Portfolio Timing

How Can a Flexible Schedule Increase the Saver’s Chances of Success?

In a previous article I detailed the historic range of outcomes experienced by the saver undertaking a dollar cost averaging (DCA) program over various time periods and investing in the S+P 500. It is well known that the stock market has provided a wide range of returns. That uncertainty is, in a sense, the definition of risk, and the reason why the stock market investor has in most cases been rewarded with higher returns. While some people claim that the risks faced by the saver are not as great as the risks of the market itself, I demonstrated that the range of returns for a DCA program was larger than the market returns in all periods studied. Put another way, sequence risk trumps lump-sum risk.

Most people try to reach retirement by undertaking an extended DCA program. Yet despite the popular chorus that “stocks provide superior long term returns,” there seems to be much uncertainty. So what is the saver to do? Let’s start by discussing a 30-year savings time frame. Below are the probabilities, reprinted, of realizing a certain internal rate of return over 30 years.

Image and video hosting by TinyPic

Even if you assume for a moment that the future will look something like the past (which is far from certain), how do you create a plan from that chart? One option would be to choose a relatively high probability of success, say the 80th percentile, and use that IRR for planning purposes. But what about the other 20% of the time? If you end up in a particularly bad year are you just SOL? At some point it starts to feel like gambling with your retirement.

Another option would be to chose the 100th percentile, and use the lowest IRR for planning. At least that way (if history comes close to repeating itself) you would be almost certain to meet your goal. But is this really a good option? With a real return of only 1.075%, you better be a serious saver to reach retirement in 30 years. Forget about taking the kids to Disney land when they’re young.

It’s important to realize that the very worst outcomes existed not only because of a bad sequence of returns, but perhaps more importantly because the final market value was at a real low point. For instance, the worst outcome for the 30-year saver was during a period ending in 1921. The real price (2012 dollars) of the S+P 500 that January was 87.16. The market hadn’t seen prices that low since 1878.  The all time January high up to that point was 271.5 in 1906. Five years after the low point, the market would rebound and almost double to 164.61 in 1926. In short, 1921 was a really bad year to get out the beach chairs. If our hopeful retiree had called it quits a few years earlier or later, they would have been in much better shape.

Despite the best-laid savings plans, we all might need to be flexible when it comes to precisely when we retire. Let’s try to quantify the benefits of flexibility.

If we examine each 30-year series for the saver, it is obvious that the high points (including contributions) will not always come at year 30. Once again, that specific year may be a low point for the market. If we look at the high points for each series, we can start to eliminate some of the worst potential outcomes. By using those high points as our new low points, things don’t look so bad. For example, the lowest high point for a 30-year DCA series was $245,385, reached by saving $5000 real dollars each year. Computed over a 30-year program, that comes to an IRR of just over 3%.

Just to be clear, when I make this type of comparison, I am not saying the actual IRR was 3%. Why? Because the value of $245,385 may have come in a year before the 30-year program ended. But what does it matter? In that case, the IRR is actually higher. Regardless, what the saver is really concerned with is the value they hit, not what the IRR may actually be. Another important point here is that if this high point comes before the 30-year program is up, the saver would have to know in advance to get out while the getting is good in order to be sure of their return. That is somewhat of a problem, but one I will address later on.

So, historically speaking, we could have done much better than the very worst outcomes merely by ending early in some cases. And what if we can be a little bit patient? The situation improves with each year we wait. Below is a table showing the lowest high points if we waited 5 extra years to retire. And again, I compare that value to its corresponding IRR computed over a 30-year period. I also list the IRR computed over a matching number of years. This second IRR is obviously lower, but again, does it really matter if the actual IRR is not so high? It’s up to you. Isn’t the actual number the important part? Maybe it depends.

Image and video hosting by TinyPic

So let’s say that at the beginning of our savings program we wanted to retire in 30 years but figured we could wait an extra five years if we really had to. To be 100% sure that we would reach our savings goal we could have used a real return of 5.018% for planning purposes. Based on our needs, we could then determine how much we should save each year in real dollars to reach that goal.

Using this approach, we can avoid the feeling that we’re gambling with our retirement, yet still use a reasonable real rate of return for our planning. In this case, we would have had about a 77% chance of reaching our goal by year 30 or earlier, and a 100% chance of hitting it by year 35. Not too bad.

But this brings us back to the issue of timing. If you use this type of approach, you must have the fortitude to get out when you hit your “number.” What you do with your money at that point is a topic for another day, but if you want to maintain the certainty of success, you need to bail early, even if you are years ahead of schedule. In some cases you would have hit your goal in only 17 years. What do you do at that point? It’s up to you. You could retire early, for one. Or, after pulling out of the market, you could keep working to continue saving (or not), placing your money in a safer vehicle in order to fund a richer retirement or other goals. There are many possibilities.

Of course this analysis scorns reality for a number of reasons. First, we cannot know that the future markets will resemble those of the past. Second, it might be wise to lower our estimates in the current era of higher valuations. Third, we would make various adjustments both up and down to account for the fact that a reasonable portfolio would hold a wider array of asset classes. The inclusion of high quality bonds might lower estimates of our IRR, while the inclusion of other factor exposures such as the small and value premiums, as well as a possible rebalancing bonus from other equity exposures, might raise our estimates.

Whatever estimate seems right to you, the important points are to use a number that will give you a high probability of success, withdraw from the markets when you reach your goal, and be prepared to wait a few extra years if need be. This is certainly not the only recipe for success, but it is one way to plan.

We can also extend this type of analysis to other timeframes. Below are graphs for 25, 20, 25, 10, and 5-year periods with up to 5-year waits.

Image and video hosting by TinyPic

Image and video hosting by TinyPic

Image and video hosting by TinyPic

Image and video hosting by TinyPic

Image and video hosting by TinyPic

For the 25 and 20-year periods things still look relatively good. Simply by focusing on the lowest high points for the series you were certain to at least keep your purchasing power over 20 years. And if you were willing to wait five more years, you could have hit a goal based on a projection using an IRR of 3.699%. The picture is a bit better for the 25-year periods.

For shorter time periods the analysis starts to seem pointless. In the worst case, for the five-year periods you could have used an IRR of 12.921% if you were willing to wait five extra years. However, waiting to year ten for a five-year plan is hardly the same as waiting until year 35 for a 30-year plan. In the former case, you are doubling your waiting period. On top of that, you may have hit your goal, but your actual IRR was -5.799%!

The chorus will often echo that an investing time frame of five (or sometimes ten) years or less means stay away from stocks. I would respond that even periods of up to 15 years might avoid stocks. If the expense you are saving for is highly necessary, and the timeframe is inflexible, then even periods of 20 or 25 years may be unsuitable for stock market investing. (What this means for parents with a 15-year timeframe hoping to send children to college is up to you.)

Along with risk tolerance and a few other factors, a person’s timeframe is often a major determinant of the overall asset allocation of their portfolio. I would argue that the flexibility of that timeframe should play a very important part in that discussion, or the possibility exists for the saver to be very disappointed.

Malcolm Lewis, January 3013


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s