What was most striking to me when talking to advisors and panelists at a recent conference was the almost doctrinaire use of the 4 percent "safe rate" as a benchmark for retirement withdrawals--a number several studies have enshrined. As thorough as some of those studies were, we need to reconsider this conclusion based on real-world applications of the withdrawal process and the changing demographic and economic circumstances of retirees. Advisors need to further review variables such as asset class selection and inflation assumptions. Indeed, the percentage issue is just one of the myths about retirement income planning we need to dispel.
Many advisors still select portfolios that are dominated by fixed income and/or dividend-paying investments to fulfill retirement income needs. The insistence on income and income-oriented investments for retiree portfolios may turn out to be a major pitfall for a successful retirement strategy. Portfolios that are primarily fixed income-oriented should not be the standard strategy for clients seeking retirement income.
Bonds are certain; they are not safe. They pay a certain interest rate for a certain period and pay a certain value at maturity or when called. Unfortunately, the trade-off for certainty has historically been a much lower return than other, less certain investments. As our clients retire earlier and live longer, portfolios consisting of bonds or mostly bonds won't work to provide a sustainable and growing income. Bonds are a diversifier for retirement portfolios; they are not a retirement income solution.
In fact, the greatest risk for longer-lived retirees is not aggressive portfolios but under-investing. Investors focus primarily on portfolio risk when their greatest danger is outliving their money because of returns that are too low. The real wildcard for retirees is longevity. Current actuarial tables are based on stale data that bears little resemblance to today's actual landscape in terms of the health of older Americans.
More importantly, what retirees need is cash flow--not income. The difference is that income is associated with interest-bearing investments like bonds, and to a lesser extent dividend-paying stocks. Income-oriented portfolios are dominated by bonds and are doomed to deliver lower returns over time. Cash flow comes from the systematic liquidation of securities where dividends and interest have been reinvested.
To solve this problem, we use a system at our firm called DIESEL: Dividends, Interest and Equity SElect Liquidations. DIESEL combines quarterly rebalancing with harvesting of portfolio components to generate cash flow that funds retirement "paychecks." When rebalancing, portfolio components that are out of favor or over their allocation percentages are used as sources of cash to fund withdrawals. DIESEL portfolios do not rely on bond interest to generate spending money for retirees and hence can be invested more aggressively. Based on historical data, broadly diversified portfolios combining rebalancing and harvesting would have produced higher returns than bond or mostly bond configurations, with a volatility level that provides a stable platform from which ongoing withdrawals can be made.
DIESEL is designed to work with mutual funds. Funds allow for efficient asset allocation, reinvestment of distributions and incremental liquidity that lends itself to cash flow generation and rebalancing. The system is easily administered in a fee-based, no transaction fee (NTF) environment or a wrap arrangement.
With DIESEL, it becomes clear that one of the primary flaws of the 4 percent safe rate studies is the lack of sufficient asset class diversification. These studies used only stocks, bonds and cash represented by the S& P 500, a bond index and a Treasury bill index. Inadequately or improperly diversified portfolios can have two problems: Either insufficient return as in bonds and cash, or unacceptable volatility as with stocks. Combining the three asset classes does not solve the problem.
A more properly diversified portfolio would have contained all four major asset classes: Stocks, bonds, cash equivalents and hard assets, as well as sub-classes that further diversify by capitalization, country and investment style. Our studies showed that portfolios configured with sufficient diversification were better able to sustain higher withdrawal rates through a combination of adequate returns and low volatility.
We also reconsidered the assumption that withdrawals must be increased by the Consumer Price Index. Remember: The inflation rates from 1973 to 1981 averaged more than 9 percent per year, and safe return studies increased their withdrawals accordingly. In the hypothetical portfolios studied, these massive withdrawal increases occurred when U.S. stock market performance was in the very low single digits and historically high interest rates devastated the bond market. The concurrence of these factors dooms the safe return study's three-asset-class hypothetical portfolios to failure unless withdrawals are kept very low: less than or equal to 4 percent. The use of the actual CPI is one of the primary sources of the 4 percent safe withdrawal mantra.
The need for an inflator of withdrawal amounts in these stu-dies is clear; however the need to use the actual CPI as that inflator is less clear. Advisors must carefully consider this issue and ask: Does the CPI actually measure inflation for the average retiree? Does it make sense to increase withdrawals each year based on the CPI? And do retirees actually do that?
In our practice, clients who are taking cash flow from their portfolios rarely ask for "raises." Not once in my 30 years in the industry has a client cited the CPI as a reason for any increase in cash flow from a portfolio. In fact, according to government sources, spending actually decreases after age 70 as retirees become less active. We have observed this in our practice as well. As long as major medical and long-term care risks are properly insured, the probabilities are good that retirees will not need an increasing income throughout their retirement years.
Due to these observations, when we created the study that led to the DIESEL System, we chose not to adjust portfolio withdrawals each year based on the actual CPI. Instead, we chose to smooth out the changes by using a 3 percent constant rate which is three-quarters of the 50-year CPI average. With a constant 3 percent withdrawal increase, the DEISEL System's hypothetical portfolio was able to sustain a withdrawal for the period between January 1973 and December 2006 equal to 7 percent of the initial value. (See Box, right.) Despite the withdrawals, the portfolio value increased seven-fold over the period. When returns were randomized, the DIESEL portfolio showed a lower failure rate than other, less well diversified, portfolios.
It is important to note that during the same period, the DIESEL portfolio also depletes using a withdrawal rate tied to the CPI. However, the hypothetical DIESEL portfolio outlasts any combination of the S & P 500, bonds and cash equivalents. Furthermore, with a 7 percent withdrawal increased at 3 percent per year, any combination of S & P 500, bonds and cash equivalents runs out of money before December, 2006.
NB: Implementation of the DIESEL System requires consolidation of investment assets at one institution:
o Consolidation allows easy, one-stop shopping for all financial activities;
o Portfolios are easier to monitor;
o Asset allocation is easier to manage;
o Consolidation benefits retirees by simplifying overall money- related responsibilities;
o Simplification becomes increasingly important as retirees age and become less interested in the details of their financial processes.
DIESEL cash flow is produced by maintaining a balance in a "withdrawal account" --usually the consolidation account's money market checking account-- equal to five months of the retiree's cash flow requirement. This allocation is in addition to what would be designated for the cash allocation pursuant to the client's Investment Policy Statement. The withdrawal account is usually a money market fund that is part of the consolidation account. Retirees can get their retirement paychecks by having money electronically transferred to an operating checking account each month or by simply writing checks.
Once the client has been designated a "DIESEL Income Recipient" (DIR), agreed on a monthly withdrawal amount, and the withdrawal account has been seeded, withdrawals can begin. The account is reviewed every 90 days to determine how much needs to be added to the withdrawal account to bring it in line with the three-to-five-month threshold figure. Once this amount is determined, our portfolio managers review the account for appropriate "harvest" candidates. Generally, these liquidations are done in such a way as to rebalance the portfolio back to its target allocation by liquidating positions that are out of proportion, out of favor, or that will generate a needed tax loss.
Although "harvesting" occurs quarterly, paychecks are generated monthly. This allows a monthly withdrawal while minimizing transaction costs and allowing for smaller balances in low-yielding money market funds. Since the differences between quarterly and monthly rebalancing were minor in our study, we chose quarterly to limit transactions, paperwork and record keeping. Taxes can be withheld from these "paychecks," often eliminating the need for those pesky quarterly estimated payments and adding further to the convenience of account consolidation.
Financial professionals can provide a systematic withdrawal program that produces reliable--but not guaranteed--retirement paychecks. Such a program historically has been able to sustain withdrawals greater than the widely accepted 4 percent safe rate. A key component of the DIESEL process must be careful monitoring of portfolio values and an accompanying caveat to retirees that, despite historical support for the reliability of the system, significant or sustained portfolio declines may require reductions in monthly withdrawals.
Of course, the DIESEL system takes some work: Properly trained staff, with one group assigned to review and calculate the required funds for each DIR and another assigned to making the portfolio changes necessary to generate these funds for the withdrawal account. But isn't it worth the effort to give your clients a retirement plan designed for the 21st century?
The DIESEL Study
January 1972 - January 2007
Study Portfolios Contained:
o S&P 500 Index;
o DFA Small Cap Index;
o MSCI EAFE Index (dollars);
o NAREIT Index;
o Goldman Sachs Commodity Index;
o Citigroup Composite Bond Index;
o Three-month U.S. Treasury Bill Index.
The study portfolio was rebalanced quarterly with quarterly withdrawals taken proportionately from each asset class after rebalancing. All dividends and interest were reinvested. Withdrawals began at 7 percent of beginning value and increased 3 percent annually with no taxes, fees or transaction costs included. The study was meant to be an apples to apples comparison on a level playing field. It showed that, all things being equal, a portfolio utilizing multiple asset class and sub-class investing would have historically sustained a higher withdrawal than other portfolios.
Stephan Cassaday CFP, CFS, is president and CEO of Cassaday & Company, Inc., an independent financial planning and investment advisory firm in McLean, Va.