Monday, February 1, 2016

Expense Risk in Retirement

I have recently posted about bankruptcy risk, positive feedback loops, chaos theory and Kaplan-Meier estimators and now I'll try to tie all of those posts together. They're mostly about the unpredictability of spending in retirement compared to sequence of returns risk, the probability of outliving your retirement savings due to market volatility. A retiree can enjoy a favorable sequence of market returns, limit spending to a sustainable amount, and still suffer insolvency as a result of huge medical bills, divorce or identity fraud.

Much of retirement writing focuses on the unpredictability of market returns, but unpredictable spending is a greater risk. The most you can lose of your savings is 100% of your portfolio, but you can have unexpected expenses far greater than your savings – a medical catastrophe, for example. Retirees typically plan for a 5% to 10% probability of outliving their retirement savings, but about a half-percent of Americans 65 and older file for bankruptcy, and that number appears to be growing.

First, the big picture. Your retirement finances will likely include some combination of Social Security benefits (in the U.S., or a similar program where you live), private pensions, personal retirement savings, and more and more, working longer. Sadly, private pensions, known as defined-benefit (DB) plans, are disappearing and a sizable majority of Americans don’t have a significant amount of retirement savings to invest. The GAO recently found that older households typically have retirement savings only “equivalent to an inflation-protected annuity of $310 [to] $649 per month.”

Your retirement finances will also include expenses and that spending side of the equation is significantly less predictable than market returns. Most retirement research assumes that you will only spend a "sustainable" amount of your savings portfolio ( a "spherical cow") but in reality, you will spend whatever life costs. Spending a sustainable amount of your portfolio is "retirement savings insurance", not bankruptcy insurance.

Spending a sustainable amount of your portfolio is retirement savings insurance, not bankruptcy insurance.
[Tweet this]

David Blanchett and Sudipto Banerjee showed that retirement spending typically declines about 2% annually for retirees spending appropriately for their wealth and that even when there are large end-of-life expenses these are usually less in real dollars than spending was at the beginning of retirement. Spending crises, however, are not typical. They develop from unpredictable and uncontrollable expenses like large medical bills, housing problems, debt problems, and fraud.

Much retirement research focuses on how to make your personal retirement savings last your entire lifetime, by spending only 3% or 4% annually, for example. This does not, however, consider what happens when life costs more than your sustainable withdrawal plus your Social Security benefits. If it costs a lot more, you could end up insolvent.

You can manage your savings portfolio's longevity risk by reducing spending, but elder bankruptcies are most often the result of spending crises. By definition, you can’t reduce spending in a spending crisis. Controlling spending is the problem, not the solution.

When a household’s finances crash and lead to bankruptcy, it is most often the result of a positive feedback loop of two or three problems. Not only is the household subjected to multiple problems, but these problems feed off one another and act in synergy, “the interaction of two or more agents to produce a combined effect greater than the sum of their separate effects.”

The key point of the Positive Feedback Loops: The Other Roads to Ruin post is that a financial crisis in retirement is often the result of a spending problem that causes another problem that causes a third problem, the problems become synergistic, and then the downward spiral is nearly impossible to stop. These failures can begin without warning and can bankrupt a family in a year or less.

Your retirement plan should consider all risks to your financial well-being, not just sequence of returns risk, and you should not be overconfident even if you have a low annual withdrawal rate. The four households I discussed in that post were all flying high a year before insolvency.

Positive feedback loops in retirement finance suggest the presence of chaos. In Retirement Income and Chaos Theory, I investigated that possibility and found that there are good reasons to believe that retirement finances are chaotic. If they are, then we will never have enough empirical data to understand the underlying mechanisms. Predicting your financial future using probabilities based on past data will always be suspect and can also lead to overconfidence.

The key point of that post is not that our financial future is totally unpredictable and planning is useless, or to measure how much more random market returns are than we believe. The take-away is that, while simulations and forecasts are useful when our finances are in equilibrium (which is most of the time), there will be times when our finances can behave wildly outside expectations. We need to plan for that.

We should plan for the most likely outcomes but try to create backup plans for “worst-case” scenarios. Like a floor, for instance, made up of assets that are not exposed to the equity markets and are protected from creditors.

In the post entitled, “Why Retirees Go Broke”, I combined the findings of an outstanding paper from Dr. Deborah Thorne with a paper my son and I recently co-authored analyzing the timing of portfolio ruin to show that, while sequence of returns risk might eventually deplete your savings, it is unlikely to be a major contributor to bankruptcy.

Outliving your savings and going bankrupt are two different risks. You can outlive your savings without going bankrupt. You may not spend all of your savings if you declare bankruptcy because Social Security benefits are protected from creditors and so are some retirement account assets, especially in bankruptcy. But, poor market returns aren’t the only way to deplete your savings. A spending crisis can deplete savings even faster than market losses and can lead to bankruptcy.

(ERISA-qualified retirement accounts – 401(k)s, are a typical example – are typically protected from judgments even when bankruptcy is not declared, but protection of non-ERISA-qualified retirement accounts like IRAs is complicated. You should discuss both with an estate attorney, but you can find good explanations at and at the Strictly Business Law Blog.)

You can become insolvent even if you invest all of your savings in Treasuries and annuities because expenses will still be uncertain – owning stocks isn’t a prerequisite for financial disaster. The culprit is unpredictable expenses.

The point of the bankruptcy post is that you should not confuse mitigating portfolio ruin with mitigating insolvency. Most elder bankruptcies result from spending crises, not poor market returns. Plan for both.

Most retirement research calculates a lifetime probability of portfolio ruin, the probability that you will deplete your retirement savings sometime during your life. Spend 3% of your portfolio annually, for example, and you have only a 5% probability of outliving your savings. In reality, that probability of portfolio survival ranges from near zero in the first decade of retirement to 5% after three decades or so. Kaplan-Meier curves, as I described in Death and Ruin, show how your probability of portfolio survival changes with age.

That post also explains that unless you live past your median life expectancy, you probably won’t outlive your savings with a reasonable withdrawal rate. The biggest risk of portfolio ruin, assuming you select a reasonable annual spending rate, is longevity, not market risk.

Retirement finance may be quite different than what you’ve read. Losing your savings due to market volatility is only one risk of retirement and it isn’t the worst outcome. You aren’t likely to go broke because you spend 4% of your savings annually instead of 3%, or even to completely deplete your savings. You’re more likely to see your savings decline, then to reduce spending to avoid going broke, and finally to suffer a decline in your standard of living as a result. About half of us won’t live long enough to be exposed to sequence of returns risk, at all.

In the unlikely event that you do spend all your savings, that doesn’t mean you’ll be bankrupt. Bankruptcy is far worse. It means you have more debt than you can repay and your debts need to be reorganized. You may lose all your assets that are not protected from creditors, but you will still receive Social Security benefits and you will keep assets in retirement accounts, although traditional and Roth IRA's are only protected to an inflation-adjusted $1,000,000 (currently about $1,200,000) and this applies to all your plans combined. You won't have credit and you will probably have trouble using banks for a long time.

Sequence of returns risk will rarely be a big contributor to bankruptcy and it takes decades to erode savings. Bankruptcy will strike like a bolt out of the blue as a result of spending shocks and may cost much more than just your savings. (All four of the families I wrote about in Positive Feedback Loops lost their homes, too.) The crisis will be difficult to stop once it starts.

Some shocks are impossible to avoid, but we should plan for the ones we can. We should be aware of the ones we can’t, if for no other reason than to avoid overconfidence in our retirement planning.

Retirees should focus on a bigger picture than just spending sustainable amounts of savings and the right asset allocation.
[Tweet this]

The expense side of the retirement finance equation is at least as important as the income side and it's probably riskier. Retirees should focus on a bigger picture than just spending sustainable amounts of their savings and getting their asset allocations right.

Retirement research should, too.

In my next post, I'll review the major expense risks of retirement and suggest a few ways to deal with them.

Moshe Milevsky just published a piece entitled, "It’s Time to Retire Ruin (Probabilities)". Regular readers of this blog will recall a similar post, "Time to Retire the Probability of Ruin?", from last April, though Dr. Milevsky makes a much better argument.


  1. The tone of this post is dire, but I'm not sure. Doesn't this statement

    > Retirees typically plan for a 5% to 10% probability of
    > outliving their retirement savings, but about a half-
    > percent of Americans 65 and older file for bankruptcy,
    > and that number appears to be growing.

    mean that we're planning for more than 10 times the rate of retirement bankruptcies than we're actually seeing in the general population (5% as opposed to 0.5%)?

    1. That isn't what it means, Emily. You're equating the risk of outliving your portfolio (5%) with bankruptcy risk (0.5%), which is my main point – they aren't the same thing. Here’s an example.

      Let’s say we listed your retirement assets and they included:

      • $250,000 in equities
      • $100,000 in TIPS bonds
      • Home equity $200,000
      • Social Security benefits paying $30,000 annually
      • A fixed annuity paying $10,000 annually
      • An emergency fund of $40,000 cash

      Now, assume that your planner advised you to spend 3% of your stock and bond portfolio so you would have a 5% probability of outliving the savings in that portfolio. Then assume that you were unlucky enough to fall into the 5% and your savings portfolio is now depleted. You still have home equity, Social Security benefits, an annuity, and your emergency fund of cash, so you aren’t broke.

      Alternatively, assume that your investments were fine but you were hit with a million dollar debt from out-of-pocket medical costs or a liability claim. You would spend your savings portfolio, your emergency cash and your home equity and possibly your annuity. You would still have Social Security benefits, so you still wouldn’t be broke. In fact, with a little planning you could save some of your assets by declaring bankruptcy. But, you would be worse off than had you only lost your savings.

      The probability of needing to file bankruptcy is less than 0.5%, but the outcome is much worse than losing only your savings to market volatility.

      The 5% is the probability of losing your savings portfolio as a result of market risk. The 0.5% is the probability of needing to file bankruptcy due to a spending crisis. They’re two different problems.

      It’s a little like saying that if you’re in an automobile accident there is a 5% probability that you will be seriously injured and a 0.5% probability that you won’t survive. In the retirement example, however, it’s a 5% probability that you will lose only your savings due to market risk compared to a 0.5% chance that you will lose those savings and nearly everything else due to a spending crisis.

      Does that help?

    2. So, a 5% probability that I lose a large chunk of my money, versus a 0.5% probability that I lose absolutely everything?

    3. Roughly.

      Assuming you plan spending so that you have a 5% probability of portfolio ruin, there would be roughly a 5% probability of losing a lot of the investment portfolio from which you are spending. Whether that is a "large chunk of money" depends on how much you have saved. (For the majority of households in the U.S., that is not a large chunk of money.)

      There is about a 0.5% (0.43%, actually) lifetime probability that you would file bankruptcy. In that case, you would lose everything that is not protected from creditors, but that wouldn't literally be "everything." Bankruptcy laws ensure that you don't lose "absolutely everything."

      So, while I think your last statement accurately captures the spirit of the issue, a better statement would be "a 5% probability that I lose a large chunk of my retirement savings, versus a 0.5% probability that I lose everything not protected for creditors."

  2. Dirk, can't wait till the next post. I want to see what you identify as the major expense risks. I have tried to work on the long term care expense risk, and I assume medical is one. Fraud might be another. Can't wait to see the items you list and ways to mitigate them. Brad

    1. Brad, most of the major risks were identified in Throne's paper and listed in Why Retirees Go Broke. My next post will suggest what you can do to mitigate them.

    2. Thanks for the reference Dirk. I read all your blogs, but it sure helps to refresh. I appreciate the fact that you are a thought leader - looking beyond just typical market volatility for possibilities of ruin in retirement. I wish more planners did this. Really looking forward to your next blog with a listing of ways to mitigate spending/expense crisis in retirement. Thanks, Brad.