Ed Andrews Writes About Asset Returns and Economic Growth

Ed Andrews of the New York Times writes about "Asset Returns and Economic Growth":

The New York Times > Washington > Social Security, Growth and Stock Returns: In barnstorming the country over Social Security, administration officials predict that American economic growth will slow to an anemic rate of 1.9 percent as baby boomers reach retirement. Yet as they extol the rewards of letting people invest some of their payroll taxes in personal retirement accounts, President Bush and his allies assume that stock returns will be almost as high as ever, about 6.5 percent a year after inflation.

'For the life of me, I can't imagine why anybody would argue against young workers having the ability to invest and build a better retirement for their future,' Treasury Secretary John W. Snow said Wednesday in a speech in Bozeman, Mont.

A growing number of economists, however, including many who favor personal accounts, say Mr. Bush's assumptions are optimistic. Many believe that stock returns will be lower than they have been in the past, closer to 5 percent than 6.5 percent, and that returns on a balanced mix of stocks and bonds will be much lower than that.... The statistical battle is politically important. If investment returns are just one percentage point lower each year than predicted, a person would end up with 35 percent less money than she expected after 30 years of saving. Under Mr. Bush's plan, moreover, people would need to earn at least 3 percent a year after inflation just to make up for automatic cuts in traditional Social Security benefits.

In a paper to be presented on Thursday at the Brookings Institution, three economists who are longtime critics of Mr. Bush argue that stock returns are likely to be about 4.5 percent if economic growth slows as much as the administration predicts. 'We find it arithmetically very difficult to construct scenarios in which asset returns are at their historic average values and real G.D.P. growth is markedly slowed,' wrote the economists, Paul Krugman of Princeton University, whose Op-Ed columns in The New York Times have long been sharply critical of Mr. Bush's plan; J. Bradford DeLong of the University of California, Berkeley; and Dean Baker of the Center for Economic Policy and Research, a liberal research organization in Washington.

To make the numbers work, the economists contended in their paper, domestic profits would have to grow far more rapidly than they have in the past, or American companies would have to become huge exporters of capital to faster-growing countries. At the moment, the United States is a huge net importer of foreign capital.... [M]any Wall Street analysts warn that stock returns are likely to be significantly lower in the future for a separate reason: stock valuations are high relative to expected earnings, and they are likely to remain that way. The S.&P. 500 index is currently valued at about 20 times earnings, which translates to an expected return of about 5 percent a year....

William C. Dudley, chief United States economist at Goldman Sachs, estimates that stock returns are likely to be about 5 percent in the future, because investors are accepting lower 'risk premiums.'... 'My view is that stocks really can't deliver the same returns in the future as in the past, unless we have a major decline in stock prices,' said John Y. Campbell, a professor of economics at Harvard University and an adviser to the Social Security trustees on the issue in 2001.... Two recent computer simulations, one by Robert J. Shiller at Yale University and one by the Congressional Budget Office, suggest that even historical stock returns are no guarantee against losing money....

Stephen Goss, chief actuary for the Social Security program, defended the administration's assumptions. 'Keep in mind that we are trying to make projections over a very long time, 75 years,' Mr. Goss said. 'I would suggest that 5 percent at the moment makes perfect sense. But if you buy at another time, when the price-earnings ratio is 10, you would expect a higher return over time.'...

White House officials may be revising their assumptions. N. Gregory Mankiw, who recently stepped down as chairman of the Council of Economic Advisers, said Mr. Bush's proposed break-even rate of 3 percent on personal accounts may be too high. The yield on inflation-protected Treasury bonds is about 2 percent.

White House officials say they are open to proposals for changing the break-even point, which would raise the plan's cost, but Democratic lawmakers remain fundamentally opposed to Mr. Bush's plans.

'The basic arguments are over the extent to which people ought to be given more freedom over their risk and return choices,' Mr. Mankiw said. 'Returns on the stock market may affect the choice people make, but the question of whether they should be given a choice is broader than the issue of returns.'

Four points:

1. It's not clear to me that Dudley-Campbell is anything other than Baker-DeLong-Krugman seen from the other side. High price-earnings and price-dividend multiplies coupled with slow GDP and profit growth imply low returns. Only if GDP and profit growth is rapid can current stock market levels be consistent with high returns.
2. The issue separating Baker-DeLong-Krugman from Goss (and others, like MIT's Peter Diamond) is whether the stock market knows what it is doing. As we see it, there are three possibilities. We see all three of these as live. Goss sees only the third:
   1. The stock market knows what it is doing, and it is expecting low long-term returns (because of a fall in the equity premium and an anticipated fall in the rate of profit.
   2. The stock market knows what it is doing, and it is expecting normal long-term returns because it is anticipating relatively rapid long-term GDP and profits growth.
   3. The stock market doesn't know what it is doing, is currently overvalued, and will decline over the next decade to levels that are consistent with historical return patterns.
3. A Bush plan that had a 2% real clawback rate (or set the clawback rate equal to the Treasury borrowing rate) would be vastly preferable to the current Bush plan: it would accomplish the important task of providing a vehicle for the poorer half of America's population to easily invest in equities on reasonable terms.
4. Mankiw's invocation of the language of consumer choice and consumer sovereignty makes me very, very uneasy, for reasons I don't have time to set out here...

Comments

[Robert Waldmann]

Brad

If the current plan costs the SSA $ 1.1 trillion in present value, what would a plan with a 2% real clawback do ?

Also the real marginal cost of borrowing is not the interest rate, since the US treasury is not exactly a price taker. A plan which reduced national savings (as any voluntary personal account plan should as I have argued a few times) would increase borrowing costs by driving up the interest rate as well as by increasing debt.

[Joe]

"A Bush plan that had a 2% real clawback rate would be vastly preferable to the current Bush plan."

And a plan that had a clawback of 1% or 0% would be even better. But then the concept of "exactly zero cost for transition" might be a little tough to sell.

[anne]

http://www.nytimes.com/2005/03/31/politics/31social.html?pagewanted=all&position=

Two recent computer simulations, one by Robert J. Shiller at Yale University and one by the Congressional Budget Office, suggest that even historical stock returns are no guarantee against losing money.

Under Mr. Bush's plan, workers would be allowed to divert up to 4 percent of their payroll taxes to personal retirement accounts. But people would have to earn at least 3 percent a year after inflation to break even, because their traditional benefits would be reduced by the amount of their contributions, plus 3 percent a year in interest.

To protect people from stock market volatility, the government would also offer a "life cycle" investment account that would gradually reduce the share of stocks relative to conservative Treasury bonds as they neared retirement.

Mr. Shiller, using financial data going back to 1871, found that people who enrolled in life-cycle accounts would have lost money 32 percent of the time. The median annual return was 3.4 percent, barely above the break-even point in Mr. Bush's plan.

The results highlighted the inherent trade-off: higher returns come with higher risks....

[spencer]

If you look at the long term history of the stock market PE it swings in roughly 10-20 year cycles from well below 5 back to over 20. For example it was 26 in 1894; 6 in 1917; 20.2 in 1929; 9.4 in 1932; 25.5 in 1934;7.9 in 1942; 22.1 in 1946, 5.8 in 1946; 22.5 in 1961, 7.1 in 1980; and 20 in 1999.

Moreover, there is no central tendency. You often hear Wall Sreet srtategist talk about the average PE of the S&P being 14.4 as if there was a central tendency for the market to trend towards that average. But if you do a frequency distribution over the last 100 plus years there is about an equal probability that the pe
will be at any number between 9 and 19.

At the current PE of about 17 on trailing operating earnings the Pe is not discounting any particular scenario. Rather, it is more likely in the middle of another historic long term swings from over 20 back to under 9.

[spencer]

correction -- it shout be 30 in 1999.

[Ken Houghton]

Which means, for those not following the math associated with Spencer's post, that either:

1) the market should be around 5,500 when it finishes the swing OR

2) EPS will have to increase just under 90% before we rise above ca. 11,000.

I'm much fonder of the idea that the market has adjusted than either of those ideas (options 1 or 2 above), but the risk that (3) is accurate does rather screw the pooch on basing retirement plans solely on an investment in QQQ or DJI or SNP, as opposed to cherry-picking.

Cherry-picking, of course, is riskier than a market basket.

[wishful]

Hope it is okay to comment on this technical stuff, being a non-economist. Please don't lose sight of the fact that comparing rates of return on an insurance premium (SS) to rates of return on an investment is comparing apples to cats. That is not to say that the comparison is not an important one to make. However, I believe that this context should always be noted. Otherwise, regular people get the wrong impression. Thank you.

[anne]

Of the 3 stock market possibilities, normal returns would seem most difficult to expect because current market prices correspond to earnings leves that match and may best the record level of 2000. Unless we anticipate a continued gain in share of GDP for earnings at the expense of labor, we should not expect normal stock market returns.

We could well expect either lower than normal returns for quite a while or a sharp market decline which will allow stock purchases at far better values and a subsequently better projection of returns.

[anne]

Wishful

Social Security is indeed a social insurance program, and I do care that it remain so. Turning Social Security to a personal stock investment program is certainly going to leave a fair portion of retirees with lower returns than under the present system. Even if historical returns are matched, a portion of retirees will have lower benefits. The more difficult it is to match historical returns the larger portion of people will be harmed by lower benefits.

Essentially in Sweden retirees are guaranteed returns as large as the public pension system would have paid without private plans. There is a completely different program than we may imagine.

[anne]

Robert Shiller estimates that if historical return patterns persist about 34% of retirees will lose money will private Social Security accounts. So much then for social insurance.

[anne]

http://www.irrationalexuberance.com/index.htm

Robert Shiller:

· I have a new, March 2005, study "The Life-Cycle Personal Accounts Proposal for Social Security: An Evaluation" that uses historical data to assess the returns to investments in the life-cycle personal accounts. These personal accounts would invest heavily in the stock market for young workers, and gradually reduce exposure to stocks as the worker ages.

· One can access an Excel file with the data set (used and described in the book) on stock prices, earnings, dividends and interest rates since 1871, updated.

[wkwillis]

When the dollar renormalizes interest rates will go to ten percent (after inflation) and stock PEs will drop to match interest rates. You can figure that the Republicans will not be offering this plan when that happens because buying stock at low PEs would too obviously be a good deal for the workers.