Common Sense on Social Security
A Centrist Strategy for Social Security Reform
Seven Percent Returns Realistic?
Some of the linkages that connect GDP growth, market capitalization growth, and stock portfolio growth are not well known. Yet these hidden linkages play a vital role in shaping long run return rates for index fund investors. Real returns of 7% a year may have been possible over the past seven decades, for those investors who put their money into S&P; 500 index funds. But the driving logic that delivered 7% returns in the past is more likely to deliver 5% real returns in the future.
How to Make Sense Out of the Nonsense
Do you ever read the news and wonder if the stories make sense?
From time to time, the news media carry reports about GDP growth, announcing that GDP (Gross Domestic Product) is growing at three or four percent a year (adjusted for inflation). Equally often, the media carry stories about stock market returns. Anyone who invests in an index fund can earn compounded returns of ten percent a year, the stories declare. Now and again, the news media also run articles about work force wages. Real wages have grown, some economist will state, at only one percent a year for the past quarter century.
These stories are told and retold with measured solemnity, endowing them with an aura of Truth. The world of wage growth is a one percent world. The world of GDP growth is a three percent world. And the world of compound stock returns is a ten percent world. The oracles have spoken
The innocent reader wonders at all this. Aren't these separate worlds somehow connected to one another? If they are, how it is possible for them to co-exist in the same economy? Wage growth at one percent? GDP growth at three or four percent? Stock returns at ten percent? Isn't something out of whack?
And now, to remind us of our uncertainty, a major debate on Social Security is unfolding. In this debate, the world of ten percent returns on stocks is getting a great deal more press. Social Security specialists at the Cato Institute, the American Enterprise Institute, and the Heritage Foundation repeatedly hammer away at the same argument. If you've been investing in an S&P; 500 index fund every year, for the past several decades, it is asserted, your investments have been compounding at ten percent a year.
Even when inflation is taken out, they note, your money has still been compounding at seven percent a year. This has become the magic seven percent number that everyone cites. Put Social Security money into the stock market, somehow, and voil - seven percent returns! Compound growth is so powerful, it is asserted, that it can easily be Social Security's salvation.
Even some liberals think it's time to catch the wave. If the stock market's delivered seven percent returns for the past seven decades, they say, why can't it deliver seven percent returns for the next seven decades as well? Seven, seven, seven, seven. Why not allow the Social Security system to take advantage of the stock market's wealth-creating potential?
The average citizen isn't so sure. Seven percent stock returns? With an economy that only grows at three percent? With wages that grow at only one percent? Does this picture really make sense?
In general terms, we'll find, it is possible to make sense out of these seemingly disparate growth rates. Wage growth, GDP growth, and stock portfolio growth are all linked together, in a common framework, with each piece logically connected to the next. It's not a magician's illusion. There really is an underlying validity to the numbers we keep reading about in the news.
At the same time, there are some additional pieces to the story, pieces you don't hear about in the news, linking pieces that play a vital role in shaping the overall story.
Once these hidden pieces are brought into the open, the nature of the story changes. Conservative claims, it turns out, are full of holes. Yes, compound growth rates of seven percent have indeed been possible in the past. But no, seven percent investment growth rates are not nearly so likely in the future.
This essay has two basic objectives. First, it aims to explain the links that connect stock portfolio growth to GDP growth and connect GDP growth to wage growth. Second, it explains why it would be imprudent for Social Security reformers to base their planning on a forecast of seven percent stock market returns in the future.
The place to begin is at the beginning, by examining the key linkages that tie everything together. Start at the beginning, with GDP. Take a careful look at the boxes. Work your way across. Note the connections that link the separate pieces together.
As you do this, imagine that you've invested your Personal Retirement Account in an index fund, a special type of mutual fund whose stock portfolio exactly mirrors the stocks that are tracked by the Standard & Poor's 500 Index. Imagine that all the dividends you earn from this fund are reinvested, buying more of the same. Think about the returns you should expect. And think about why. What kind of returns would you be getting if you'd invested in an S&P; index fund twenty, thirty, or even forty years ago? What kind of return should you expect over the next forty years if you should start investing tomorrow?
In studying the linkages, you'll notice that Total Stock Market Growth is linked to GDP Growth. [Link 1]
And, as you look to the right, [Link 3] you'll see that S&P; Index Growth is linked to Total Stock Market Growth.
Not surprisingly, [Link 4] the value of an actual stock portfolio is linked, in turn, to the value of the S&P; Index.
Now pause. Look again at the box called Total Stock Market Growth. Just below, [Link 2] you'll see another box, Market Capitalization-to-GDP Ratio, which represents the relationship between total GDP and total stock market capitalization. [Note: If stock price refers to the value of a single stock, market capitalization refers to the value of all stocks collectively.] There is a volatile but unbreakable relationship between these two magnitudes.
Now go back to the first box, GDP, and follow the arrow that goes down and to the right, [Link 5] to the box labeled Dividend Payouts. It also turns out that there's quite a consistent link between total corporate dividends paid and the total size of the U.S. economy.
Link 6 originates in two boxes and leads to a third. Recall that the dividend yield on a single share of stock consists of the dividends paid to the shareholder divided by the value of the stock. The same concept also works for the economy as a whole. In the aggregate, the total Dividend Yield Rate consists of total corporate dividends paid, divided by the total market capitalization of the entire stock market. And that's where Link 6 takes us - to the nation's overall Dividend Yield Rate.
Finally, we come to the last box, Stock Portfolio Total Returns. This box represents the average return that the nation's investors are likely to receive on their index fund investments. As shown by Link 7, Stock Portfolio Total Returns are determined by two factors - Portfolio Price Growth, and the Dividend Yield Rate. As your stocks appreciate in price, your portfolio increases in value. As you earn dividends and reinvest them, your portfolio increases in value even more.
That's the overview. Portfolio returns are determined by price appreciation rates and by dividend yield rates. Price appreciation rates are determined, ultimately, by the GDP growth rate. Dividend yield rates are determined, ultimately, by market capitalization levels, relative to GDP.
Now we'll follow the same path again, in a bit more detail. Some of what you'll read may surprise you. There are indeed some hidden elements in the story, some clues that you probably haven't known about until now.
Link 1: GDP Growth and Total Stock Market Growth
There is a permanent relationship between the total size of the stock market and the total size of the GDP. This may seem hard to believe, with stocks soaring while the GDP remains earthbound. But it's true. Logic explains the relationship. The data confirm it.
First the logic. Even when stock prices soar, they're still related to something tangible. Corporate revenues. Corporate earnings. Corporate dividends. Add these same elements together, for all U.S. corporations, and you're very close to the nation's GDP. What is GDP, at its heart, after all, but the sum of the revenues of all U.S. businesses?
As business revenues grow, business earnings grow. Business dividends grow. And GDP grows. As business revenues, earnings, and dividends grow, stock prices also grow. The business factors that drive GDP growth are the same factors that drive stock prices.
Stock prices don't necessarily grow at the same rate as GDP, of course. In periods when inflation and interest rates are rising, stock prices probably won't grow as rapidly as GDP. In periods when inflation and interest rates are falling, stock price growth is apt to race ahead of GDP growth.
Nevertheless, over the long haul, it is still GDP growth that sets the growth path for total market capitalization; it is total GDP that defines the range within which total market capitalization will fluctuate.
Two sets of data are needed to show the relationship. GDP data is relatively easy to obtain. Within the U.S. Commerce Department, the Bureau of Economic Analysis (BEA) keeps careful track of the ups and downs of the American economy. For years from the late 1920's on, GDP numbers have been calculated according to a fairly consistent set of rules.
The second data set is not so easy to obtain, especially not if one wishes to go back several decades. While the desired market capitalization data for the New York Stock Exchange is available, similar data series for the American Stock Exchange and for NASDAQ don't go back nearly as far.
The Federal Reserve supplies numbers that help correct the deficiency. The Fed's Flow of Funds report contains market capitalization information from the mid-40s forward. None of the data sets published by the Fed give an exact measure of market capitalization, but they come close. Data from the Fed provides a useful supplement to the data directly available from the New York Stock Exchange, NASDAQ, and the American Stock Exchange.
When one charts these two sets of data on the same graph, the reasonably exact GDP numbers from Commerce, and the pretty darn close market capitalization numbers from the Fed and the different exchanges, the long-term relationship between total market capitalization and total GDP can't be missed.
From the mid-20s to the late 90s, the total value of GDP has grown roughly a hundred-fold (as expressed in current dollars, not adjusted for inflation). In the late 1920's, the nation's GDP was valued at about $100 billion a year. Now, in the late 1990's, GDP is closing in on $10 trillion a year.
During the same time period, total market capitalization has risen by roughly the same amount. With market capitalization, though, everything depends on the choice of the start-year and the end-year. Pick a low start-year and a high end-year, and you will see a faster growth rate. Pick a high start-year, and a low end-year, and you will see a slower growth rate. But these are quibbles, nothing more. From any reasonable perspective, the long-run growth rate for total market capitalization is roughly the same as the long-run growth rate for GDP.
Link 2: Fluctuations in the Ratio of Market Capitalization-to-GDP
Even though total market capitalization is tied to total GDP, the tie is a loose one. Total stock values rise and fall within quite a broad range. At its low points, relative to GDP, stock market capitalization has slipped below forty percent. When it has peaked, historically, stock market capitalization has topped 100% of GDP, but not by much.
From 1925 to 1995, stock market capitalization averaged about 65% of GDP, drifting below 60% of GDP just as often as it rose above the 60% mark.
But that seems like ancient history, now, in 1999. In the last three years, the market's upward climb has accelerated, sending total market capitalization to unprecedented highs. Relative to GDP, current market capitalization levels are now two-and-a-half times higher than their historic average.
This is a startling and extraordinary run-up. If the stock market's capitalization total were at its historic average, relative to GDP, the Dow would be somewhere in the high 3000's. It certainly wouldn't be closing in on 10,000.
Yet - while the run-up has been extraordinary - what we're seeing is not an example of never-ending exponential growth. This rapid upslope will, as it must, come someday to an end. It's likely that we're already nearing the top of this S-curve run-up. Like an S-curve, the Market Cap-to-GDP ratio is certain to level off at some point. The question is not whether, but when. And, once it levels off, will it slump back down, or simply go flat? No one can say.
Link 2 tells one of the hidden stories. The Market Capitalization-to-GDP Ratio is an indicator that hasn't received much public attention. Yet the picture it paints is one we can't afford to overlook.
Link 3: S&P; Index Growth and Total Stock Market Growth
Over the last seven decades, GDP has grown at 3.3% a year, adjusted for inflation. And the total stock market has grown at roughly the same rate over the same time period.
It's natural to assume, then, that the S&P; 500 index has also grown at about three percent a year, adjusted for inflation.
It may be natural, but it wouldn't be accurate. S&P; 500 index growth has actually lagged behind overall stock market growth, and by a noticeable amount.
Real growth in the S&P; 500 has been roughly 2.3% a year over this time period, definitely slower than the GDP's 3.3% a year real growth rate.
[Note: "Real growth" is econo-speak for the growth that remains, once inflation has been taken out of the picture.]
And even this number, 2.3%, may be misleadingly high. S&P; price appreciation calculations usually begin in the mid-1920's and end in the late 1990's, because that's the time period covered by Ibbotson Associates, the data source most commonly used for S&P; 500 data. But note carefully: the mid-20's start years in the Ibbotson data set were characterized by moderate Market Capitalization-to-GDP ratios. The mid-90's end years have been characterized by very high Market Cap-to-GDP ratios. As a result, there's probably an upward bias in any price appreciation calculations that rely on the time period covered by the Ibbotson data.
It might be more realistic to assess S&P; 500 price appreciation on a "comparable years" basis. In other words, if we want to start with a peak year, wouldn't it make sense to end on a peak year? If we want to start with a trough year, shouldn't we also end on a trough year?
And this is what's been done in the two charts that follow. The first is a "Peak to Peak" analysis of S&P; 500 price growth. It begins in 1965 and ends in 1995, when Market Cap-to-GDP ratios were at comparably high levels, at about 100%. The second is a "Trough to Trough" analysis, that begins with 1954 and ends with 1990, when Market Cap-to-GDP ratios were at comparably lower levels, roughly 55%.
Both charts normalize Total Market Capitalization, the S&P; 500 Index, and the Dow at 100 in the starting year. Both charts take out the effects of inflation. And both charts show the respective changes in value, versus the normalized start year value of 100, for the Total Market, the S&P; 500, and the Dow.
The results may surprise you.
From 1965 to 1995, Total Market Capitalization grew at a compound annual growth rate [CAGR] of 3.0% a year, adjusted for inflation. The S&P; 500 index grew at a compound rate of only 1.5% a year for the same peak-to-peak time period, lagging total market growth by quite a substantial margin. And the Dow's growth rate lagged even more, trailing along at 0.7% a year. By comparison, GDP grew 2.9% a year, almost the same as Total Market growth.
The results are similar in the trough-to-trough comparison.
From 1954 to 1990, Total Market Capitalization grew at a compound annual growth rate (CAGR) of 3.3%. The S&P; 500 index grew at only 1.9% a year. And the Dow grew at only 0.9% a year. Real GDP growth, by comparison, was 3.4% a year.
Do these results catch you off guard? The S&P; 500 index is commonly touted as an excellent proxy for the market as a whole, implying, of course, that the S&P; 500 grows at the same pace as the overall market. In reality, as the analysis so clearly shows, the S&P; 500 growth rate lags the total market's growth rate.
As Wharton's Jeremy Siegel has pointed out, this is to be expected. Whenever new companies are listed on the stock market, their stocks add something tangible to the total value of the stock market, but don't add anything to the value of the S&P; 500 index. Whenever an existing company sells new stocks, the total market will grow in size. The S&P; index will not grow.
And it's also possible that S&P; 500 firms, as a group, are so well established in their industry that most of them are no longer high growth firms. Certainly that's the case with the blue chip industrials listed on the Dow.
As a result, the growth rates for most stock indexes will almost always lag behind the total market's overall capitalization growth rate.
Link 4: Index Growth and Portfolio Growth
The numbers reported by an index don't represent real value to an investor, of course. To get real growth in value, one must own real stocks. In a real stock portfolio.
Will the compound growth rate for a stock portfolio based on an index be identical to the compound growth rate of the index itself? The answer is almost, but not quite.
Index fund managers do charge a management fee for their services. Index fund management fees are not as high as the fees charged by a conventional mutual fund, but fees they are, and they must be paid.
In estimating long-run portfolio returns, a deduction has to be taken for the fees paid to the fund manager. Two-tenths of one percent of assets, per year, is typical. If the S&P; index appreciates at 2.3% a year, the related stock portfolio is likely to appreciate at 2.1% a year.
Link 4 may not be the most critical link in the picture for someone who's invested in an index fund, and who's being charged a yearly management fee of only 0.2% on assets. It is, though, a significant link for those who elect to invest their money with stock-picking mutual funds. Those funds collect annual fees of 1% to 1.5% of assets. Those fees won't seem large during periods when the Market Cap-to-GDP Ratio is accelerating upwards. But they'll be quite noticeable if the Market Cap-to-GDP ratio flattens out or heads south.
Link 5: GDP and Corporate Dividend Payouts
Price appreciation is only part of the story for the index fund investor. Average price appreciation of 2.1% a year - or less - doesn't sound like a very strong rate of return. And it's not.
Reinvested dividends are the story's second strand. And a powerful strand they've been, at least in the past.
Over the past seventy years, average dividend yields on S&P; 500 stocks have been 4.6%. In other words, every $100 in stocks held in an S&P; 500 index fund yielded $4.60 a year in dividend payments.
When stock market experts talk about ten percent nominal returns on stocks, or seven percent real returns on stocks, they often neglect to mention the importance of reinvested dividends. They shouldn't. Seven percent real returns on the S&P; 500 consist of a mere 2.3% a year in price appreciation, augmented by reinvested dividends of 4.6% a year. (In math terms, 1.023 x 1.046 approximately equals 1.07)
In other words, from a seven percent return perspective, two-thirds of the heavy lifting has been done by reinvested dividends. Only a third of the growth can be credited to price appreciation.
But I'm getting ahead of the story. Link 5 points to a very strong relationship between the total amount of corporate dividends paid out to shareholders and the total size of the GDP. This ratio doesn't change much from one year to the next, no matter what's happening to the stock market as a whole.
The following chart shows the year-after-year relationship between dividends and GDP, using data only from the New York Stock Exchange. Year after year, corporate dividends paid to shareholders by firms carried on the NYSE are almost equal to 2 percent of GDP.
The New York Stock Exchange represents about seventy to eighty percent of total stock market capitalization. If one were to add in the dividends paid out by firms listed on other stock exchanges, total corporate dividends paid by firms listed on all U.S. stock exchanges would run somewhere between 2 and 2 percent of GDP.
This ratio is almost entirely independent of the overall value of the stocks listed on the stock market. In 1978, the first year shown on this chart, the Market Cap-to-GDP ratio wasn't much above 40%. In 1996, the final year charted here, the Market Cap-to-GDP ratio was close to 120%. In relative terms, the market was three times more valuable in 1996 than it was in 1978. Yet total dividends paid by NYSE firms had risen only from 1.8% of GDP to 2% of GDP.
This is an important finding, though it's often hidden from view. The total dividend cash flow paid to shareholders by companies listed on the nation's stock markets is quite constant. To reiterate the key point: Total dividend payouts run somewhere between 2 and 2 percent of GDP, and we can count on them staying in that range no matter what's happening to stock prices or to the Market Cap-to-GDP Ratio.
Link 6: The Dividend Yield Rate
The total dividend paid to the owner of a single share, divided by the value of that share, equals the dividend yield on that share. In almost exactly the same way, as Link 6 shows us, the total dividend payout made by all corporations, divided by the total market capitalization of those corporations, equals the dividend yield rate for those corporate stocks, taken collectively.
Now think what this means. Rain or shine, dividend payouts run 2 to 2 percent of GDP.
When it rains on the stock market, total market capitalization can sink to 40% of GDP or less. When market capitalization slumps to those levels, average dividend yields can be extraordinarily healthy - up in the five to six percent range.
When the market is at its historic average, with market capitalization equaling 65% of GDP, dividend yields on the S&P; 500 will average 4.6%.
When the sun shines on the stock market, dividend yields change dramatically. With market capitalization now equaling 160% of GDP, dividend yields have plummeted correspondingly.
Just think about the numbers. Corporate dividend payouts holding steady at roughly 2% of GDP. Total market capitalization at nearly 160% of GDP. What's 2% divided by 160%? Roughly 1%. That's all. Aggregate dividend yields in today's market have sunk into the 1% range.
This isn't just a theoretical point. The historical data bear it out. Consider the results of the following analysis. S&P; 500 dividend rates from 1945 through 1997 were sorted into five separate buckets, related to the Market Capitalization-to-GDP levels. What the analysis shows is just what Link 6 tells us it should show.
For those years in which the Market Capitalization-to-GDP Ratio fell between 38% and 59%, S&P; 500 firms had an average dividend yield of 5.18%.
For those years in which the Market Capitalization-to-GDP Ratio fell between 60% and 79%, the dividend yield averaged 3.61%.
And so on. The higher the Market Capitalization-to-GDP ratio, the lower the average dividend yield.
As today's stock prices keep rising, today's aggregate dividend yields keep falling.
Is this important? Who cares about dividends, anyway, when stock prices are rising as smartly as they've been doing over the past few years?
Social Security's would-be reformers ought to care, if they're hoping for a significant boost from the 21st century stock market. Can the stock market really deliver 2.3% price growth, and a 4.6% dividend yield rate, in the decades to come?
It wouldn't be incorrect to respond by saying, "Yes, if..." If the Market Capitalization-to-GDP Ratio collapses back to its historic levels, if dividend yields recover to their historic levels, then, yes, anything's possible. If, by chance, the total stock market should return to its historic Market Capitalization-to-GDP Ratio of 65%, then S&P; 500 dividend yields of 4.6% would certainly return as well.
But such a dramatic change seems most unlikely. The stock market has become much too popular. There's so much more money floating around, and it's driving stock prices up. Dividend yields of 4.6% are almost certainly a thing of the past. Link 6 points out a hidden truth that many Social Security forecasters would prefer to downplay. But we, as citizens, ignore this linkage at our peril.
Link 7: Stock Portfolio Total Returns
The final link, Link 7, ends with total returns to the index fund investor. That's you, as you may remember. Under many of the Social Security reforms that are being proposed, you can expect to see a portion of your payroll tax deposited in your very own Personal Retirement Account, and then deposited in an S&P; 500-like index fund. There it will stay until your retirement, with all dividends reinvested, growing in value every year.
What kind of returns can you expect to see in your stock portfolio? Low returns? Moderate returns? High returns?
Your total returns ultimately depend on two main drivers: Price Appreciation, and Dividend Yields. The faster stock prices grow, the more your portfolio will be worth. The higher your dividend yields, the more quickly your portfolio will grow as a result of reinvested dividends.
The analysis presented here is not meant to encourage too much optimism. Over the long run, total returns of five percent seem more likely than total returns of seven percent.
Price appreciation. Price appreciation is a tricky forecasting problem. There's been such rapid price appreciation in the past few years that many people, by now, probably assume that the same high rates will continue indefinitely into the future.
This would be a mistake. Sooner or later, and probably sooner, the Market Capitalization-to-GDP Ratio will stop rising.
Let's assume in this analysis that the economy's Market Cap-to-GDP Ratio stabilizes in some new range, relative to GDP. This vital ratio probably won't stabilize in the 40% to 100% range, its habitual range between 1925 and 1995. And the MC/GDP Ratio probably won't stabilize in the 130% to 160% range either as a long-term forecast, that seems unrealistically high.
Shall we assume that total market capitalization ultimately stabilizes in the 80% to 140% range, with an average value of 100%? No one can know for sure, but such an estimate seems reasonable. Move the range higher, and the long-range spread between stocks and bonds virtually disappears. Move the range lower, and the spread between stocks and bonds becomes temptingly wide, pulling stock prices back up.
Whatever the range turns out to be, it's sure to have a bottom limit, an upper limit, and a long-term average.
And it's the long-term average that concerns us here. What's the average amount of price appreciation likely to be, given the market's average MC/GDP Ratio over the next few decades?
To answer this properly, we'll begin with GDP. How fast can GDP be expected to grow, in the years to come?
Price appreciation: GDP Growth. Over the long run, GDP growth depends on two main factors - the population growth rate, and the productivity growth rate. Past GDP growth of 3.3% a year has been driven by population growth of 1.3% to 1.4% a year, and productivity growth averaging about 1.8% a year.
What can we expect population and productivity to do in the future?
Population growth is likely to slow down, perhaps by a little, perhaps by quite a bit.
From an environmental perspective, this is an excellent development. The world's population is already bumping up against the planet's total carrying capacity. The more quickly the world's population stabilizes, the better it will be for our grandchildren and our grandchildren's grandchildren.
Productivity growth, though, may recover a bit. It certainly sagged over the past quarter century. But the evidence suggests that this is changing, as a new digital sector takes its place alongside the economy's traditional manufacturing and service sectors.
Two percent a year productivity growth may not, in the future, be out of reach for the U.S. economy. But if population growth sags from 1.3% to 0.5%, total GDP growth isn't likely to keep pace with the 3.3% mark that the economy racked up over the past 70 years. Long-term GDP growth ranging between 2.5% and 3.0% seems more likely than real GDP growth of 3.3%.
Now, as Link 2 reminds us, Total Market Capitalization will, over the long run, grow at about the same rate as GDP. If the GDP growth rate comes out somewhere between 2.5% and 3.0% a year, for the next several decades, total market capitalization is likely to grow at the same rate.
Will index growth keep pace with market growth? As Link 3 reminds us, the answer is, "Probably not." There's a lag effect at work, and it's built in. No matter how broad the market index fund one chooses, there'll still be a lag factor separating total market growth from index growth. Index growth in the 1.5% to 2.0% range is more likely in decades to come, a bit below its 2.3% average between 1925 and 1995.
And then, as Link 4 reminds us, there'll be a slight additional loss due to management fees.
The prudent forecaster might not want to be too pessimistic, but it's difficult to make an argument for long-run index portfolio price appreciation above 2% a year. Long-term price appreciation forecasts as low as 1.5% might be quite appropriate.
Dividend Yields. To forecast likely dividend yields, we need only look at the two factors linked to dividend yield - total dividend payouts, and total market capitalization levels. Corporate dividend payouts will remain fairly steady. Total market capitalization may not sustain its current highs, at 160% of GDP, but it's not likely to sink back into the 65% of GDP range, either. Let's assume - who can know for sure - that the average Market Cap-to-GDP ratio in the decades to come turns out to be about 100%.
In that event, dividend yields, for the market as a whole, won't be nearly as high as they've been in the past. Aggregate dividend yields in the 2% to 3% range would be consistent with the overall logic of the market.
We can mark this estimate up a little, to take account of the fact that share repurchasing has now, for many companies, become a normal business practice. When a company repurchases shares, it has the same basic effect as a dividend payment - it puts cash back into the hands of shareholders. If share repurchasing is counted, total dividend yields of three percent plus would be a reasonable forecast.
Price appreciation a shade below two percent. Dividend yields (with share repurchasing factored in) a shade above three percent. A total real return of five percent.
This is, of course, meant to be an index fund average. But if Social Security is given the job of establishing Personal Retirement Accounts for the entire workforce, the population-wide average is the only number that counts. Get one hundred forty million Americans all investing in index fund accounts, and the best you can hope for is that the overall average will be reasonable. From a Social Security perspective, a five percent average is not an unreasonable forecast.
A forecast of seven percent, though, would definitely be unreasonable. It's not likely that we'll see average real returns of seven percent for index fund investors over the next seven decades, not with population growth rates slowing down, not with high Market Cap-to-GDP ratios, not with low dividend yield rates.
Recapping the Connections
I began this essay by promising an explanation for three sets of numbers - wage growth, GDP growth, and stock portfolio growth.
To begin with, all three growth rates have to be adjusted to take out inflation. Real wage growth of 1%, real GDP growth of 3.3%, and real portfolio value growth of 7%.
Real wage growth has been as low as 1% for two reasons. The main driver of real wage growth is productivity growth, and productivity growth has been weak for the past quarter century. An additional factor affecting real wage growth is the way in which productivity gains are shared between high salary workers and low wage workers. In the 50's and 60's, productivity-driven wage gains were shared fairly evenly by both groups. In the 80's and 90's, high salary workers have done quite well in this economy, capturing more than their share of total wage growth, while average earnings for low wage workers have often lagged behind productivity gains. As a result, news stories focusing on lower income workers commonly report very slow wage growth.
GDP growth has been a bit higher, at 3.3%, for reasons that should now be evident. While real wage growth depends solely on productivity growth, GDP growth depends not only on productivity growth but also on population growth.
Total returns to stock investors have been as high as 7% for reasons already explained in some detail. While stock price appreciation has lagged the GDP growth rate, reinvested dividends have more than made up the difference.
Now you can see the logic that connects the world of 1% wage gains to the world of 3% GDP growth and the world of 7% stock returns. The disparities turn out not to be so mysterious after all.
And with this logic in hand one of the key risks in the Social Security debate can be seen much more clearly. What if Congress accepts the 7% figure as the best estimate of future stock returns? And builds its Social Security reform package around it? We're in for trouble if that happens. One hopes the Congress and its Social Security forecasters will take a more realistic view of long-run stock market returns.
Copyright 1999, by the Collaborative Democracy Project. You are welcome to quote from this essay, or use its charts, provided you let us know ahead of time, and give us proper credit.
A Non-Profit Successor to the Collaborative Democracy Project
Page Version 1.02
Revision Date April 13, 2006