The 4% is interesting in the context of big market swings around retirement time, when the capital for the 4% is set and first year withdrawal established. Consider the following scenario.
Bill, Mary and Pat all have a target retirement age of 60. They and their spouses have had long and enriching careers and they have saved a lot and are nearly ready to retire. At the beginning of 2007, each had saved a $5MM and had it invested in a total stock market index fund (let's suppose they and work in California where such a sum is fairly modest given 30 or so working years and high COL). Each saves $100,000 per year by being frugal, using Republic Wireless and biking to Costco for dinners out.
Bill turns 60 in 2007. He retires Jan 1 2008 with savings of 5.37MM (S&P 500 appreciated 5% in 2007, plus he saved another 100K that year). He lives the rest of his life on $215,000/year (4%).
Mary turns 60 in 2008. Her $5.3MM Jan 1 2008 fell to $3.3MM by the end of the year (just like Bill's 'cuz the S&P 500 tanked 37% that year). She retired Jan 1 2009 on $140,000 per year even though she worked a year extra while Bill saved nothing, blew $215,000 on vacations and fine dining, worked on his blog and caught up on daytime tv.
Pat retired Jan 1 2010 with $4.52MM (S&P surged 25% during 2009). Pat will live on $181,000 annually, even having worked and saved two more years than Bill.
Hardly seems fair or sensible, does it? Should we ignore big swings in the market near retirement time and just assume a 10% annual appreciation the last year or two to smooth out any wild gyrations which would give us a crazy high or low starting value from 4%? Doing so (using 10% in each case instead of the actual) would put Bill at $224K, Mary at $250K and Pat $280K. Great for everyone, especially Mary! What would you do?