Increasing the Safe Withdrawal Rate, Part II
In my last post I discussed how the most reliable way that a retiree can increase the Safe Withdrawal Rate (SWR) is to work longer, thereby decreasing the number of years a portfolio must support someone in retirement. In this post I will discuss the one situation where it is possible to have your cake and eat it too. That is, increase your SWR, not have to work longer, but also maintain a 90%+ probability that your portfolio will last 30 years.
However, this one situation is outside anyone’s control because it is related to the valuation of the equity markets at the time of retirement. That is, if you retire and begin your retirement income withdrawals when the stock market is at a low valuation, you can increase your SWR without increasing your risk of portfolio exhaustion within 30 years.
A really good report was written on this subject by a financial planning advisor in 2008 named Michael Kitces. It is a fairly long report, but I highly recommend it; you can click here to read it. For those who do not want to read it, I will summarize the report’s findings.
Michael Kitces’ report goes through a lot of rigorous analysis of market returns for 105 rolling 30-year periods from 1871 through 1975 (the last 30-year period ending in 2004) to determine the 100% SWR assuming a portfolio asset allocation ratio of 60/40 (stock/fixed income). His conclusion was that a 4.4% initial withdrawal rate (adjusted annually for inflation) would not have exhausted your portfolio in any rolling 30-year period from 1871 to 1975. Mr. Kitces is quick to point out that future market returns may be different from past returns, which implies that a 4.4% initial withdrawal rate is not a guarantee. But it has very a high probability of success. Based on this information, Mr. Kitces adopts 4.5% as his “baseline” SWR.
Mr. Kitces’ report goes through much analysis and discovers a very strong correlation between the 15-year real returns of the US stock market and the SWRs for the same starting years. He also verifies what is widely known, that the long term total returns of the equity markets is greatly affected by the market’s aggregate valuation at the beginning of the time period. That is, the higher the market value at the time you purchase the aggregate market (such as the S&P 500 Index), the lower the long term market return will be and vice versa.
In arriving at his final conclusions, Mr. Kitces introduces a market valuation metric the Price/(Real Earnings for 10-years) commonly abbreviated as “P/E10”. This metric was popularized over the last decade by Yale University economics professor, Robert Shiller. The P/E10 is a metric that is derived by dividing the current S&P 500 stock index inflation-adjusted value (the P in P/E10) by the index’s previous 10-year average of real earnings (the E10 in P/E10). The “E10” smoothes out the annual ups and downs of the S&P 500 index stocks’ real earnings. The P/E10 is considered a better market valuation tool to assess SWRs due to the long term nature of retirement income withdrawals.
Using the P/E10 measure as the market’s valuation metric, Mr. Kitces arrives at his conclusions about when it is possible to increase his baseline 4.5% SWR. I have summarized his recommendations in the table below. So if the P/E10 metric is below about 20, one can increase their SWR above the baseline 4.5%.
Rules for Increasing Safe Withdrawal Rates
Starting P/E10 Metric |
Initial Safe Withdrawal Rate |
Greater than 20.0 |
Baseline 4.5% |
Between 12.0 and 20.0 |
5.0% |
Below 12.0 |
5.5% |
What is the P/E10 for today’s market? This metric is not an easy calculation to make because it requires tracking the CPI then applying it to both the current S&P 500 index price and the previous 10 years average earnings for the index. Fortunately Robert Shiller has a web site where he publishes his latest P/E10 metric. The earnings are usually a quarter or two old, but since the earnings are a 10-year average, a couple more quarterly earnings reports would not significantly change the “E10” value. You can view Robert Shiller’s Excel spreadsheet P/E10 calculations for yourself by clicking on this link http://www.econ.yale.edu/~shiller/data/ie_data.xls. When the spreadsheet opens, go to the tab titled “Data.” The latest P/E10 metric is the last row in column K. Currently the calculated P/E10 metric, based on an S&P 500 Index value of 1,324 is 21.94.
Unfortunately, with the current P/E10 metric at about 22, we are not in one of the periods with low stock valuations. So if you are planning to retire this year, it would be wise to stick with the 4% to 4.5% SWR that is recommended by most financial planners.
To what level would the S&P 500 index have to drop so that one could increase the SWR above 4.5%? Assuming the “E10” stays the same as it is currently calculated by Professor Shiller, the S&P 500 Index would have to drop below about 1,200 to get the P/E10 below 20.0. This is certainly possible in the near term. So a 5% SWR may be possible if you are nearing your retirement date. But according to Mr. Kitces’ rules, to increase the SWR to 5.5%, the S&P 500 Index would have to drop below about 725.
Before you start rooting for a big drop in US stock indexes in order to increase your SWR, you need to keep in mind everything that a big market drop would mean. Let’s suppose that you are one year away from your planned retirement date and the S&P 500 stock index is at about 1,350 (the value as of this writing) with a P/E10 of about 22.3. Let’s also assume that you have a retirement portfolio of $600,000 with an asset allocation ratio of 60/40 (stocks/fixed income). The table below illustrates the impact of a drop in the S&P 500 stock index that, in turn, lowers the market valuation as determined by the P/E10 metric.
S&P 500 Stock Index |
P/E10 Metric |
Portfolio Value |
Allowable Initial SWR |
First Year Income |
1,350 |
22.3 |
$600,000 |
4.5% |
$27,000 |
1,100 |
18.2 |
$533,300 |
5.0% |
$26,700 |
700 |
11.6 |
$426,800 |
5.5% |
$23,500 |
This table illustrates that when the S&P 500 index price drops and, in turn, lowers the P/E10 metric; it also lowers the value of your original 60/40 portfolio. So, even though you can increase your SWR, you will be applying the higher SWR to a smaller portfolio value. The above table indicates that it is possible that your first year income would actually be higher if the S&P 500 index price (and the associated P/E10) does not drop even though your SWR is lower.
This example shows why, as you get closer to your retirement date and you feel you have saved enough money to provide the required income at the baseline withdrawal rate, it is important to pay attention to your equity allocation. This example also shows why you should set up a fixed income ladder in the event that there is a prolonged drop in the equity markets just before retirement.
How would I handle my retirement assets if I was the person exemplified in the table and the current P/E10 metric was over 20? Assuming I could live on the $27,000 first year income (and if I could not, I would not be planning to retire in one year); I would dial back my equity allocation to about 40% to protect what I have. If the US equity market experiences a big drop to, say, around 1,000 (a 26% drop), this would change my asset allocation from 40/60 (stock/fixed income) to about 30/70, I would then rebalance my portfolio back to my 40/60 target allocation. Given today’s corporate earnings, a drop of the S&P 500 stock index to 1,000 would represent a much better market valuation than we currently have. So, in this case, I might even consider increasing my equity allocation to 45% or 50% depending on my overall market outlook.
If, after the 26% drop, the market looked like it was stabilizing, at that point I might consider increasing my withdrawal rate to 5%. The market would be at a much better valuation, so withdrawing 5% would not represent greater risk of portfolio depletion. In this case the market index dropping to 1,000 would mean that my portfolio value has dropped about 10% to $540,000. So a 5% withdrawal rate would provide the same $27,000 withdrawal amount allowing me to maintain my planned dollar withdrawal amount going into retirement. I know this does not actually increase the initial annual dollar amount I can withdraw, but if I had not changed my allocation from 60/40 (stock/fixed income) to 40/60, the higher 5% withdrawal rate would mean I would be withdrawing a lower dollar amount in my first year of retirement as indicated in the above table. This is what I meant above by “protecting what I have”. If the market experienced a strong uptrend in the first few years after retirement, I would likely not increase my 5% withdrawal rate, but I might consider in any given year increasing my withdrawal dollar amount more than that year’s inflation adjustment (always remembering to re-balance your portfolio if the market valuations get too high).
To summarize, if the US equity market indexes have a low valuation as indicated by the P/E10 metric, you can increase your SWR above the typical recommendation of 4% to 4.5% to 5% or 5.5%. However, the current P/E10 metric indicates that the equity markets are over-valued based on historical norms, and does not warrant an increase in the baseline 30-year SWR.
There is one other way to increase the SWR without decreasing a portfolio’s survival rate. This last option is the one I use and it requires some flexibility. I will describe this approach in my next post.
Did you enjoy this post? Why not leave a comment below and continue the conversation, or subscribe to my feed and get articles like this delivered automatically to your feed reader.
Comments
No comments yet.
Sorry, the comment form is closed at this time.