Is the 4% rule safe enough for our family?
We plan to leave our corporate jobs in less than two years, and in order to do that, we need to be confident that we can live on our investments. So the question for us is how safe is the often quoted rule of 4% withdrawals, and does that rule still hold in the crazy market we find ourselves in today.
I covered what the Sustainable withdrawal rate (SWR) is and where it comes from in a previous post. So if you haven’t read that yet head on back and take a gander. However, one constant question you hear from folks is, “is it different now?” Due to market overvaluation, or low-interest rates, or low dividend yields, or high inflation, etc. is the SWR of 4% going to continue to hold into the future? And what SWR can we expect if we’re retiring today or very soon?
So is the 4% rule safe enough for the Grizzlies?
Short answer – Pretty safe. The longer answer is that it very much depends on the current market conditions such as PE ratios, dividend yields, and interest rates and those conditions do point to being slightly more cautious than not. But still pretty safe. REALLY LONG answer below. Beware – math! There are TONS of articles about this around the internet, I’m not the first one to look into it. However, this is one area that I wanted to verify for myself since it’s critically important for my family. And I haven’t seen a rigorous look at what to expect in the current market conditions. So I took some of the brain power I normally use at work and applied it here.
How will we answer it?
MATH! or more specifically we’re going to construct a relatively simple linear regression model for determining the likely current SWR given various market data such as PE ratios, current interest rates, and dividend yields.
What Market Data?
I’m going to use for different metrics to capture the current market environment. Each with a slightly different rationale. We could include more such as government debt to GDP, GDP growth, consumer spending, etc. But I found these four explain almost all the variation in the expected SWR and I’d prefer to keep the model simpler rather than adding unnecessary complexity. So the for four metrics are:
- Cyclically Adjusted SP 500 PE Ratio, often called the Shiller PE Ratio after Robert Shiller a Professor at Yale – This Price earnings ratio is based on average inflation-adjusted earnings from the previous 10 years relative to the current price of the SP 500 index. It’s generally thought to be a good measure of market value and is less sensitive to wild swings that the current years PE ratio of the SP500.
- One Year SP 500 PE Ratio – simply, just the trailing 12 months of earnings vs. the price of the SP 500 Index. This is included to capture the current market sentiment more accurately than the Shiller PE ratio.
- 10 Year Treasury Yields – the Average yield on 10 year US Gov treasury notes. This is to capture the current interest rate environment which will have two primary effects. One, it determines the yield on the bond portion of our portfolio. Two, interest rates have an inverse correlation with stock prices. I.e. in low-interest rate environments like that of today you should actually expect PE ratios to be higher than average.
- SP 500 Dividend Yield – used in order to capture the current investment yield environment and the expected return shareholders can expect.
What do we do with this info?
Our model is a fairly simple linear regression of the data points above vs. the SWR as calculated in the Trinity Study. For the non-math types a linear regression is simply an attempt to build an equation that looks like this:
That best matches the historic data we have. Y is the projected SWR in any given year. The X’s are the data sets we’re using – in this case our various PE ratios, Dividend Yields, and interest rates. A,B,C,D are the coefficients were going to try to estimate using the historic data.
Does this actually help us predict the SWR?
Yes. The SWR rate in any given retirement year can be pretty accurately predicted by looking at these data points. The graph below shows the actual SWR for a 60/40 stock/bond portfolio alongside the SWR as projected by the model.
It’s a pretty tight match. I can get into more detailed statistics on exactly how tight a match it is, but the important takeaway is that if you have these four data points you can predict the SWR within about +/- 1.2% points.
What does this mean for the current SWR of a portfolio?
The right-hand side of the orange line above are the projected SWRs for the last few years based on actual data. What’s our current SWR for 2016? Around 4.2%. What this means is that based on everything we can see about the market right now the 4% SWR probably still holds, but it is riskier than in the past. Additionally, there are new contenders for the lowest SWR. Both 1999 and 2006 are shaping up to have lower SWR than any other time in the data set. This could improve if markets improve over the next years. But there may be implications for early retirees that want to maintain absolute safety.
What does this mean for our family and yours?
It means that while financial independence is still a very real possibility, we have to be a little more cautious than if we were leaving our jobs in 1981 or at the very bottom of the stock market crash a few years ago! We have to be more careful with our spending plans and adjust our withdrawals accordingly. Or be slightly more willing and able to make money in other pursuits outside of corporate America.
Bottom line is that these findings are not changing out plans for early retirement and we still feel relatively confident in our plan.
Where can I find your calculations?
If you’re interested you can find them at the link below (you’ll have to register):
Let me know if you’re interested in any of the details behind the methodology. Happy to offer explanations!