Monday, July 6, 2015

Early Retirement: Spending Sooner

In my first post on this topic, The Risk of Retiring (or Being Retired) Early, I provided some thoughts about the risk of retiring early and perhaps extending what might already have been a long and expensive retirement. In my second installment, Retiring Early: Lost Savings, I reviewed the risk of retiring early and consequently saving less. As I mentioned in that second post, limiting savings at the end of retirement has a significant impact, but when we stop saving early, we typically also start spending from savings. The cost of early spending is greater than the cost of forgone savings and the two combined are substantial.

(As always, click on a chart or table to see a larger version. Hover your mouse over any yellow text.)

Here's the chart from that second post showing the cost of forgone savings without the simultaneous cost of early spending, in other words, imagine a retiree who could stop saving, retire, and avoid spending savings until age 70. (Note a minor change to the chart from my last post: this one graphs balances at the beginning of each year whereas the last post assumed the worker retires at the end of the year.)

Chart 1
As you can see, were this imaginary retiree able to retire at age 55 and not touch retirement savings until age 70, her portfolio value at age 70 would be about $265,000 less than it would be if she had kept saving until age 70. She would save $156,000 less over those 15 years and lose $109,000 in interest on those forgone savings.

Typically, however, when a worker retires and stops saving for retirement, he also begins spending from savings. Chart 2 below shows the resulting portfolio balances when a retiree simultaneously stops savings and starts spending at a given retirement age. It also shows the amount of spending supported assuming a constant-dollar annual withdrawal of the portfolio balance at the retirement age.

The amount of the sustainable withdrawal percentage is calculated using Milevsky's formula for sustainable spending (download PDF) using the life expectancy from the "male" columns of the following table. Note that females have slightly lower SWR's because they have longer life expectancies. My first post on this topic noted that the longer you postpone retirement, the greater your expected savings will be and the larger the percentage you can safely spend annually.

Table 1.
For example, if our sample retiree stops saving at age 65 (the inflection point in the purple line on Chart 2), he would have accumulated $1,181,178 by age 65 and Milevsky tells us we can assume he could spend 4% of that amount, or $47,247 annually beginning at age 65. If he saves until age 70, he accumulates $1,712,935 and Milevsky tells us he can spend 4.6% of that amount annually because he has a shorter life expectancy.

These probably seem like hugely different outcomes, and they are, so let me walk through one example of retiring at age 65 (purple curve on Chart 2 above) versus waiting until age 70 (teal curve on Chart 2 above) in Table 2 below.

Table 2.
Retiring at 70 allows the retiree to contribute $56,000 more to savings in this example. Five more years of 7% annual returns with no withdrawals provides over $530,000 more in portfolio savings. Together these amounts create a portfolio at retirement five years later that is $531,756 larger. Because the 70-year old has a 4-year shorter remaining life expectancy, he can spend 4.6% of this portfolio, according to Milevsky, which is 15% more than the 4% he could spend at age 65. The increase in spending from 4% of $1.18M to 4.6% of 1.71M is more than $30,000 a year.

A lot of this difference comes from the huge growth in the portfolio the last few years of retirement resulting from compound earnings. These portfolios grow exponentially and each year that you delay spending affects your savings balance more than it did the year before. (This is why most financial planners urge you to be very cautious with your investments the last decade of your working career.)

A substantial amount of the sustainable spending difference also comes from the increased SWR – the retiree gets to spend a larger percentage of a larger portfolio. In this example, the additional spending increases $21,269 a year from a larger portfolio at retirement and another $10,278 from an increased SWR.

This scenario is an example and there is no guarantee that your portfolio will grow at all in the final five years of your career, let alone that it will grow as much as 7% annually. The intent is only to show how changes in retirement age affect retirement spending. How much it affects spending depends on market returns and life expectancy, things we can't predict.

The earlier you retire, the less money you can spend after you retire. A significant portion of the reduction of retirement spending can be attributed to the fact that you stopped saving earlier, and a larger portion of additional cost is attributable to spending savings earlier. Toss in the lower sustainable spending amount at younger ages because we have to plan for a longer retirement and the body blows add up quickly.

So far, those body blows from retiring early include:
  • a longer (and consequently more expensive) retirement, 
  • fewer years to save, 
  • lost returns on those forgone savings, 
  • lost returns on savings that we spend at an earlier age, and 
  • a lower sustainable withdrawal rate (or life annuity payout) due to the longer expected lifetime in retirement.
Of course, you can turn that frown upside down by looking at the flip side of those bullets as advantages to delaying retirement: a shorter, less expensive retirement, more years to save, etc.
There is still at least one major financial risk to consider when deciding to retire early, or evaluating the impact of forced early retirement, and that is the impact on Social Security benefits. I'll cover that next time in Early Retirement and Social Security Benefits.

Note: The assumptions for these calculations are the same as in the initial post, The Risk of Retiring (or Being Retired) Early. I assume the worker will earn the "typical" annual incomes shown in the charts in that post and will save 10% of earnings every year. I assume he or she will earn a consistent 7% annual return on all savings (an optimistic assumption in current capital markets). All calculations are in nominal dollars except for expected market returns used for the Milevsky formula, for which I assume a 5.6% real annual return with 11% standard deviation.


  1. Just out of curiosity, who is the "retiree able to retire at age 55 and not touch retirement savings until age 70" and what is an example of where the money comes from to pay the bills? Part time post-retirement work? Pension?

    1. I doubt there are many in real life. Hard to imagine a pension that pays out at 55. Getting hard to imagine a pension, period. A part-time post-retirement gig might do the trick for someone willing to live extremely cheaply. (Rob Greenfield, maybe?) But if we keep working, are we really retired?

      No, the point is to show the difference between the cost of discontinuing savings and the cost of starting to spend savings early. It's hypothetical, like SWR assuming that people will keep spending the same amount each year even when it's clear they're going broke. We don't really think it will happen, but we assume it has demonstration value.

      Thanks for writing!

    2. "As you can see, were this imaginary retiree able to retire at age 55 and not touch retirement savings until age 70, her portfolio value at age 70 would be about $265,000 less than it would be if she had kept saving until age 70. She would save $156,000 less over those 15 years and lose $109,000 in interest on those forgone savings."

      Dirk I have read your blog for some time and like it but honestly are you advocating to be the richest corpse in the cemetery? I retired 6 weeks short of my 57th birthday and haven't spent a dime of savings. It has grown substantially but I do live a very frugal lifestyle. Between a pension and SS at 62 I bank $1700 to $1900 a month. There are a few months I take from bank account for bills like this one due to property taxes but then there are a month or two that I save $2000 to $2200. I think having 15 years in retirement and having less at 70 is OK but I realize I am atypical. Without a pension and retiree medical I couldn't have retired that young.

    3. I think I have answered this a few times already, but once more to be clear. When you retire is a personal and often non-financial decision. But if you do get to make that decision and not have it made for you, I think you should fully understand the financial risks before you do. That is the sole purpose of this series of blogs. I really can't be more clear.

      I retired in my early fifties and I'm perfectly happy with my decision so the only thing I'm advocating is that you understand what you're getting into. It was riskier than I understood and it's a hard bell to unring.

      Thanks for writing!

    4. So, Will, two people have responded that they have been able to retire early and not touch savings for a long time. In both cases, the answer seems to be some combination of being fortunate enough to have both a pension and Social Security benefits and, in one case, paid retirement medical insurance. Living frugally appears to be another piece of the equation. Of course, they aren't done so we don't know how life will turn out for them.

      I agree with your original sentiment, however, that most retirees won't be this fortunate. I created the scenario as a demonstration, but apparently, it isn't totally out of the realm of possibility.

      One thing I have learned from my blog is that when you have thousands of readers, you can almost always find a couple of them who have done what you suggest they probably cannot.

  2. Dirk, great series of articles on the benefits of retiring late! The assumption of 7% return may be unrealistic though, considering one is likely to be at the bottom of a V-shape glidepath (Pfau & Kitces 2014) near the retirement date in order to reduce sequence-of-returns risk.

    1. Thanks, Isaac. I hope I made it clear at several points in the series that a nominal 7% return on an investment portfolio seems unlikely in today's capital markets (though not impossible – no one knows), no matter what your asset allocation. If you assume low returns when you're making up examples, though, the differences get hidden.

      I'm not a fan of glide paths, by the way. In Wade's words, "a glide path is what we would recommend to a retiree about whom we know nothing except their age, or to a mutual fund. A custom asset allocation is always better."

      Glide paths are academically interesting. Outside the lab, optimize your allocation annually. I do like the idea of starting out retirement (and ending your career) with a relatively low (40% to 50%) equity allocation. I retired in 2005 and a healthy helping of cash and bonds pulled my buns out of the fire in 2007.

      Thanks for writing!

  3. Dirk; In your response to Mr. Seldon you said it was hard to imagine a pension that pays at 55 or any pensions at all. Although I didn't retire until I was sixty, I started drawing my first pension at 38. The USAF has paid me a pension for the past twenty-nine years with 27 Cost-of-Living adjustment to boot. After the AF I worked for an established aircraft manufacturer who started paying a pension when I retired in 2008. My wife and I currently spend 85% of our fixed income leaving our savings (both deferred and taxable) to grow even though our cowardly 35% equity,50% fixed income, and 15% cash isn't recommended by many advisors today. Our spending runs between 55k and 60k a year - which makes for a pretty nice retirement.

    1. Excellent point! I was referring to private sector pensions, which are going the way of the 8-track tape player, and I should have been more specific. Defined benefit retirement plans are still fairly common in public sector jobs, usually to the exclusion of Social Security benefits.

      You won't get rich in public sector jobs (unless maybe you're a U.S. Congressman or Senator), but they can look pretty sweet in retirement. Many states have public pension plans for teachers instead of Social Security, too, though many appear to be rethinking that.

      My primary point to Will was that the example was hypothetical.

      By the way, I have no problem with your 35% equity portfolio at all. There's nothing cowardly about avoiding risk that you don't need to take.

      Thanks for writing!