Bloomberg Anywhere Remote Login Bloomberg Terminal Demo Request


Connecting decision makers to a dynamic network of information, people and ideas, Bloomberg quickly and accurately delivers business and financial information, news and insight around the world.


Financial Products

Enterprise Products


Customer Support

  • Americas

    +1 212 318 2000

  • Europe, Middle East, & Africa

    +44 20 7330 7500

  • Asia Pacific

    +65 6212 1000


Industry Products

Media Services

Follow Us

Bloomberg Customers

Retirement: What Price Risk?

By David Blitzer One of the points in the debate about Social Security is whether it's safe to transfer financial risks to individuals -- risks they may not have the wherewithal to accept and manage. Moreover, it's not clear if a typical worker can save enough to retire comfortably even if the risks are well managed.

The current Social Security debate is the latest phase of a longer-term shift in the way America provides for its citizens' retirement. Over the past two to three decades, a gradual and uninterrupted move has been under way from defined-benefit (DB) plans to defined-contribution (DC) plans.

In DB plans, the investment risks are borne by the plan sponsor -- usually an employer -- while in DC plans the risks are shouldered by the individual. Most of the new retirement programs of the recent past, including IRAs and 401(k)s, are DC plans. The nature of investment-risk management in DB and DC plans differs in two ways. In DB plans, the sponsor is usually a large organization with significant financial resources outside of its pension activities. Many companies cover major retirement expenses out of their business earnings.

PREEMINENT FACT. Second, and often missed, is that the risks are spread over a large number of beneficiaries and over time. If a retiree dies before his statistical life expectancy, funds that were expected to cover his needs will cover other retirees. In effect, risks are shared among the beneficiaries. In a DC plan, this is typically not so except where individuals use their savings to purchase life annuities.

Social Security -- a DB plan with additional insurance attributes -- is a risk-sharing device. Private accounts, as proposed in the current debate, put all the risks on the individual account owner. A key question is whether an individual can build a nest egg large enough to provide for his or her retirement with a cushion against inevitable investment risks. If it's easy and cheap to build a huge nest egg, no one need worry about risk.

But of course, the preeminent fact in investing is risk. Anyone who doubts that should look at some of the recent articles marking the fifth anniversary of the bursting of the technology bubble. Five years ago this month, the Nasdaq stock index was above 5,000. Today, it's around 2,000 -- meaning 60% of its value vanished. Although not all investors need to focus their efforts on what proved to be one of the more volatile approaches to making -- or losing -- money, any analysis of long-term prospects to investing for retirement must incorporate risks as well as returns.

60-YEAR CYCLE. The examples in this article are based on two models using the Monte Carlo system -- a mathematical model for computing the odds or probability of an outcome by testing thousands upon thousands of possible results. The first focuses on the accumulation phase -- saving and investing to build a retirement fund. This analysis is probably familiar to most people, although the range of possible outcomes may be surprisingly large. The second focuses on the retirement phase to determine the probability that a retiree will not outlive his money.

Both models share certain assumptions. The analysis is done in real, inflation-adjusted, terms. An average return of 6% is 6% before inflation -- or more like 8% to 9% in today's economy. Investment risks, returns, and inflation are based on annual data for the period 1945-2004. The choice of any historical time period is arbitrary. But using a 60-year period should be enough to even out most unusual events. The 1945-2004 period includes two major bear markets -- 1973-74 and 2000-02 -- as well as booms in the 1950s, 1960s, and 1990s.

The models use annual data and assume that stock market returns are normally distributed. There's some debate among analysts about how valid the normal distribution assumption is and, to some extent, the models may understate downside risk.


In these examples we consider an individual whose investing, or accumulation, phase extends more than 40 years, from age 25 to age 65 followed by retirement. To keep the examples simple, the accumulation phase considers two investment strategies: 1) 60% stocks, 30% bonds, and 10% cash (60/30/10); and 2) 100% stocks.

Table 1 shows the inflation-adjusted returns and risks for the strategies, including a third strategy of 30% stocks, 60% bonds, and 10% cash used in retirement.

Table 1: Investment Strategies



Return (%)

Risk (%)













Data for returns from S&P Financial Communications (2004)

Table 2 shows the size of the fund at age 65, assuming $1,000 invested annually. For an individual earning $40,000 annually -- roughly the median personal income in the U.S. -- setting aside $1,000 annually means saving 2.5% of one's pretax income. Investing $4,000 annually would represent 10% of pretax income. As we will see, this range roughly represents failure vs. success in planning for a modest retirement.

Table 2: Accumulation Phase Results





+ 1 Std Dev

-1 Std Dev













The risks in investing are large. An investor following a typical investment strategy of 60/30/10 -- common among DB plans -- could end his 40 years of investing $1,000 annually with somewhat more than $1 million, or nothing at all. The average fund is about $175,000, and two-thirds of the outcomes should be between $285,000 and $62,000.

For a more aggressive investor who is 100% in stocks, the average is $377,000, much larger than the more typical 60/30/10 approach. But the range is even larger -- the best results could exceed $6 million and the range covering two-thirds of the outcomes runs from $777,000 down to nothing. (Where the model shows net losses we assume that investors do not borrow to keep losing but just end up with a zero balance.) What these savings programs will do depends on the distribution phase during retirement.


The second model covers retirement. Each year, money is withdrawn from the fund and the remainder stays invested. As long as the fund balance is greater than twice the annual withdrawal, the process is considered safe. If the fund falls below this critical level, the retiree is assumed to be running out of money. The investment results depend on the portfolio strategy chosen -- in this example, a relatively conservative strategy of 30% stocks, 60% bonds, and 10% cash is assumed.

Here, we'll look at three cases: First, a less thrifty individual who annually set aside $1,000, or 2.5% of his $40,000 income, and hopes to draw out $30,000 each year in retirement; and two even thriftier souls who annually set aside $4,000, or 10% of their income, and who hope to draw out $30,000 or $40,000 each year in retirement. In developing these numbers, we assume the initial retirement fund is the average size found in the accumulation phase -- the less thrifty saver starts with about $175,000 and the more thrifty with about $700,000. The return and risk are shown in Table 1. Table 3 summarizes these results.

Table 3: Probability of Having Money Left (%)

Year 5

Year 10

Year 20

Year @ 85% Safe

Retiree #1 -- $4,000 per year in/$30,000 per year out





Retiree #2 -- $4,000 per year in/$30,000 per year out





Retiree #3 -- $4,000 per year in/$40,000 per year out





The individual whose 40 years of savings amassed $175,000 won't be able to sustain a $40,000-per-year lifestyle for very long. At the end of Year 5, there's only a 38% probability that he has more than two years (or $80,000) remaining in his fund. In fact, by Year 3 he has less than an 85% probability of being solvent. By Year 10 he will have exhausted the fund.

The thriftier saver who starts with $700,000 is far better off. While this is three-quarters of his preretirement income, he's no longer saving $4,000 annually, so he's actually looking at living on $36,000, or 83% of his preretirement spending. He's expected to have a 100% probability of solvency at 5 and 10 years, and a 94% chance at 20 years. His risk of hitting the critical level of two years of income reaches 85% in Year 25. Assuming he retired at 65, his life expectancy at retirement is 15.8 years, and he's unlikely to outlive his money.

Further, as the last three lines of Table 3 show, spending $40,000 per year would not raise significant risks through Year 10, and even at Year 17 (age 82) he would face only a 15% chance of hitting the critical fund level of two years of remaining income.

Retirement Rules of Thumb

Over a wide range of different returns and risks, the most important factor in determining whether a retirement fund will last long enough is what proportion of the fund's initial principal is withdrawn each year. In simple terms, if the annual withdrawal is about 4% or less of the initial fund, the money will probably last 20 years. Increasing the expected returns and risks does little to alter this 4% rule of thumb.

Clearly, if one expects to earn returns of 10% with risks typical of 4% conservative investment policies that normally offer returns of 4%, the results would look attractive -- but they wouldn't be what one might rationally expect, given the investment experience of the past century or so. Table 4 shows the probability of having enough money in Year 20 for the three investment strategies used above for different annual withdrawals. The withdrawals are measured as a percentage of the initial fund:

Table 4: Annual Withdrawal as % of Initial Fund





























One might expect that as long as the initial proportion withdrawn is close to the return, the fund should be stable -- that is, when the return is about 4% one can withdraw 4% annually, when the return is 6% one can withdraw 6%, and so forth. Because risks increase as returns increase this isn't the case. Withdrawing 4% annually -- implying that someone retiring on $40,000 a year has a $1 million nest egg -- is reasonably safe across all three investment strategies.

However, someone who withdraws 6% of the initial fund each year is much more likely to run out of money even if the expected return is close to 6%. In Year 20 he or she has a 79% chance of solvency compared to a 4% withdrawal program. The 4% rule of thumb holds across a far wider range of strategies and plans than those shown here.

Can the Median Worker Retire?

Our analysis began with the question of whether a worker earning $40,000 annually -- close to the median personal income -- could put away enough to retire. Assuming 40 years to accumulate savings for a retirement of 20 years and retirement spending of $30,000 to $40,000 a year, the answer is that some can, but it's far from assured. The average fund accumulated by someone setting aside 10% of his income would be almost $700,000 and that would support a retirement income of $28,000 using the 4% rule of thumb -- almost there. As Table 3 illustrates, withdrawing slightly more is only a modest increase in the risk.

However, this is an aggressive program with a savings rate substantially higher than most Americans manage today. A recent Federal Reserve release shows the savings rate, including consumer durables and housing, ranges from 2.4% to 10.1% over 2000 to 2004. Savings rates excluding consumer durables are lower.

However, not all aggressive savers will retire with ease. These numbers use the average nest egg. Given risks in the market, half the savers will have nest eggs smaller than $700,000 and will find their positions far less attractive. Table 2 showed that the savings for a person one standard deviation below the average would be $61,950 for each thousand dollars saved annually. For someone saving 10% of his $40,000 median income, this works out to about $250,000 in the fund and, at the 4% rule of thumb, this is retirement income of $10,000.

BIG IFs. The statistics argue that almost 17% -- those more than one standard deviation below the mean -- of the aggressive savers would retire on $10,000 or less. Current Social Security provisions would provide this worker with retirement of about $16,000 depending on his or her age, according a December, 2004, actuarial table from the Social Security Administration.

In the end, it's still a question of who bears the risks. On average, personal accounts may be attractive. If the investments risks become rewards, letting each person bear his own risks and rewards would be very remunerative. However, if the risks aren't shared and risks don't become rewards, the results for many will be disappointing.

Editor's Note: Related video clips can be found under "Hot Topics" at Blitzer is a managing director of Standard & Poor's and chairman of the S&P 500 Index Committee

blog comments powered by Disqus