By Michael Finke
Baby boomers are facing retirement flush with gains from the longest sustained bull market in American history. According to the traditional 4% rule, they’ll be able to withdraw 4% of their swollen 401(k) balance each year, increase their annual spending to match the rate of inflation, and be reasonably certain that they won’t ever run out of money.
The low yield on bond investments today is creating a retirement income planning challenge. If retirees follow a safety-first strategy and buy a ladder of safe inflation-adjusted government bonds (TIPS) with an average yield of 50 basis points above inflation, they would indeed be able to safely spend 4% of their portfolio for about 27 years. Since about half of higher-earning couples will have one spouse live beyond the age of 92, this means that this half will achieve absolute retirement income certainty if they retire at age 65. That is, the certainty that they’ll run out of money in retirement.
OK, so retirees shouldn’t just invest in the safest government bonds. They should invest in a mix of corporate bonds and equities that will provide the extra return they’ll need to fund spending over a much longer period of time. Since we don’t know exactly what stock and bond returns will be in the future (unlike TIPS), we can see how they did in the past and project these returns in order to make us feel more comfortable that we’ve saved enough for retirement.
Many financial advisors rely on Monte Carlo projections that use historical asset return data to provide a reasonable estimate of retirement income adequacy. Even many large financial services companies use these same data to estimate the amount of savings a plan participant needs to safely fund a retirement income goal.
Unfortunately, there is a small problem with the assumption that past is prologue when it comes to asset returns. That’s because today’s retiree faces a market for stocks and bonds that doesn’t look like any that existed in the past. Stocks and bonds are expensive. And expensive assets make for an expensive retirement. It may be time to rethink some of our assumptions about the amount of money we need to support a comfortable retirement.
Asset Prices Matter
The reason that the retiree investing in TIPS ran out of money after only 27 years is because TIPS rates are abysmally low. As I write this, TIPS with a duration of 10 years are fighting to reach 50 basis points of real return, and even the longest duration TIPS aren’t hitting 1% above inflation.
There have been a few other periods in the United States when real, after-inflation bond yields were this low or even negative. The Great Depression, World Wars I and II, and the late 1970s. Each of these periods has one thing in common—scared investors who didn’t want to buy stocks. Such a flight to safety shifted assets from risky to low-risk investments, pushing up the price of bonds and driving down the price of stocks. In each of these periods, the stock market price-to-earnings (P/E) ratio fell below 10. In other words, businesses were valued at, say, $8 in stock price for each $1 of profits.
Low stock prices ended up bailing out even a Depression-era retiree invested in a 50/50 portfolio over a 30-year retirement. Stock prices rebounded in the mid-20th century and retirees never ran out of money if they followed the simple 4% rule (and dutifully passed away before their 31st year). But what if stocks didn’t rebound?
The difference in future performance between a stock market priced at 10 times earnings and one priced at 28 times earnings is, historically, pretty significant. In order for an expensive stock market to sustain the kind of returns witnessed in the 20th century in the U.S., one of two things would need to happen. Either the growth rates of profits would need to rise faster than they had in the past, or interest rates would have to fall even lower on bonds pushing P/E ratios to perhaps 35 or 40.
The growth rate story is a tough one to make. Historically, even when the economy grew at a much faster rate than today, the subsequent 10-year return on a portfolio when stocks were valued as high as they are today looked a lot like the rates of return investors are getting on TIPS. According to calculations from David Blanchett, head of retirement research at Morningstar, the historical 10-year annualized return on stocks when priced at current levels has been more than 300 basis points below the long-run average.
Indeed, the relation between 10-year stock returns and beginning valuations is nearly linear. As the prices of stocks fall to below a P/E ratio of 10, as they did in the past when bond yields fell to near zero, the future 10-year annualized returns were more than 300 basis points above the historical average. Could it be that there is such a thing as a demand curve for stocks? When prices go up, the required rate of return demanded by investors goes down.
Moreover, it may not be entirely coincidental that the largest and wealthiest cohort of baby boom retirees is currently bidding on the financial assets they’ll need to fund retirement. If that demand shifts to the right, both the risk-free rate will go down (which we can see in TIPS) and the equity premium will fall. Both of these show up in the high P/E ratio.
Bond returns are even more predictable. Unlike a stock portfolio, bond portfolios actually lose value when their yield rises. So if yields fall further, bond prices rise slightly but yields remain low. If yields rise, the reinvested bonds will now earn a higher rate but the longer-duration bonds fall in value. Even if one has carefully laddered a bond portfolio, today’s bond rates predict more than 90% of the variation in 10-year bond portfolio returns. Low rates today mean low rates over the first 10 years of a retirement.
And why are the first 10 years so important? Monte Carlo simulations of retirement sustainability over a 30-year time horizon show that portfolio performance during the first 10 years is actually more important than portfolio returns over the next 20 years when predicting failure rates. If your portfolio doesn’t perform anywhere near the historical average over the first 10 years of retirement, then failure rates rise to uncomfortably high levels.
Better Monte Carlo
In three papers published last year by David Blanchett, Wade Pfau and me, we do our best to use the right sets of historical returns in order to predict the risk of following a 4% rule over a 30-year retirement. Instead of just assuming that future rates of return on bonds and stocks will look like historical returns, we instead use models that incorporate today’s stock and bond valuations and then project future returns from that higher-priced starting point.
Think of it this way. We know with almost perfect certainty what bond yields will be over the first 10 years of retirement by looking at bond yields today. Stocks are much less certain since today’s stock prices only predict about 25% of the variation in stock return over the next 10 years. But even though annual returns are random, today’s valuations give us a much more accurate idea of what the average will be. If I left my drunk cousin in the kitchen, he may wander into the bedroom or the living room. If I last saw him in the front yard, then he’s probably going to be in the street or on my front porch. But chances are slim that he’s going to wander into the street if I left him in the kitchen.
In order to estimate failure rates, we project the future returns on a bond portfolio based on the historical average and dispersion of bond returns when they start at today’s yields. Then we project future returns from a stock portfolio based on today’s valuations. While these projections won’t be perfect, and indeed we simulate a dispersion of potential outcomes based on historical portfolio variation, they will probably be much more accurate than a Monte Carlo analysis that assumes that we’re starting from the kitchen when, in reality, we’re hanging out in the front yard.
The front yard analogy is appropriate for a retiree worried about getting run over by wandering out into the street. Using this simulation technique, Blanchett finds that even at higher levels of equity exposure the best retirees can hope for is a 60% probability that they’ll be able to sustain a 4% inflation-adjusted spending rate over a 30-year time horizon. As I mentioned before, with no stocks a retiree has a 100% chance of failure.
Is the news of low asset returns all bad for a retiree? Yes. There’s no sugarcoating the impact that a highly probable low-return environment will have on portfolio sustainability. And the news isn’t any better for workers who are saving for retirement in low-return assets. The only way to deal with lower returns is to either save more or to retire later. This may mean some difficult conversations with clients about whether their ideal retirement age is a reasonable match with their expected lifestyle.
The good news is that retirees do not have to follow the 4% rule. They can spend more by partially annuitizing—especially through products that hedge against later life income risk such as deferred income annuities (DIAs). Varying spending in response to portfolio returns can reduce the chance that retirees will run out of money if they experience a bear market, and delaying Social Security can provide an important bump in inflation-adjusted income. All of these strategies are part of a more sustainable retirement income plan. A low-return environment makes exploring them even more important for clients.