Briefing Book on Social Security Solvency

A. How to critique the Republican strategy

1. The GOP’s Rosy Assumptions Scenario

2. How to analyze Index Fund Equity Returns
        - Twentieth Century: Historic Index Fund Returns
        - Twenty-First Century: Likely Index Fund Returns

3. Applying a realistic ROI for stock returns

4. Adjusting corporate bond estimates

5. Applying realistic assumptions for management fees and inheritance leakage

6. Applying a realistic assumption for Family Emergency Leakage

7. Applying a realistic assumption for Treasury interest rates

8. Eliminating the “Borrowing OK” assumption

9. Indexing Benefits to Inflation, not Wages – its weakness in a faster growth economy

1. Rosy Assumptions Scenario. We begin with an idealistic success scenario for Personal Retirement Accounts (PRA’s), using assumptions that are fairly close to Mr. Bush’s current game plan.

In the rosy assumptions scenario, what happens?

(PRA Annuity Payments are in the lighter color, SSA benefit payments in the darker color)

• PRA’s make negligible contributions until the mid-2030’s. By 2075, PRA’s generate 44% of the monthly total, or $783 a month in 2003 dollars, while the regular Social Security benefit schedule generates $989 a month, for a combined total of $1,772.

• Total assets owned by PRA’s rise to about 90% of GDP. Half those assets are stocks, half are bonds, so stock ownership equals about 45% of GDP. (Bear in mind that, for the 1925 – 1995 time period, total stock market capitalization averaged approximately 65% of GDP)

2 - How to analyze Index Fund Equity Returns.

As we'll see, Republicans greatly overstate likely stock returns to index fund investors.

Here’s the proper method for analyzing Index Fund Real ROI’s. Read the chart from the bottom up, starting from the lower left. (Potential index fund management fees omitted in this analysis.) This chart analyzes historic returns to an index fund based on the S&P 500 stock index. The next chart applies a similar analysis to future stock market returns.

> GDP growth of 3.3%. Workforce Growth of 1.3% combined with Productivity Growth of 2% to generate GDP Growth of 3.3%.
> Index Fund Lag Factor of 1%. Index Fund Capital Growth lags GDP growth, especially for index funds that favor well-established firms and omit young firms in new, faster growing sectors of the economy. Capital Growth in the S&P 500 index lagged GDP growth considerably. (To avoid bias in measuring the historic lag factor, it makes sense to choose Start and End Years with similar Market Capitalization to GDP Ratios.)
> Index Fund Capital Growth. Capital Growth in the S&P 500 index averaged 2.3% a year, a full percentage point slower than GDP Growth.
> Market Capitalization as Percent of GDP. This ratio ranged from 35% to 105%. Short-term changes in this ratio affect short-term Capital Growth rates. They also affect Dividend Yields. Over the long run, Market Capitalization growth essentially tracks GDP growth.
> Corp. Dividends as Pct of GDP. Dividend-paying corporations tend to grow their dividend payouts at a very steady, constant rate. Total Dividends paid by New York Stock Exchange firms were roughly 2% of GDP.
> Dividend Yields. Across the entire stock exchange, if Corporate Dividends equal 2% of GDP and the total Market Capitalization equals 40% of GDP, Dividend Yields will average 5%. In the S&P 500, Dividend Yields are a little higher – Dividend Yields for the S&P 500 averaged 4.6% from 1925 to 1995.
> Index Fund Equity Returns. Average Capital Growth in the S&P 500 of 2.3% a year (after inflation) combined with average Dividend Yields of 4.6% a year produced Index Fund Equity Returns in the S&P 500 averaging 7% a year.

Here’s how the same analysis looks from today’s 21st Century perspective.

> GDP growth of 1.8%. Social Security actuaries forecast long-run Workforce Growth of 0.2%, and Productivity Growth of 1.6%, which combine to generate 1.8% GDP Growth.

> Index Fund Lag Factor. The Index Fund Lag Factor may be smaller, going forward, if the relevant index funds are substantially broader than the S&P 500. Still, not all the growth sectors in the economy show up immediately in any index fund. Index funds also lag GDP growth due to the dilution effects from stock bonuses to executives. A prudent Lag Factor estimate will fall in the 0% to 1% range.

> Index Fund Capital Growth. Capital Growth over the short run will be high during periods when the Market Capitalization to GDP Ratio is rising, negative when it’s falling. Over the long run, Capital Growth will at best be even with GDP Growth, 1.8% by Social Security estimates. It could be a full percentage point slower.

> Market Capitalization as Percent of GDP. While this ratio ranged from 35% to 105% between 1925 and 1995, strong investor interest in the market seems to have pushed the long-run Market Cap to GDP Ratio permanently higher, by a very substantial amount.

> Corp. Dividends as Pct of GDP. If corporate profits are strong, if stock dividends are added to cash dividends in calculating total Corporate Dividends as a percent of GDP, the figure for Corporate Dividends as a Percent of GDP might rise above 2% just a bit. This chart shows a range, 2% to 3%.

> Dividend Yields. If Dividends are in the 2% to 3% range, and Market Capitalization to GDP Ratios fluctuate between 100% and 200%, a reasonable estimate for long-range Dividend Yields is 1.5% to 2.0%.

> Index Fund Equity Returns. Index Fund Capital Growth ranging between 0.8% a year and 1.8% a year, combined with Dividend Yields ranging between 1.5% a year and 2.0% a year, produce a “prudent investor” estimate for overall Index Fund Equity Returns that’s hardly better than bond returns, in the range of 2.3% a year to 3.8% a year.

A prudent forecast for long-run equity returns will be much lower than the GOP’s claimed 6.5%.

3. Now we’re ready to peel away the ideal assumptions in the GOP case. First we trim the likely real ROI for stocks from 6.5% a year to 3.5% a year.

One could certainly go lower than 3.5%, but 3.5% is a prudent estimate.

• Monthly annuity payments from PRA’s drop from $783 to $536. (Social Security’s adjusted benefit schedule remains constant.)
• Total PRA assets peak at about 70% of GDP, not 90% (not shown)
• Social Security debt stays the same, reaching 10% of GDP. (not shown)

4. Next we adjust the expected bond return rate. Economists think real bond returns could go as low as 1.6%. We’ll use 2.2% as a reasonable guess.

• Monthly annuities from PRA’s drop from $536 to $439. (Social Security’s adjusted benefit schedule remains constant.)
• Total PRA assets peak at less than 60% of GDP.
• No change in Social Security debt

5. Now we raise management fees from 0.2% to 0.3%. And we assume 35% inheritance leakage. For PRA owners who die before retirement, 65% of their PRA assets go into spousal PRA’s; the other 35% goes not into PRA’s but to children and grandchildren as spendable cash.

• Average monthly annuities from PRA’s drop to $383.

6. Next we adjust the assumption for Family Emergency Leakage, raising it from 0.0% to 0.5%. With as many as a million families a year driven into bankruptcy by catastrophic illness, it’s reasonable to assume Congress will allow such families to tap their PRA funds. And it’s also reasonable to assume that a great many of them will do exactly that.

7. To be on the safe side, we no long assume 3.0% interest for Treasury bonds. We synch Trust Fund earnings on Treasury bonds with the interest rate that PRA’s earn on corporate bonds.

• Monthly benefits from Social Security fall to $953, as the target benefit rate must be trimmed from 55% of its current target to only 53%, in order to protect solvency.

• With no borrowing allowed, it’s essential to make sure that Social Security has enough receipts and enough assets to meet obligations.
• In this simulation, we maintain solvency by pulling up the target date for achieving the new benefit level. We pull it up from 2045 to 2038, then stabilize at the new level.
• What happens? Ultimate benefit numbers in 2075 remain the same. Note the middle years though. The benefit plus annuity total is well below $1,000 a month.
• Given that the No Borrowing scenario is safer from an overall federal budget perspective, it is the more prudent assumption.

9. Bush’s team has suggested an important change in the benefit formula. Instead of indexing future benefits to growth in wages, as is done now, they’d index future benefits solely to inflation.

> If Social Security switches from wage indexing to inflation indexing, future retirees would receive no benefit increases whatever no matter how rapid the nation’s GDP growth.
• Let’s start with the $953 per month in regular Social Security benefits that we’ve seen in the previous examples.
• So long as the starting benefit formula is indexed to wages, with benefits held at a level that ensure lasting solvency, the 2075 average would be:
          - No increase in GDP growth • • • • • • • • $ 953 per month
          - Half a percent faster GDP growth • • • $1,315 per month
          - One percent faster GDP growth • • • • $1,845 per month
• Were the starting benefit formula to be indexed to inflation, though, average monthly benefits would be frozen at a predetermined level indefinitely, no matter how rapid the GDP growth rate.

Recap: Realistic assumptions take much of the shine off Mr. Bush’s Personal Retirement Accounts strategy

Remember. You can reproduce all this analysis by using our online Simulation Tool.

B. How to Critique the Democrats' Approach
C. Solvency Strategy for Republicans
D. Solvency Strategy for Democrats
E. Faster GDP Growth: Impact on Both Strategies

Briefing Book Summary

HOME

Revision Date 2005-05-02