Market cycles make predictions hard for those who like to make such predictions. I believe predictions are impossible, especially over the rest of your life when it comes to retirement.
This doesn’t mean we all throw up our hands though. It means taking a measure of what is prudent based on facts as we know them each year. I call this Dynamic Updating.
There is a lot of debate about whether the “4% Rule” should be lowered because stock returns and bond yields are lower compared to the past. So those who argue this would also probably be those who would have supported using the higher-returns-influence on retirement income estimations during the late 1990’s.* Why? That would be the flip side of arguing for lower income estimations because returns are now lower. One can not have it both ways – good income during good-return times and lesser income during worse-return times.
But, as the diagram simply illustrates, returns are over or under a long term average. Jeremy Siegel points out here, average rates of returns and their deviation, continually change over time. The truth is, nobody knows what future rate of returns will be.
- One way to incorporate expectations of lower future returns is to lower retirement income expectations as suggested by reducing the withdrawal rate to 3% – that’s a 25% reduction in retirement income based on expected poor future returns.
- Note that this approach may lead to a long(er) term reduction in spending, not only for this year, but at least as long until the pendulum swings and proponents begin to argue for withdrawal rates higher than 4%!
- This tends to be the approach used by the Safety First School of Thought (sometimes called liability matching).
- Another way, is to simply adjust the long term returns average, AND standard deviation, at the beginning of each new year (Dynamic Updating).
- Note that this approach may lead to short(er) term reduction in spending, and tends to also allow for faster adjustments, either up or down.
- This tends to be the approach used by the Dynamic School of Thought.
So let’s look at an example of how retirement income may be affected by Dynamic Updating of the facts as they’re know at the beginning of each year. The Slide Share slide shows a prudent income for this year may be $3,867/month (under those facts and assumptions). Should that couples portfolio decline to the projected balance shown (from ~$860,000 to ~$760,000), then their prudence income should be reduced to $3,477/month. This is just under 10% pay reduction (($3867-3477)/3857) … which is temporary (AND, only IF the portfolio value goes down) … at least until the portfolio value may be reviewed and updated over the following subsequent years. Less than this 10% reduction may be taken earlier – and this earlier change would also change the decision portfolio values – all improving the odds the portfolio may last longer because less is withdrawn while the price per share is lower (in other words avoiding selling more shares at depressed prices – you run out of money when you run out of shares!). ** Thus, one can’t, and shouldn’t, make longer term choices based on short term events or on expectation – only on facts.
Why reduce spending? Because this is the most effective manner to preserving retirement resources for future spending needs.
So – the consensus is that retirement spending should be reduced. The questions are 1) by how much, and 2) when?
When you add the latest data to a long series of data, the mean and standard deviation doesn’t change as much as the differences between the last two data points. The longer the data series, the more stable the mean and standard deviations … this is called regression towards the mean. Actual data (what I mean by, and how I, use historical data for real returns and standard deviation***) has a regression towards its’ mean. Expected returns don’t (at least I haven’t seen a long term data series that records historical expectations – and such data should also regress towards the true historical mean; otherwise expectations would not be of any value since they diverge from fact) … thus, expectations keep changing and tend to be as volatile as short term data series.
Using data is NOT making a prediction! Regardless of approach … the coming year may go up, or may go down, and spending decisions would be needed about what adjustments may be needed, and by how much that adjustment may need to be, under any retirement income scenario. The profession spends lots of time discussing data use for simulations or calculations, but seems to miss the larger picture that the objective of the exercise is to establish Decision Rules that relate to what to do when something happens. That is what is illustrated in the Slide Share example.
People think about retirement income being steady as if they had steady income throughout their working years. In my experience, that isn’t so. You made adjustments then … you should expect to continue to do so throughout retirement as well. The goal is to remain as close to your standard of living as possible as much as future reality allows. Just as in the past, facts may call for adjustments.
* Industry professionals will remember that returns expectations were so high back then that a limit was placed on the maximum rate that could be used. 12% is the maximum return for illustration purposes – and even higher rates were sometimes used before the limit was imposed … demonstrating my point that expectations often follow along quite closely with more recent market results (recency bias).
** Additionally, adding a “Reserve Bucket” (a “Distribution Reservoir“) from which monthly cash flow comes from, aids the longevity of the entire portfolio by allowing for turning off, or turning on, distributions from the “Long Term Portfolio.” The Benefits of a Cash Reserve Strategy in Retirement Distribution Planning by Shaun Pfeiffer, Ph.D.; John Salter, Ph.D., CFP®, AIFA®; and Harold Evensky, CFP®, AIF®, Journal of Financial Planning Sept 2013; login required.
** *Volatility (commonly measured by standard deviation) takes away from prudent withdrawal rates what returns give. Capital Market Expectations, Asset Allocation, and Safe Withdrawal Rates by Wade D.Pfau, Ph.D., CFA, Journal of Financial Planning Jan 2012 (Figure 2); login required.
Note: Although this article is about downward spending adjustments, increased spending is also possible during better times. Decide then to spend it or keep in invested, instead of making more permanent decisions today that are harder to undo in the future.
Note: The line illustrated in the graphic above slopes up and to the right depicting the tendency for markets to have positive long term averages. However, shorter term results may result in a loss of value. Where that loss is temporary, or permanent, depends not on the market – it depends on what you do! What Bernstein calls turning “shallow risk” into “deep risk” by ill-considered actions.
Note: Everyone is in the same boat in 2014 – regardless of if the markets were better when one person retired compared to poorer markets when another person retired. It doesn’t matter when someone retired. Markets move everyone up and down the same – assuming the same exposure to them; more affect with more exposure to stocks, and vice versa.
Photo By F l a n k e r (Own work) [Public domain], via Wikimedia Commons