How I used the 4% rule over the past 29 years

When I quit my engineering job with a Fortune 500 chemical company back in 1994 at age 38, I hadn't seen William P. Bengen's landmark work on "Determining Withdrawal Rates Using Historical Data", FPA Journal, October 1994. I didn't come across the article until after I'd been retired for about a year.

I knew that large charitable foundations and college endowments managing their investment portfolio as a perpretuity usually limit their annual spending to 5% of assets. But of course, the annual dollar amount of that 5% spending went up and down with the value of the endowment. I was looking for a withdrawal rate that I could reliably increase with inflation over time, so I knew that had to be less than 5%. I chose 4% as a wild guess and felt comfortable with that since it was less than the dividend and interest income being generated by my retirement portfolio at the time.

My initial 1994 retirement portfolio consisted of about 20 individual stocks, mostly in Tech and Drugs, with the fixed income portion containing a mix of FDIC-insured CDs, and US Treasury securities. I avoided mutual funds because of the higher fees and expense ratios at the time. Even Vanguard's S&P 500 index fund had an expense ratio of around 0.20% in 1994. My long-term, buy and hold portfolio where I typically only made one or two trades per year had annual trading costs of 0.02% to 0.03% of assets. Since today you can buy an S&P 500 index fund with an expense ratio as low as 0.015% of assets (e.g. Fidelity 500 Index Fund), I'd do that rather buy individual stocks.

The late 1990's were a wonderful time to start retirement. The S&P 500 increased by 200% from 1994 to the market peak in 2000, the NASDAQ by almost 700%. My retirement portfolio had a couple of big winners in DELL and Pfizer which left me with about 6-1/2 times my initial 1994 portfolio balance at its peak, prior to the Dot.com bubble bust in 2000 and the market crash.

At retirement in 1994, I had about a 70% stock, 30% fixed income asset allocation, but planned to do some part-time consulting work, just to keep my resume fresh if I needed to return to the workforce. Over the first 2 years of my retirement, the consulting work averaged out to be about 25% of a full time job.

My retirement portfolio doubled within the first 2-1/2 years, and I decided it probably wasn't necessary to do any more consulting work. Then it doubled again in the next year to 4X the initial balance, bringing my annual withdrawal rate down to about 1% of assets. A 70/30 allocation meant I had 30 years of spending in fixed income which I thought was nuts. So I said, "A 60/40 allocation with a 4% withdrawal is 10 years of spending in fixed income, so I'll hold it there at 10 years and put the rest in stock." At a 1% withdrawal rate, that left me with a 90% stock, 10% fixed income allocation. I was at nearly 95/05 at the market peak in 2000. After taking about a 50% haircut in portfolio value during the 2000 market crash, I was back to 90/10.

I've had confidence over the years in the long-term stock market return data and just maintained my asset allocation through thick and thin. I don't time the market. Stocks go up and down, but the money you lose to financial advisor fees, commissions, trading costs, and taxes is gone forever. I learned my lesson during the Black Monday market crash in 1987 when the DOW dropped by 22.6% in one day. I stayed out of the stock market for the next 2 years. I'm at least 25% poorer as a result, but education is often expensive.

One thing that appealed to me when I investigated William P. Bengen's method using a larger data set from Yale economist Robert Shiller was the distribution of the terminal value of the portfolio after 30 years. As you can see from the table below, in 95% of the one hundred 30-year payout periods examined you ended that 30 year period with a higher balance than you started -- a higher balance despite 30 years of annual withdrawals. There's about a 50/50 chance you'll have 4X your starting balance and a 10% chance you'll end 30 years with nearly 8X. As long as you don't happen to retire on the eve of the next Crash of 1929 and the Great Depression, there's a good chance that the 4% rule will make you rich.

Resources for more information

How much do you really need to retire?, Quora.com post from w.w. Lenzo Ph.D Quantum Physics, Stanford University 1990

Life Expectancy vs. Income in the United States -- Opportunity Insights Harvard University

PortfolioCharts.com - site that shows lowest cost ETF/mutual fund choices for a retirement portfolio

Bengen, William P, Determining Withdrawal Rates Using Historical Data, Journal of Financial Planning, October 1994, pp 171-180, Volume 7, Number 4.

Investment Company Factbook 2019, ICI.org

You�ll Spend Less As You Age -- Time Magazine, Feb 26, 2014