MSN Money recently carried a well written article on safe withdrawal rates. The author highlighted recent studies which called into question the "standard" 4% SWR which is often used for financial planning. One of the studies cited suggesting that WR above 1.8% may not be safe while the other suggested that a WR of 7% may be sustainable.
While both studies highlighted a number of assumptions and qualifications to their analysis, for present purposes the point is that such a wide range of outcomes is possible depending on what the inputs are. Since all of the inputs involve making assumptions or guesses about the future and experts routinely get their predictions wrong, it is very hard to do anything other than be ultra conservative when financially preparing for retirement. If I had been content to blindly rely on a 4% WR I'd be writing this post from the beach (maybe not given the weather at the moment, but you get the idea). I'd also be fretting about running out of money later in life.
So what affects the sustainability of a given withdrawal rate? The main variables and how we intend to deal with them:
1. expenses - how much we spend and how those expenses grow (or reduce) over time either due to inflation or personal circumstances. We've monitored our expenses closely for several years now and are very comfortable with our ability to maintain our standard of living once we retire. There is a substantial safety margin here. I arbitrarily added 20% to our budget, ignored the fact that our children will eventually become independent and assumed that Mrs Traineeinvestor will retire at the same time that I do. We also have the ability to cut some expenses without much grief if we need to;
2. net return on investments - how much our investments return. My basic assumption is that our risk assets (equities and real estate) will earn a real rate of return equal to the net of everything income they generate and that over the long term that rate of return will increase at about the same rate as our expenses. This may or may not be realistic but I would be uncomfortable assuming a higher rate of return;
3. sequence of returns - risk investments do not generate "average" returns. Some years they do better and some years they do worse. Getting a few bad years at the start of a sequence can destroy a retirement plan that doesn't allow for this possibility. Since I can't predict the future any better than the next person, I plan to keep at least two years' of living expenses in cash or near cash (ccurrently we are sitting on close to five years' of expenses which is too much). This should allow us a buffer to draw against in years when the cash flows fall short of our needs;
4. investor behaviour - apart from the risk of simply selecting bad investments, there is the not insubstantial risk of panicking and selling at the bottom of a market rout (which will happen from time to time). Without the crutch of a pay cheque this risk is not immaterial. At the risk of sounding delusional, I don't expect to be consistently right in my investment decisions but when markets get to extremely levels, I hope I can resist the pressure to panic sell at the bottom or insanely buy at the top. Having a cash buffer and not being dependant on selling assets to meeting living costs is a useful tool in dealing with this risk;
5. duration - how long do our savings have to last? In our case 50+ years. For all practical purposes this might as well be forever and I have planned on the basis that the real value of our investments should not decline over any length of time (market fluctuations aside).
Given my thinking on these issues, asking "what is a "safe" WR?" is largely irrelevant - I'm simply aiming to invest in largely risk assets to offset long term inflation and spend at least 20% less than the net dividends and rent received from those assets with a meaningful cash reserve to fall back on should their be any disruptions to those cash flows.
As an aside, if I had a solid COLA'd defined benefitt pension, welfare entitlement or similar, my confidence levels would be higher as these reduce risk. But I don't and, over pensioned civil servants aside, not many other people will be able to rely on these going forward either (which is a good thing).