• 4% may be insufficient for your needs
• depending on your age or state of health 30 year plan may be too long/short
• this model of constant real (inflation adjusted) requirements may not be consistent with your needs; e.g. you may a higher initial spend-rate followed by a lower rate until say home-care or nursing home needs are higher again starting in the late 80s

the assumption is that you start and stay with the same scheme throughout retirement
So innovative planners are always in search of options on how to increase or at least customize withdrawal rate for retirees to try to meet their perceived and/or actual needs. I’ll describe below some of the proposed approaches. I’ll start with a more complex model by William Bengen, because this model explains many of the variables involved in determining the outcome.

I. William Bengen, in an August 2006 article in the Journal of the Financial Planning Association entitled “Baking a Withdrawal Plan ‘Layer Cake’ for Your Retirement Clients”, as well as his excellent recent book aimed at financial planners entitled “Conserving Client Portfolios in Retirement” enumerates a number of withdrawal schemes:
1. Lifestyle– whereby an initial withdrawal percentage is identified and then adjusted for inflation each year, so that your real income remains unchanged throughout retirement (or until you run out of money) 2. Life phase– whereby your retirement is considered to be multi-phased with early retirement phases requiring higher withdrawal rates (reflecting higher spend rates when you are younger/healthier) and lower rates in later retirement years (and then higher again just before death). 3. Performance based- whereby a fixed percentage of portfolios are starting value each year can be withdrawn. So the funds available for spending each year are a function of what the performance of the portfolio each year. So you can expect a significant fluctuation in both real and nominal withdrawals. A floor and ceiling scheme in the real value of the withdrawn dollar values reduces fluctuations somewhat.
Bengen also introduces a SAFEMAX concept, which is the maximum initial withdrawal rate that will sustain a portfolio of a certain asset mix for a minimum number of years. Number 3, the performance based scheme with floor and ceiling enhancement, turns out to be preferred and leads to superior SAFEMAX and lower income volatility.
Bengen calculates the SAFEMAX rate for 30 year longevity of the:

• Life-style scheme to be 4.15% and 4.4% for 60:40 mix of large company stock (LCS) and intermediate term government bond (ITGB) and 4.4% for the 20:40:40 small company stock (SCS), LSC and ITGB. (He also provides the SAFEMAX figures 10, 20 and 40 year longevity are 8.9%, 5.2% and 4.2%, respectively.)
• Life-phase scheme to be 4.8% for the 20:40:40 mix when one takes a 3% reduction below CPI per year between years 11-20; i.e. to get 4.8% vs. 4.4% or about an extra 8% dollars during years 1-10 there is a 20-25% dollar reduction between years 21-30
• Floor & Ceiling scheme results in 4.9% and 5.16% for -5% and -10% floors respectively (independent of the ceiling); clearly a preferable approach than the life-phase one which result in considerable extra pain.

Bengen’s paper actually looks in detail at other factors entering his ‘recipe’ for the withdrawal plan, including: asset allocation, time horizon, target estate size and probability of success ( i.e. that of not running out of money in selected time horizon).
II. Jonathan Guyton’s work is discussed in a WSJ article entitled “Spending in Retirement” describing how the 4% rule of thumb for annual withdrawals can be increased to 5.5% for a 30 year horizon with 65:25:10 stock:bond:cash portfolio if certain guidelines are adhered to, such as: (i) if value of portfolio after you have withdrawn money is lower than at start of year, do not make an inflation adjustment the next year and (ii) when losses/gains are large then “guardrails” are used, i.e. when portfolio withdrawal rate ( or current withdrawal rate, the term used by Bengen) increases/decreases by more than 1% from 5.5% starting value, then add the inflation adjustment but then decrease/increase withdrawal rate by an additional 10%.
III. Jonathan Clements In another interesting WSJ article “Survive retirement even if short on savings” suggests a strategy to spend down assets with reduced risk of exhausting them for a 65 year old with limited resources. The approach has the potential of increased income if portfolio performs well. Take 85% of the portfolio and draw 1/20th the first year, 1/19th the second year and so on. The other 15% is invested into stocks and inflation indexed bonds and if the individual lives past 85 can buy an annuity or just spend down the accumulated assets. (Sounds like a credible approach, though no reference is made as to how such a scheme may have performed for historical or statistical scenarios)
There is no reason to stay throughout retirement with the same withdrawal strategy. In fact, Bengen in his book “Conserving Client Portfolios During Retirement” introduces the Current Withdrawal Rate (CWR) which is “the ratio between the withdrawal made in any year and the value of the portfolio at the beginning of that year.” He then proposes monitoring of the CWR on an ongoing basis and taking corrective action if it increases beyond some threshold. (His thresholds, at age 60, 70, 80, 90 and 100, are 5.5%, 5.9%, 6.7%, 9% and 20%, respectively. You can find the rationale for these figures on pp.104-105 of his book).
In summary, these schemes, or more likely more than one of these combined or used opportunistically as needed at different points during retirement, are food for thought for developing/changing a workable withdrawal plan for each individual. Changes will be warranted by significant depletions/increases in asset base due to sustained bear/bull markets, individual’s health situation (life expectancy) and estate considerations. The bottom line is that depending on the strategy you choose (level of risk that want to take) the above recommendations range from 4%-5.5% for a 30 year horizon and the indicated asset mix, based on historical data.