter century or more. If you can wait a quarter century, stocks do not look like a sure thing relative to Treasury bonds, but they do look like a 90% thing.

Posted by DeLong at January 4, 2005 01:57 PM

Trackback Pings

TrackBack URL for this entry:


This months Financial Analysts Journal published an article by Peter Berstein that had a table of real GDP and stock market returns over 20 year moving averages since 1870.

The data show that there is a strong positive correlation between real growth and stock returns that leads to the conclusion that if growth is as wek as the 1.7% the SS commission projects that it is extremely unlikely that the
stock market returns or profits they project can be achieved. The SS commission appears to be
internally inconsistent.

One final point. I have always loved Samuelsons comment that the US stock market since 1926 is a sample of one.

Posted by: spencer at January 4, 2005 02:27 PM

That is all ridiculous. Let's look at a typical case: Kodak. You hold onto that stock from 1950 to 1990. Looks fabulous.

2000, dead in the water.

2004 near bankruptcy.

Panam Airlines: invest in 1950 to 1980. Great.

1985. Bankrupt. Stock worth zero.

I could use thousands of such examples. Long investing is very dangerous, they use for their models only those that haven't fallen into bankruptcy which makes the investment 100% worthless.

Posted by: Elaine Supkis at January 4, 2005 02:42 PM

"If you can wait a quarter century..."

If a person has lifetime job security they can afford to wait 25 years. What about the typical
working person ?

Posted by: bhaim at January 4, 2005 02:52 PM

The return of the Vanguard S&P Index Fund since August 31, 1976 has been 12.38%. The return over the last 10 years has been 12.00%, even with the fierce bear market. A return of 7% over the coming generation should not be excessive and would be pleasing relative to the expected return from bonds.

Posted by: anne at January 4, 2005 03:06 PM

Graphs from the Economist article are here:

SS investment in equities would take place over a long period of time, so historical risk premia should apply. There is some chance that SS buying will bid up the price of shares proportionally more than it will improve the actual productivity of the economy, leading to a lower risk premium:

I suppose the portfolio should also include foreign holdings, to take advantage of rapid growth in emerging markets.

Posted by: steve at January 4, 2005 03:56 PM

I, too, tend to start from the equation

r = E/P + (rs-r)(E-D)/E + rs(E*-E)/E

but eventually the guy behind me starts honking and I just tell them I'll have a number 2 with a Diet Coke.

Posted by: alkali at January 4, 2005 04:37 PM

We should remember that equities are like any other scarce resource subject to supply and demand. If demand for shares increases, their prices will also increase, even if there is no change in the "intrinsic value" = sum of future dividend payments. Eventually the supply of shares can increase, as perhaps the rate of business formation speeds up. But, it seems obvious that the growth in capitalization of the broadest index of equities cannot exceed GDP growth for any length of time, so it would be surprising if this rate of value creation could accelerate drastically.

From this perspective, it seems that social security privatization is likely to bid up equity prices and depress their future returns. Imagine the following analogy: one day, foreign investors wake up and decide to increase their portfolio allocation to US equities. The result may be a buoyant stock market, but to what extent does this increase real value creation in our economy? Does it create enough value (i.e. future earnings and dividend growth) to justify the amount by which prices are bid up? (An even simpler analogy: I have a chicken, which produces eggs at a fixed rate. Demand for egg-laying chickens increases, driving up the price of my chicken. Will it lay eggs any faster as a result of its increased price?)


Posted by: steve at January 4, 2005 04:40 PM

All I can say is, I'm glad you're not managing anyone's money. That is really the most amazingly obfuscated, over-optimistic prediction I've seen in months.

When you buy a stock, you get the dividend, plus capital gains when you sell. If P/E ratios stay the same, the capital gains growth rate is equal to growth in earnings per share (EPS). Simple.

Plug in the numbers. Historically, the dividend payout ratio of the S&P500 has been about 55%. With a current earnings yield of about 5%, they COULD be paying out 2.5% today (they are actually paying out less, but you could argue that the remainder is going towards share buybacks, which also return money to shareholders via capital gains)

Historically, growth in EPS has been about 2% less than growth in GDP, due to share dilution.


Historically, real GDP growth has been about 3.5%.

So, assuming P/E multiples stay the same, we have:

r = 2.5% + 3.5% - 2%

r = 4%

This is what you would expect if P/E multiples stay at the currently elevated level of 21. If they return to their historical average of 16 over the next 10 years, you should expect 2.7% per year lower, for a real return of

r = 4% - 2.7% + ((2.5% + 3.1%)/2 - 2.5%)

r = 1.6%

A 20 year TIPS returning a real 2% looks pretty good right now.

Anne, your mistake is that you assume the P/E multiples will grow as much in the future as they have in the past. VERY unlikely.

Posted by: ErikR at January 4, 2005 04:41 PM

[another comment spam makes it through]

Posted by: at January 4, 2005 04:52 PM

I agree with EricR's comments above.

To make matters worse, the P/E of the market (the S&P 500 or whatever proxy you like) is understated by a good deal. First, options are not properly expensed. If they were, the P/E of the S&P 500 would be (i.e. actually is) 10% higher. Second, companies that provide pensions make unrealistic asssumptions about the future returns of their pension funds. If companies used reasonable estimates, the P/E of companies that have pensions would be (i.e. actually are) another 10% higher.

Posted by: Jeffrey Miller at January 4, 2005 06:00 PM

Wow. This was a little technical for me. If I understand correctly, the argument is that the stock market gives higher returns, but you're saying that's not necessarily true. That's a good argument.

I think the more compelling argument to me is the shifting of risk for retirement from the government (defined benefit) to you (defined contribution). I guess that's the "ownership" society for you. I call the gambler's society.

What the privatizers don't discuss is how many people fail to put away money, even with incentives. At my work, the company gives a 50% match up to 6% pre-tax savings. If you're not putting away 6%, you're actually giving up money. Still, many people, who are making a decent wage, fail to put away any money at all. They feel they can't afford it now. I tell them they won't even notice the difference (it's pre-tax, so it would lower their taxable income), but they still won't do it.

Posted by: Unstable Isotope at January 4, 2005 06:13 PM


"[S]ocial security privatization is likely to bid up equity prices and depress their future returns. Imagine the following analogy: one day, foreign investors wake up and decide to increase their portfolio allocation to US equities. The result may be a buoyant stock market, but to what extent does this increase real value creation in our economy?"

I agree completely, there is every reson to believe that investing payroll tax revenues in the American stock market will buoy prices in the short term only. An important argument.

Posted by: anne at January 4, 2005 06:50 PM

Jeremy Siegel and Rob Arnott debated future equity returns in the April 19, 2004 issue of Forbes.

Arnott is the Editor in Chief of the Financial Analysts Journal. His work is required reading for the Chartered Financial Analyst (CFA) program.

Here's a summary:

"[According to] Siegel, the long-term outlook isn't that great. He is forecasting only a 5% to 6% return from stocks after inflation over the next 25 years. He sees dividend payouts rising, adding to total returns. He also sees modest economic growth, and profits rising at the same rate as economic growth. Of course, Siegel does not forecast an increase in valuations as we had in the 1990s.

Arnott is much more bearish. In addition to arguing that earnings are overstated, he thinks economic growth will be fairly low in future years. He believes stock investors will be lucky to earn a 1% return after inflation over the next 10 to 20 years. He's an advisor to the PIMCO All Asset fund and is recommending investments such as emerging market stocks and bonds, REITs, timber, TIPS, and European convertible bonds."

Source: Bob Carlson's Retirement Watch

Posted by: bhaim at January 4, 2005 07:15 PM

Anne - a 12% return since 1976 corresponds to a real rate of return of about 7.7% after adjusting for inflation.

If you back out the roughly doubling of P/E ratio since 1976, that brings you down to a return closer to 5% since 1976. We've had an extra 2.7% annual return since 1976 just by expanding the P/E multiple.

If we get a 5% real expansion of profits over the next 29 years, like we did during the past 29 years; but instead of having the P/E expand, it retracts; then we'd have a 2.7% annual HEADWIND for the next 29 years. That would bring the real rate of return for the next 29 down - oh, I haven't done the math to confirm this one - but I think down to around 2.3%.

But the actuary assumptions call for slower growth of population and slower growth of productivity during the next 29 compared with the past 29. So, there's a pretty fair chance that we'll have less than a 5% annual real growth of profits.

Not to mention the chance for something like a flu pandemic, nuclear war, or various other externalities that could wreck your whole spreadsheet.

Equity is still a good place to put long term capital if you can accept some risk.

That's possible if you've got a guaranteed, inflation adjusted lifetime source of income to complement your personal retirement savings.

Posted by: Charlie at January 4, 2005 07:48 PM

Things I see missing...

These funds will be invested by the public, which has a very small amount of information about different firms. Not only will Fortune 50 and Blue Chip stocks get a market-insensitive amount of investment, but also fly-by-night firms will find millions of new pigeons.

Will it be legal to invest the accounts in overseas funds? In ADRs? What is our interest if billions are invested in country X and country X starts nationalizing?


As the (recently) late Rep Matsui of California said, the long term outlook for Social Security was worse in 1983.

Now, how many 50 (not to mention 75) year projections are you going to rely on? Sure, the mortality tables will be good, but what about Federal receipts and outlays?

Garbage In, Garbage Out.

This also applies to the White House and Congress.

Posted by: Josh Narins at January 4, 2005 08:01 PM

The mention of a 25 year window of course brings up another issue that the Administration hasn't addressed. Where is this cutoff for those "close" to retirement? How will their plans affect people with different time windows until retirement. I still have years before retirement, even with the proposed upward creep of the age to receive full benefits. But not 25 or more of them. So where does that leave me and others like me who are technically the trailing edge of the boomer generation?

Posted by: Jim S at January 4, 2005 09:03 PM

[another comment spam makes it through]

Posted by: at January 4, 2005 10:18 PM

[Be polite.]

Posted by: at January 4, 2005 10:33 PM

Jeffrey Miller, what a good point. The S&P is trading at a P/E of 20.57 today. Adding 10% gives a number over 22. A really undervalued market would trade at a P/E of 6 or 7. That would put the S&P at around 400, possibly below. Granted that calculation is imperfect because it doesn't account for earnings growth. But nevertheless, the downside risk for the equity market seems enormous right now. At the end of 2025, we may be lauding the guru that publishes "Rydex Inverse Short Funds for the Long Run".

Posted by: Hula Man at January 4, 2005 10:45 PM


I think that if you analyze the Vanguard S&P Index Fund return over the last 10 years not on the basis of someone having invested a lump sum in it 10 years ago, but rather on the more realistic basis of someone having been investing a fixed amount per month throughout that period, you will find the rate of return rather lower.

When the S&P500 was down around 800 a few years ago, I did an analysis on such a basis and found that to have even broken even, an investor would have had to have been investing thus since before mid-1975. Anyone who had begun investing a fixed amount per month at any time after that would have had a negative return.

As the S&P500 is still far above its long-term growth trend line, and its P/E is far above historically normal levels (even without adjusting for overstatement of earnings through understatement of pension obligations, etc.), it will not be surprising if it returns to and even undershoots its long-term trend, returning once again to 800 territory. If it does, then losses on the amounts invested during the bounce of the last two years will push the breakeven date back even further.

Those who think the price bounce of the last two years has been a true recovery should ponder the fact that the trajectory of US stock prices over the last ten years market has been nearly identical to that of the Japanese market in its bubble and post-bubble years.

And see also:

Posted by: jm at January 4, 2005 10:46 PM

Those who believe that we will continue to see the 7% return found by Dr. Siegel are ignoring the obvious----any observable truth about stock performance over the long term is no longer true once it is generally accepted.

This is because stocks are, for all intents and purposes, a finite resource, and the moment a "truth" is discovered, demand for stocks that take advantage of that "truth" increases, artificially raising the price of those stocks, and decreasing their yield. Its Capitalism 101 at work---increases in demand without an increase in supply results in higher prices.

The moment Siegel's findings (which have been bastardized to "you can't lose in the stock market long-term" in the public mind) became accepted, they stopped being true. The perception of the stock market as a place of "risk" was replaced by one of the stock market being a "sure thing." Much (if not most) of the rise of the stock market from 1994 onward can be attributed to Siegel's findings, as more and more people assumed that stocks were better than CDs as the place to put their IRA and 401K funds. (between 1990 and 2000, assets in 401K plans jumped from $385 billion to 1.83 trillion. And the Dow went from 2633 at the end of 1990 to 10656 as of November 16, 2000.)


Posted by: paul_lukasiak at January 4, 2005 11:59 PM

Any-one want to point me to a paper or give me the basic reasoning why extra investment won't tend to reduce the returns on other investments?

I suppose especially if the cost of the extra input is reduced won't the agregate return on all invesments approach the return expectations of the most recent?

What is to prevent the extra investment from utterly obliterating some exisiting investments streams through over investment in a commodity like field and why after that obliteration would the investments made obsolete factor at all...the new low cost plants would be built until they matched the lower cost of capital the now obsolete plants which otherwise would have had continuing life if interest rates weren't artificially lowered would be worthless if their operating costs couldn't match the lower cost plants plus the lower cost plants low loading with their low interest rates.

It seems to me that the bent to force, or entice more money into investments would help the economy through greater society efficiency and short term employment and cash turn-over. The benefit to the economic efficiency and short nearer term activity would come at an expense to investments in place and with returns only equal to the below real (after inflation and after efficiecy enhancment) inflation interest rates prevalent during constuction.

In the very basic, the idea, lets shift a huge portion of peoples money into equity investments seems to me a plan that will necessarily lower returns on investments.

I believe that the generally held belief is that more investment benefits an economy...however does more investment benefit investors?

By forcing more americans to be investors would we be making the retirments safer?

As it stands now social security holders are investors anyway, the investment being made in government bonds "held in the Lockbox ;) ".

I don't really buy the invesment notion but I do buy the notion that the deficit spending that is acrruing in invesment accounts of our nation and foreign banks, is working its way into investment here. That investment will hopefully lead to a growth in our economy that will permit tranfer payments to the elderly to appear a smaller portion of the total economy than they would be without economic growth owing to the extra investment.

But thats not to say the invesments themselves will pay off but the invesment in the economy will pay off.

Anyone want to help me with what I'm missing?

Posted by: Tom Norian at January 5, 2005 12:34 AM


Strikes me that and increase in supply can be even more detrimental.

If I own a office buiding and due to the fine performance of rents and occupancy I've had over 20 years the "P/e" goes up (or as is more prevalent in RE, the cap rate goes down), I could realize more from my investment, although its value would have changed very little to me if my intention were to hold it for another 30 years.

Some of that "increase in value" (change in market value?) is likely also attributable to falling interest rates in the last decade.

Where the notion of expanding P/e's or falling interest rates is not neutal to a long term holder is when those higher values, lower borrowing or access to capital costs stimulate new constuction.

Even if I had kept my rents at a level where building another building in my locality wouldn't make sense at my initial 10% above inflation premise at a time of 4% inflation, 4% real return and 2% extra risk premium, as the inflation rate real return and notions of risk premium dropped suddenly new investment might be stimulated.

However, the new office building might be built upon premises of demand increase that weren't there, or buy stealing a major tenant of mine induced into a long term lease.

The over capaicty if created, would decrease the value of both buildings per square foot, but would certainly undo mine more ...the long term hold would be less valuable to me even as the market value had increased due to P/E expansion.

I think that some of the market investment market cycle could be attributable to the idea that extra invemestment stimulated creates modestly over investment that if isn't quite hugely excess in capacity, is enough so that it will continually dissapoint the expecations initially held.

And of course if interest rates bump up a bit as a mild inflation eats at the value of the income streams that were sometimes contractually entered...well it wouldn't be a wonder to see P/E ratios narrow in a way that created a major 20 year headwind.

Even with a headwind of course could still beat holding treasuries, espeically in a taxable account.

Posted by: Tom Norian at January 5, 2005 12:57 AM

> The return of the Vanguard S&P Index Fund since
> August 31, 1976 has been 12.38%. The return over
> the last 10 years has been 12.00%, even with the
> fierce bear market. A return of 7% over the coming
> generation should not be excessive and would be
> pleasing relative to the expected return from
> bonds.

It is said that the average Japanese family has 2 years' salary in passbook savings. What would happen if every family in Japan woke up this morning and decided to drop out of the rat race for a year and live on their savings?

Similarly, the returns you quote are only possible if everyone does NOT invest in them. If 100% of the population starts investing in the Vanguard Total Stock Market Index, its rate of return must immediately fall to average productivity growth over the investment period.

Less fees and commissions of course. Supersize that Enron sir?


Posted by: Cranky Observer at January 5, 2005 05:22 AM

Given the volatility of equities, it is not safe to withdraw an amount equal to the expected average annual return. You'd be selling shares at depressed prices at some point, and would exhaust your portfolio faster than you might expect.

This volatility drag is probably on the order of 1.5%-2% a year. For example, if you expect your portfolio to return 5% a year, you shouldn't withdraw more than about 3% a year if you expect it to last for the reasonably long-term.

Posted by: richard at January 5, 2005 07:02 AM

Since 1996, the trajectory of US stock prices has been nearly identical to the trajectory of Japanese stock prices, with a lag of 10 years. Didier Sornette at UCLA has seen this as possible evidence that economies operating in a bubble regime behave the same as physical systems in analogous regimes.


In Q1 1995, the first of Japan's bear-market rallies ended with a 3-month decline which took the TOPIX down from 1600 to nearly 1200, followed by a 3-month bounce that ended at 1200. Though by year-end '95 it had recovered to 1600, that was just the second of the three bounces that would end with the TOPIX below 800 in 2003.


Hula Man:
On a business trip to Tokyo in the late '80s, I found that the President of the small export firm I was visiting (an overseas-Chinese resident there) was spending about 80% of his time closeted with stockbrokers. He said that he could make more money in stocks in a week than he could make in his real business in a year. At the hotel that morning, I had switched on the TV to find NHK (Japanese equivalent of UK BBC) broadcasting a talk-show with housewives listening to a stock analyst explain that "no one has ever lost money in the Japanese stock market if they've stayed invested for more than three years." A few days later, the central character of an evening drama was a female stock analyst living in a multimillion-dollar condo. The TOPIX peaked at 2,884.80 on Dec. 18, 1989.

Cranky Observer:
Most of that Japanese familiy's passbook savings will be in accounts at the Japanese Postal Service, which will have "invested" the funds in Japanese treasury bonds and the bonds of various governmental and semi-governmental agencies and authorities, financing the numerous money-losing expressways, bridges, tunnels, dams and airports that speckle the landscape, and the US-Treasury-buying that has kept the yen down to subsidize exports. As these "investments" have sub-zero returns, when the Japanese citizenry attempts to redeem those savings in coming years, the government will have to sell more bonds to bail out the postal savings system, and Japan will face the same kind of general-fund crisis we're talking about for Social Security. "Supersize that Enron sir?" Onegai shimasu.

Posted by: jm at January 5, 2005 07:06 AM

Again, compared to what? The guaranteed negative returns to young people being taxed 12.4% of payroll, say.

And this:

"stocks are, for all intents and purposes, a finite resource"

is risible. Existing companies issue new shares, and new companies come along all the time. Microsoft, Amazon, ebay....

World-wide there is something like $45 trillion of publicly traded investment. Turning SS into a true investment program, isn't going to have much effect on those returns.

Posted by: Patrick R. Sullivan at January 5, 2005 07:20 AM

Thoughtful and useful comments. Thank you.

When Vanguard began its S&P Index Fund on 8/31/76, the price earning ratio was below 10. The p/e ratio did not climb above 10 till 1983. At the market peak on 8/31/87, the p/e ratio was above 17. The S&P began the 1990s at a p/e above 15 and finished at a p/e above 35 if this number can be trusted. Now, we are at 20.

The Vanguard S&P Index Fund benefited by the p/e ratio going from below 10 to 20. Also, until recently dividends were far higher than they are today and I trust stock buybacks less than dividends in adding to stock market returns.

So my long term sense of the market is rather than look for a 12.4% return as the Vanguard Fund records from inception, a 7% return seems conservatively reasonable.

Posted by: anne at January 5, 2005 07:29 AM

What is discouraging is to hear discussions of plans for Social Security reform on PBS or NPR that take the premise for granted that Social Security must be changed and must be changed now. Conservatives are insistent and moderates hedge, so the idea that there is a Social Security crisis grows apace. Discussion of changing the indexing of benefits from following the wage index to the price index, gives no sense of how benefits will be limited by this subtle move.

Posted by: anne at January 5, 2005 08:00 AM

What about the capital-to-output ratio?

If we get the total return equal to GDP growth rate plus 1.8% dividends, and reinvest all those dividends, and get the 4.8% real return from the article, how big is the stock market going to be?

Right now, total market value of Wilshire 5000 is some 13.1 trillion, or 1.11 times the GDP. If it grows 1.8% faster than GDP, it will reach 3 times the GDP in 2061, and 5 times the GDP in 2090. But most macroeconomists estimate the total K/Y at somewhere around 3. And total K includes much more than stocks.

So which of the following is true:

(a) The amount of capital in the present economy is grossly below steady state (which also implies that growth will slow down as we approach the steady state).

(b) Most of the dividends cannot be reinvested in the stock market, so the total returns will be much lower.

(c) The stock market will really grow that fast, and we are close to a steady state. We will close the gap by selling loads of bonds to foreigners. This means the stock market will be much more levered than today, i.e., the 4.8% return will be bought by higher risk.

(d) I am missing something humongous and obvious to everyone else.

Posted by: enfant terrible at January 5, 2005 08:22 AM


I really needed that post this morning; I'm still laughing.

Posted by: matt at January 5, 2005 08:31 AM

Sorry if my point wasn't clear. I wasn't talking about returns on Japanese passbook savings, interesting though that subject may be. I was trying to find a realistic example to show why an action that can reasonably be taken by some members of society cannot (necessarily) be taken simultaneously by ALL members of society. In my example, one overworked Japanese salayman can say "to hell with the future" and take a year off. If ALL Japanese workers did that on the same day, the economy would collapse the next morning. And I mean really collapse, as in post-apocalypse.

Same thing with private investments in Social Security. A few million can do that, sure. But we cannot ALL do that because over the long run (say 10 years)
ProductivityFunction(Input) = Output
per the laws of thermodynamics and conservation of energy. Regardless of how sophisticated one's investment strategy an entire society cannot make something from nothing.


Posted by: Cranky Observer at January 5, 2005 08:52 AM

JM, thank you for your colorful anecdotes on Japan. When looking at parallels, what seems most horrific to me is the bubble shift, from stock bubble to real estate bubble, that occurred in Japan seems to be happening in the US now. It appears that the our citizenry has a tendency to embark on naive quests for mythical "can't-miss" asset classes. Even after having their stock market faith broken in the nineties, investors have become quick to repeat old mistakes by again squarely placing real estate in the "can't miss" asset class category. If the zeal and concomitant hangover of Japanese investors is in fact a parallel for our country - and that is a big if - then the Japanese example shows that a belief in the long term riskless returns of real estate is a dangerous activity.

There is a very good chance that real estate investors, sucked in by low interest rates, will get their backs broken over the next decade and a half with declining housing prices and negative equity, just like the Japanese.

If we do follow Japan's path, what will happen to the American business and the stock market as equity is destroyed? For most Americans, housing values represent the greatest percentage of their net worth. If housing values recede for a long time, people may feel poorer, and there is a good chance that spending will slow. And with that, social security privatization or no social security privatization, the stock market could get whomped. The prospects are grim enough to posit that even 5% *nominal* annual returns are highly unlikely over the next two decades. In 20 years, people may find themselves elated with 0% returns, breathing a sigh of relief that they simply did not lose money.

Posted by: Hula Man at January 5, 2005 09:16 AM

The situation is Japan is far more optimistic than presented. Travel through Japan, especially with Japanese friends, and you will find that people live more comfortably. Growth slowed in Japan after 1993, but deflation coupled with fine Japanese saving levels and high employment at good wage levels have kept Japanese households quite comfortable. My Japanese friends repeatedly comment on our faulty perception of Japanese well-being.

Posted by: anne at January 5, 2005 09:31 AM

Anne, I don't see your point. The point of citing the Japanese argument is to highlight that a similarly long decline in stock market (and housing) equity could happen here and mull over the consequences - for Americans. The Japanese are a far less economically stratified nation than the US, and, as you point out, Japan has liquid household savings. Unfortunately, we have very little household savings, and a range of classes. As for our low and lower middle classes, the potential for an extended period of future pain looms large.

Posted by: Hula Man at January 5, 2005 10:03 AM

In the 19th century, after inflation, U.S. stocks returned 1.5% more per year than bonds.


The essential argument on the outlook for long-term stock returns is between those who think:

1). Stock prices are a function of supply and demand, stock markets are relatively efficient even in the short term, and the 1926-2004 period for which we have the best data on financial asset returns is long enough to give us a good reference frame to project returns well into the future. (This is the consensus view).

2). Stock prices are a function of supply and demand, but stock markets can be radically mispriced for long periods of time. Historical stock returns can be seen, eventually, to derive from fundamentals: yield, p/e contraction / expansion, and earnings growth (which is more or less a constant). In the short run the market is a voting machine, in the long run it is a weighing machine. This much smaller group includes Peter Arnott, Jeremy Grantham, Bill Bernstein, Warren Buffett, ErikR (see his excellent post above), and myself.

If you fall into camp 1, some degree of SS privatization is a good idea. Stocks are bound to go up significantly more than bonds. If you fall into camp 2, if SS privatization had been implemented in 1974 or 1942, fabulous. A clear winner. Doing it right now is nuts, a decision based on pure wishful fantasy (making it a typical Bush administration big project). Quite likely to be not just worse than the current system but disastrous, particularly if benefits are indexed to CPI instead of wages and only U.S. stocks used as an investment option--which I strongly suspect will be the case.

If you aren't familiar with recent issues of the -Financial Analysts' Journal- (if not, you are more blessed than you know) ErikR briefly explains why above. I'll also wager that Bill Bernstein will tackle the issue in his next -Efficient Frontier- newsletter.


Posted by: Nicholas Mycroft at January 5, 2005 10:03 AM

"is risible. Existing companies issue new shares, and new companies come along all the time. Microsoft, Amazon, ebay...."

ten years ago, NYSE composite index was slightly over 2500. Today, its slightly over 7000. ten years ago, the Wilshire 5000 was at about 4500. Today, its at 11,700. Precisely what percentage of these increases was the result of new issues?
And what percentage is the result of greater demand for stocks driving up prices?

when an existing company issues new shares, it dilutes the value of existing shares.

and for every "Microsoft" or "Amazon", there are lots of Pan Ams, Enrons, etc....

Posted by: paul_lukasiak at January 5, 2005 10:11 AM

Hula Man

"The point of citing the Japanese argument is to highlight that a similarly long decline in stock market (and housing) equity could happen here and mull over the consequences - for Americans."

The intent was just to make clear that the Japanese are faring surprisingly well for all the stock price decline since 1990 and property price decline since 1993-1994. Paul Krugman has argued similarly. The Japanese are less dependent on stock market and property price levels than Americans are. There may be severe bear markets to come for us, but I can not imagine a stock market or property price decline here of the depth and length of the decline in Japan. We would use policy tools far more quickly to protect these markets.

Posted by: anne at January 5, 2005 10:46 AM

posted by im
When the S&P500 was down around 800 a few years ago, I did an analysis on such a basis and found that to have even broken even, an investor would have had to have been investing thus since before mid-1975. Anyone who had begun investing a fixed amount per month at any time after that would have had a negative return.
With the S+P fixed amount purchases and reinvested dividends, the investor would have done considerably better than short term treasury bills between 1975 and 'when the S+P was down around 800 a few years ago'.
This is not to say I disagree with the general thrust of your argument, just that you made a mistake in your model or that you disregarded reinvested dividends.
In the 27 years of 1975 to 2002 the S+P only had six down years. They were 1977, 1981, 1990, 2000, 2001, and 2002. The S+P was up in 2003 and 2004.
I do believe the stock market is overvalued. Of course, I believe that real estate and dollars are even more overvalued.
Here is a URL for the S+P returns. http://www.mutualofamerica.com/articles/CapMan/October03/SandP500.htm

Posted by: wkwillis at January 5, 2005 11:04 AM


"When the S&P500 was down around 800 a few years ago, I did an analysis on such a basis and found that to have even broken even, an investor would have had to have been investing thus since before mid-1975. Anyone who had begun investing a fixed amount per month at any time after that would have had a negative return."

Take a look at Barra.com or MSCI. The depth of the bear market in 2002 would have taken returns back to 1995, and this only for a short short while. The S&P Index has been a wonderful investment since Vanguard began the fund in 1976.

Posted by: anne at January 5, 2005 11:50 AM



Posted by: anne at January 5, 2005 12:18 PM

It seems to me that this discussion of aggregate returns misses the point.

I don't have any data, just a "gut feeling", but I'd bet that if you looked at the lifetime records of all the individuals who have invested their own money in the stock market, you'd find a small number of spectacular winners, a slightly larger number who have had the discipline to do better than inflation, and a majority who didn't even break even, even with employer matching.

If the point of Social Security is mitigating risk, this looks very bad to me. If the goal of the government is to shift risk away from corporations to individuals, then it's an excellent plan.

Posted by: Ted at January 5, 2005 12:18 PM

John Bogle has done a number of studies looking at the actual returns of mutual fund investors through the bull market years from 1982 on. The flow of money to and from funds shows the returns in fact to be most discouraging. The problem has been discipline discipline discipline, both by the investors and by active managers of funds. Of course, costs of investment are a problem as well. Investors in stock funds have fallen far far short of the S&P returns for the period.

Posted by: anne at January 5, 2005 12:48 PM

I think Japan as a beacon of what to expect in a post-bubble new demography age shows different things depending on what you care about.
The good news (for Japan) is that the country weathered the transition quite well. I am not nearly as sanguine about how the US will handle this.
However the country doing well says little specifically about the return on financial assets, and especially the stock market. The point is that, to a remarkable extent, these returns have been decoupled from the overall welfare of the population; and if as an outsider (and a conplete mercenary with zero interest in politics) ALL YOU CARE ABOUT is the stock market returns, you're presumably disappointed in your investment in Japan 10 yrs ago.

Now, what does this mean for America? I'd argue that neither viewpoint is very pleasant. If you live in some force-field protected castle somewhere and care only about US stock market returns, well lagging Japan by 10 yrs does not provide an inspiring sight. And if you care about the social welfare of the US (which, given the US' legendary propensity to blame its problems on others and then attack them, means the welfare of the whole world) you should be even more scared. Both the financial structure and the social cohesion of the US are far more rickety than those of Japan going into the 1990s.

Posted by: Maynard Handley at January 5, 2005 01:43 PM

Maynard Handley

Interesting comments.

Posted by: anne at January 5, 2005 01:50 PM

Let's ask a general question: can the growth of the stock market as a long run trend continue forever? Has not the growth of both capitalist economies and the fractions of them represented in large public companies been premised on ever expanding populations and resource use? Yes, there have also been gains in efficiency, but let's not pretend that economic growth has been solely driven by improvements in efficiency of resource use. If growth is finite, then is a general upward trend in the past a good basis for supposing a general upward trend in the future, much less one of the same magnitude. We may be on one leg of a bell curve assuming the other has the same slope. Of course, if growth itself runs into problems, it is problematic for SS under either system, but I want to question this general assumption.

Posted by: Martin Bento at January 5, 2005 02:41 PM

Anne, wkwillis, others:

The "mid-1975" in my January 4 10:46 PM post was a gross typo. It should have been "mid-1995". Anyone who began buying an S&P 500 fund in mid-1975 would have so much profit from the first 20 years that losses from purchases in the bubble years would hardly matter to them. It is only since 1995 that the S&P 500 has become seriously overpriced.

But if the US market continues to track the Japan market with a 10-year lag, mechanically putting a fixed amount every month into an index fund -- a great strategy for the 20 years beginning 1975 -- will not be a good strategy at all for the 20 years beginning 1995.

Posted by: jm at January 5, 2005 10:36 PM

Tokyo Stock Exchange TOPIX index, 1968 to Present

Trajectory comparison:

Posted by: jm at January 5, 2005 10:44 PM

anne, jm, brad,

jm says, "mechanically putting a fixed amount every month into an index fund -- a great strategy for the 20 years beginning 1975 -- will not be a good strategy at all for the 20 years beginning 1995."

Although I agree that blindly putting money in the S&P 500 might not be such a good idea today, with some folks talking about depresssion and others talking about stagflation, I would like to make one point here:

I pulled Marty Zweig's 'Winning on Wall Street' off my shelves and turned to a chart (p. 38 in my 1997 edition) that tracks a "deflated" Dow..

Zweig shows that the period from roughly 1965 to 1982 was a big looser all the zig-zag way down for 17 years. Sounds like a secular bear market to me. A person investing in that index during much of that period wouldn't have seen daylight in their investments until the latter part of the 1990s. And only would have made money then if they were lucky enough to sell out during what some now believe to be yet-another top similar to the tops of 1965,1929,.... Assuming that what we are experiencing is the beginning phase of a secular bear market. I suspect that the S&P doesn't fare much better.

What (and why?) are your scenarios going forward? And: aren't long-wave trends important things for investors to be concerned about?

Posted by: Dave Iverson at January 6, 2005 08:54 AM

Thanks! We must continue this discussion. Suffice it to say, that beginning at the valuation levels we find now there is need for considerable caution. Remember, as well, we do not have the benefit now of dividends running at 4% to cushion the market.

Posted by: anne at January 6, 2005 11:28 AM

Since there is no current thread running on trade issues, but some here (is anyone else still here?) sound like they might find this interesting:

I post this because of my vexation with numerous commentators who hold that Chinese wages are so low that no reform of the yuan/dollar exchange rate could possibly affect our trade balances.

Posted by: jm at January 6, 2005 10:13 PM

Although Japan's economy is not doing nearly as badly as some commentators have made out, neither is it doing particularly well from the average citizen's viewpoint.

Unemployment is now up to 9.8% for the 20-24 age bracket, 7% for the 25-29 age bracket, and still rising (twice the rates in 1992). The large percentage of the employed in this age cohort who can find only part-time work is regularly viewed with alarm. Note that this is in a country where the average IQ is significantly higher than in the US, and the education system has much higher standards for performance in many areas. Note also that there are no restrictions on age discrimination in hiring, and most job ads specify a maximum age of 35, so if you haven't landed a full-time salaried position by 35, well...

One symptom of the general economic distress is the low (and still-sinking) birth rate. While average age cohort size in the 20-39 age bracket is about 1.75 million, in the under-10 bracket it is only 1.17 million.

The above age cohort stats come from an op-ed in last week's issue of Nikkei Business, in which the Chairman of Itochu (a major corporation) comments on the incoherency of the current government's so-called reform initiatives. One of those is privatization of the Japanese Postal Service, which operates a savings account system that holds a large fraction of the common people's savings (350 trillion yen, i.e., about $3.5 trillion), invested mainly in bonds of the government and various semi-governmental agencies and authorities; he asks the question, "What's going to happen if after privatization they decide not to keep on buying government debt?" and points out that the result would be a crisis in national finance, and that the semi-governmental agencies and authorities in particular would be cut off from financing.

Among the policy initiatives being pushed is an increase in the "consumption tax" -- basically a national sales tax -- to 11.3% (industrialists are pushing for something in the 18% range). Other adjustments are already biting:


Homelessness, while not as large a problem as in the US, is not insignificant.

Consider that these problems are surfacing in a country that has been engaged in hyper-mercantilistic exchange rate manipulation to protect it's manufacturing industries (to the tune of about $800 billion cumulative to date, on a GDP and population base half ours). What happens if (when?) they can't keep it up?

Posted by: jm at January 6, 2005 11:38 PM

Post a comment

Remember Me?