Perhaps you’ve heard this saying about investing: “If you’ve won the game, stop playing.” It’s a famous quote from Bill Bernstein. Bernstein was saying that if you’ve accumulated enough wealth to safely make it through retirement, then don’t tempt fate by continuing to invest in risky assets. Think of it as the retirement version of premature sports celebrations.
It’s a sobering thought. The night after you confidently toast to your retirement success, your stocks could take an unprecedented tumble. So, why not avoid that potential nightmare and convert those risky stocks into something “safer” like government bonds or cash?
To answer this question we first need to define retirement “success”. The best way to define your own success is to prepare a well-thought-out investing plan. At a more generic level, Bernstein offers the 25x rule-of-thumb, which says that your retirement nest egg should have 25 times your annual living expenses after net annual Social Security income. The 25x rule is just the inverse of the 4% rule, which you can read more about here. Once you’ve reached the 25x threshold, you’ve won the game and can stop playing.
But Bernstein notes that successful retirement investors don’t have to stop playing the game entirely. He suggests keeping that 25x nest egg in safe assets like bonds and cash, and then betting on stocks or other risky assets with any money accumulated above that threshold. A related idea is the conventional advice to gradually glide down your portfolio from mostly stocks to mostly bonds as you approach and move through retirement. You can either do this yourself or use “Target Date Funds” that automatically perform this “glide path” function for you as shown in this graph from Morningstar.
Complicating Success
The often unspoken problem with the 25x rule, and similar rules-of-thumb, is that they assume we can accurately estimate our future expenses in retirement. Such estimates usually start with an existing pre-retirement budget and then adjust for various retirement factors that may either increase spending (e.g., more travel and health care costs) or decrease spending (e.g., lower housing costs and a sitting-on-the-porch lifestyle). And there are many ways to account for all these relatively foreseeable spending uncertainties.
But here’s the really unspoken part. What about the unforeseeable uncertainties? Borrowing from NASA, Donald Rumsfeld famously observed that it’s the future “unknown unknowns” that are the most difficult to address.¹ Here’s a list of some things that you probably don’t expect to happen and haven’t baked into your investing and retirement plan:
- You get divorced in retirement, and your assets are suddenly halved.
- You commit a crime (unwittingly of course) and incur legal costs or fines.
- You get in a serious accident and don’t have a sufficient umbrella policy.
- You have serious health issues and choose to pursue uninsured therapies.
- Your children have these types of unexpected expenses and ask for your financial help.
- The government starts taxing wealth like they do in Scandinavia.
- Inflation stays at 8% or greater for a decade or more.
You probably haven’t considered these sorts of financial disasters because you regard them as implausible. But many of them are surprisingly possible. For some examples, consider that divorce rates are highest for those in the early retirement window from ages 55 to 64, with 43% of those marriages ending in divorce. Similarly, only 10% of homeowners have an umbrella policy. And prolonged high inflation is not unprecedented. Inflation stayed above 8% from about 1973 to 1983.
In a post two years ago, I estimated a roughly 80% chance of one or more major financial setbacks occurring throughout a typical 60-year span as an adult. So, even with proportional adjustment for a 30-year retirement period, there’s still a roughly 40% chance of at least one major setback during retirement.
More recently, I’ve posted about the surprise of and financial damage from my divorce during early retirement. That’s one reason I’ve started thinking more about how financial surprises can upset the craftiest of well-crafted retirement plans.
Because a mindful investing plan focuses on long-term stock investing even deep into retirement, I’ve continued to “play the game” well past the 25x threshold for many years now. And now that my assets are halved, I’m so glad that I kept playing. Before my divorce, we had reached the point where we didn’t concern ourselves with annual budgets, and we basically spent money when and where we desired. I never thought I’d spend $35,000 on a rather extravagant two-week family vacation, but we did just that a few years ago without much fretting.
Since my divorce, I’m reverting to a comfortable yet less free-spending lifestyle. But if I had stopped playing the game 5 or 10 years ago, I’m not sure I’d be able to say the “comfortable” part.
Balancing Risks
The point of not playing is to avoid stock volatility or a poor sequence of returns from eroding your wealth too fast during retirement. Beyond the risks of unexpected setbacks, Bernstein’s advice seems to also downplay the other well-known risks of holding only bonds and cash in retirement.
In the past, I’ve explained the risks of bonds and cash, as compared to stocks, through the concept of permanent-loss risk as summarized in this graph.
Historical return data suggest that the longer you hold stocks, the less risk you have of losing inflation-adjusted money in your investment. In contrast, bonds and cash generally produce lower nominal returns, and these low returns often fail to keep pace with inflation, which erodes the purchasing power of your nest egg over time. So, over two or three decades of retirement, stocks that aren’t cashed in to fund living expenses become progressively unlikely to lose money, while the 35% to 40% chance of losing money in bonds or cash stays about the same, no matter how long you invest.
To Stop Playing Is Also Risky
To illustrate the risks of low returns, let’s consider what happens over time if you keep playing the game vs. not playing. I could do all sorts of fancy Monte Carlo simulations using historical return data for this comparison. But I think the likely outcomes of playing vs. not playing in retirement are pretty evident with just some simple constant growth projections.
Here’s a graph of a $1M stock nest egg that produces a constant historical average return for stocks of 9.4% per year as compared to a $1M bond nest egg that produces a constant historical average return of 4.8% per year. Both cases include annual withdrawals that start at 4% of $1M and are inflation-adjusted thereafter at the historical average of 3% per year.
The stock nest egg grows over time because the stock returns outpace the rate of inflation-adjusted annual withdrawals. The bond nest egg shrinks because it can’t keep pace with the withdrawals. In both cases, the retiree would still have money at the end of 30 years. But if I were the bond investor, I’d be chewing my nails towards the end.
And what happens if we add in a financial setback like long-term high inflation? Here’s the same comparison but using an 8% rate of annual inflation.
The stock investor still manages to make it through 30 years, but the bond investor’s nest egg peters out after only 18 years. Granted, 8% inflation for 30 straight years would be entirely unprecedented, but it still shows how surprises can be fatal if you stop playing the game for an entire retirement.
I’m sure some readers are calling foul at this point because I used constant growth to make stark and deterministic contrasts between stocks and bonds. But I assure you, if you compared the outcomes of a hundred random historical 30-year periods of stock returns versus bond returns, you would get largely the same answers but in probabilistic terms. That is, only in a small minority of random periods would the bond nest egg last longer than the stock nest egg. But I agree that bonds would occasionally win the race.
Low Expected Returns
To further highlight the risks of not playing consider that most forecasts of future returns for stocks and bonds over the next decade are far lower than the historical averages I used above. The consensus estimate among professional investors is that nominal annual returns will be more like 5% for stocks and 1.5% for bonds.² And if high inflation persists for a few years, that will further and substantially erode the inflation-adjusted returns of both stocks and bonds.
If it’s hard for a bond retirement nest egg to last 30 years with a 4.8% nominal annual return, imagine how hard it would be with a mere 1.5% nominal return. Again, a decade of low bond returns does not equate to 30 years of subpar bond returns, but it’s a tough way to start your retirement.
Conclusions
All this highlights why conventional retirement portfolios involve a balance of stocks and bonds. The bonds help guard against the risks of stock volatility while the stocks help guard against the risk of relatively anemic bond returns. While I still believe holding a lot of bonds in retirement is a bad idea, it at least attempts to balance multiple risks.
And that’s why I find the idea of not playing the game a bit dangerous, particularly for those would-be retirees who are still below or right around their 25x thresholds. For these folks, to stop playing ignores the risk of unexpected financial setbacks during retirement and ignores the very possible risks of poor long-term bond returns.
It boils down to the fact that we can’t know for sure when we’ve won the game because we can’t predict the future exactly either in terms of unexpected setbacks or likely returns. The fear of spiking the ball before we reach the end zone is valid, but it also seems too narrowly focused. Even if we’ve amassed a huge lead in the game, it seems equally risky to send the entire offense off to the locker rooms at end of the third quarter.
1 – Rumsfeld was speaking during the early days of the U.S. invasion of Iraq. And given the mountain of unforeseen regional disasters that have played out since then, he couldn’t have been any more prophetic. The state of the Middle East today is a perfect example of how even wise precautions about the “unknown unknowns” don’t help you avoid the worst dangers.
2 – This is based on my 2021 update of expected return forecasts from around the web. I plan to update these forecasts in late September or early October like I do every year. Because bond yields have risen over the last year, I’m expecting the nominal returns forecasts for intermediate bonds will be more like 3%. That’s better, but still not great. And because the stock market has mostly headed down or sideways this year, I also expect that forecasts for large-cap U.S. stocks may go up slightly to around 6%.
