By Bert Whitehead, MBA, JD
Dear Bert: I have read in various publications that the safe withdrawal rate from an investment portfolio during retirement is around 4% if you want your money to last. Could you please comment?
While I recognize that the 4% withdrawal rate has become the standard wisdom in financial planning, I respectfully disagree. Keep in mind that most financial planners are actually investment managers, and so minimizing the withdrawal rate keeps more assets under management for them, and correspondingly higher fees.
It seems to me that a withdrawal rate must take into account the after-tax return to the client. This is highly individualized, so I don't think you can know this unless you are intimately knowledgeable about the client's tax situation. It also focuses attention on the actual tax efficiency of the portfolio. Because Functional Asset Allocation is very tax efficient, I am able to keep almost all my clients with under $3 million in their investment portfolio in a 15% marginal tax bracket. This obviously impacts their appropriate withdrawal rate.
Most financial planners figure that a balanced portfolio in retirement with 60% interest earning and 40% equities will earn ~7% over a 15-20 year period. This is historically true, and is the number I use. But then they assume an 'inflation rate' of 3% and in one way or another come to the 4% withdrawal number.
I disagree that the inflation rate is the driver in retirement. While inflation may occur in general, the overall rate has little relevance to the actual rate an individual client experiences. The CPI (Consumer Price Index) is heavily weighted by education, housing, and medical costs -- none of which are significant to most retired people (especially with health insurance which most of my clients have, even absent 'universal coverage'). CPI may be meaningful to individuals who live at a subsistence level, but most people who have a financial advisor are affluent to some degree. Just ask yourself: "When gas went to $4.00 a gallon, did that affect my living standard?"
The driver for most clients is not cost-of-living, but their "standard-of-living." During accumulation stages a client's standard-of-living generally increases at a higher rate than inflation, usually in tandem with increases in their earned income. So using CPI through the accumulation period grossly underestimates the amount a client actually spends. Upon retirement many financial planners say that clients only need 80% of their pre-retirement spending. I find that, at the beginning of retirement, clients need 100% of their pre-retirement spending.
Retirement spending normally remains flat for the first 10-20 years of retirement, as the standard-of-living stabilizes. It is critical that planners monitor client spending during the first few years of retirement. If standard-of-living increases at the same rate as when they were working, they will certainly end up living beyond their means. However, there is a selection bias I've noted with financial planning clients. They tend to be savers rather than spenders. Most often I find that a client needs permission to spend because they are so accustomed to being frugal and are afraid that they will run out of money during retirement.
Interestingly, the actual expenses needed to support a client’s standard of living starts dropping around their late 70's and early 80's. They have 'been there, done that, have the T-shirt.' They don't need to buy new cars, or to keep up with fashion demands. If we exclude gifts to charities and children, the amount they need decreases year by year, regardless of what the CPI does or how their portfolio performs. Recent studies published in the Journal of Financial Planning have corroborated this phenomenon.
Two other points about estimating withdrawal rates…
Many financial planners use software that depicts an inflated future as a single estimated percentage increase of past expenses. As you well know, I consider financial planning to be a process, not an event. Clients are generally very capable at adjusting their behavior. If there is a lean year in the stock market, they put off an expense until times get better. The software projections don’t show how smart clients can be!
Another point is that investment managers define their value as “return on investment.” However, clients tend to view supporting their lifestyle in terms of liquidity, or simply “will I have the money?”
As you know, our approach (Functional Asset Allocation) uses 15 year bond ladders with US Treasuries to assure that a client's pension, social security, and cash flow from the bond ladder is sufficient to meet their living expenses, including income taxes. This approach requires that we manage a client’s living expenses, preferably for 5+ years before they retire so we can determine the amount needed from the bond ladder. Note that Treasuries are not included in a portfolio to generate yield, but rather to provide guaranteed cash flow. "Safety Trumps Yield" is our mantra for this portion of the portfolio. We stress liquidity, not performance.
The 15-year span enables the stock market to fully cycle, so that the bond ladder can be replenished during prosperous years. It gives clients assurance that they will not have to change their life style for 15 years, so they don't fret over stock market down cycles and resist capitulating during severe market drops. Even over the past 'lost decade' we were able to rebuild client's bond ladders during the up years of the market cycle, e.g. 2003-2004 and 2009.
Finally, we also factor in savings of 10% of a client's income each year, which is reinvested in their portfolio. Obviously if clients save 10% each year (which they are accustomed to during their working years), they will by definition continue living within their means. This eliminates the need to estimate their life expectancy and makes Monte Carlo theory, which calculates the probability of future investment returns, largely irrelevant because they will never run out of money.
In summary, 'withdrawal rates' that are based on combating inflation are much too simplistic to determine a client's real annual cash flow requirements. The driver for income in retirement is not a ‘withdrawal rate‘ that depends on the Consumer Price Index, but rather changes in expenses needed to support their personal standard-of-living.