Tag Archives: Retirement Savings

Is Your Target-Date Fund Too Risky?

If you’re planning to retire in or around 2020, and you have most or all of your assets in a target date fund, is that fund too risky? It might be given current stock market valuations.

I recently published an article on how various allocations served a hypothetical investor retiring in 2000. Any backtest begun that year would admittedly be unflattering to stock exposure, but retirees must think in worst-case scenarios because they are at risk of running out of money. And stocks may not be much cheaper now than they were in 2000.

In that article, I used the following chart to show how each hypothetical portfolio performed using the so-called 4% retirement rule, whereby the retiree withdraws 4% from his account in the first year of retirement, and boosts whatever the dollar value of that initial withdrawal by 4% each year thereafter.

It turned out that a pure stock portfolio couldn’t withstand the 4% rule given the amount of declines in two bear markets – from 2000 through 2002 and in the 2008-early 2009 period. The original $500,000 would have declined to a little more than $100,000 in the 18 year period. A balanced portfolio did much better; it would be down to a little more than $400,000. A still more conservative portfolio – 30% stocks and 70% bonds – would have remained intact. In other words, the more bonds a portfolio had, the better it held up despite the fact that stocks outperformed bonds on a compounded annualized basis – 5.4% for stocks  versus 5.1% for bonds.

And now most 2020 retirement funds have more than 50% stock exposure, potentially setting up their investors for a bumpy ride and loss of capital. On our list of some of the largest funds with 2020 dates, only the American Funds offering and the JP Morgan entry have less than 50% stock exposure.

Stocks reached a Shiller PE (price relative to the past decade’s average real earnings) of 44 in 1999, and they are at 32 now. The 44 reading seems far away, but, besides that extravagant reading during the technology bubble, the metric has been over 30 only one other time – in 1929.

Moreover, the median stock, on a variety of valuation metrics, is more expensive now than it was in 2000. For example, GMO’s James Montier recently showed that the median price/sales ratio is higher now than it has been in any other time in history. During the technology craze, small cap value stocks and REITs, for example, were left for dead, and investors prowling for cheap stocks could buy them, and wait. They wound up delivering boffo returns for the next decade. From 2000 through 2009, when the S&P 500 Index delivered no return, the Russell 2000 Value Index delivered an 8.3% annualized return to investors. But now there are arguably no cheap parts of the market.

Valuation metrics aren’t crash predictors; they don’t tell you a crash will occur next week, next month, or next year. But it’s reasonable to anticipate that the higher valuation metrics go, the more likely a significant decline or significant volatility become. And big declines hurt retires withdrawing from their accounts dramatically.

It’s also true that foreign stocks are cheaper, but they’re not that cheap. GMO’s most recent asset class return forecast shows no region of the world is poised to deliver inflation-beating returns. That means target date funds may be putting their client assets unnecessarily at risk. In 2013, Jack Bogle argued that target date funds were too heavily weighted in bonds, potentially crimping investor returns. With a Shiller PE above thirty and bond yields creeping up, the opposite might be the case now.

Financial planner and author Michael Kitces has shown that the Shiller PE works well as a financial planning tool, indicating what future returns stocks might deliver over intermediate time frames — around 8-18 years. Though a bit short on details, Kitces argues that the metric can help retirees facing sequence of return risk by encouraging them to adjust their spending. But it’s unclear why the metric can’t influence gentle portfolio modifications as well. When the Shiller PE is over 30, the likelihood of robust returns — or even returns that can beat bonds, despite low yields — is diminished after all. Nobody should ditch all their stock exposure; markets can always surprise investors. But retirees face such a harsh outcome if their portfolios suffer big declines during the first decade of retirement that modest stock exposure — even less than 50% stock exposure — appears the most prudent course. Unfortunately, judging from their allocations, target date funds may not be aware of the risk they’re imposing on their shareholders.

Target date funds are allocated based on investors’ distance from their spending goals. Even setting aside the difficulty of the retirement spending goal, which run over years and decades, distance from goal shouldn’t be the only consideration in answering the allocation question. Target date funds should also consider valuation and sensitivity to volatility.

Divorced From Reality: Prices & Fundamentals

There are many ways of assessing the value of the stock market. The Shiller PE (price relative to the past decade’s worth of real, average earnings) and Tobin’s Q (the value of companies’ outstanding stock and debt relative to their replacement cost) are likely the two best. That doesn’t mean those metrics are accurate crash indicators, or that one can use them profitably as trading signals. Expensive stocks can stay expensive or get more expensive, and cheap stocks can stay cheap or get cheaper for inconveniently long periods of time.

But those metrics do have a good record of forecasting future long-term (one decade or more) returns. And that’s important for financial planning and wealth management. Difficult though it is sometimes, everyone must plug in an estimated return into a formula for retirement savings. And if an advisor is plugging in a 7% or so return for a balanced portfolio currently, he or she is likely not doing their job well. Stocks will almost certainly return less than their long-term 10% annualized average for the next decade or two given a starting Shiller PE over 30. The long-term average of the metric, after all, is under 17.

Another way of looking at how expensive the market has gotten recently is to look at sales of the S&P 500 constituents and relate it to share price. Companies are always manipulating items on income statements to arrive at a particular earnings number. Recently, record numbers of companies have supported net income numbers with non-GAAP metrics. That can be legitimate sometimes. For example, depreciation on real estate is rarely commensurate with reality. But it can also be nefarious, as Vitaliy Katsenelson recently argued in criticizing Jack Welch’s stewardship of General Electric, which Katsenelson characterized as being more interested in beating quarterly earnings estimates rather than in creating long-term wealth. And that’s why sales metrics can be useful. They are less easily manipulated.

So I created a chart showing sales per share growth and price per share growth of the S&P 500 dating back to the end of 2008. From the beginning of 2009 through the end of 2016, companies in the index grew profits per share by nearly 4% annualized, a perfectly respectable number for a mature economy. But price per share grew by a whopping 14.5% over that time. Over that 8 year period, sales grew less than 50% cumulatively, while share prices tripled.

Anyone invested in stocks should worry about this chart. How do share prices get so divorced from underlying corporate sales? One likely answer is low interest rates. But there must be other reasons because we’ve had low interest rates and low stock prices before – namely in the 1940s. That was after the Great Depression, and stocks were still likely viewed as suspect investments. Today, by contrast, stocks are not viewed with much suspicion, despite the technology bubble peaking in 2000 and the housing bubble in 2008. Investors still believe in stocks as an asset class.

And yet, the decline in rates over the past four decades has been breathtaking, as has Federal Reserve intervention. James Montier of Boston asset manager, Grantham, Mayo, van Otterloo (GMO), has studied stock price movements over the past few decades and found that a significant percentage of upward price movements have occurred on or before Federal Open Market Committee (FOMC) days. Montier estimated that 25% of the market’s return since 1984 has resulted from movements around FOMC days. Moreover, the market has moved higher regardless of what the FOMC decision was.

If this is a stock market bubble – and the data shows unusually high prices relative to sales and earnings – it is a strange one. One doesn’t hear the anecdotal evidence of excitement – i.e., cab driver’s talking about their latest stock purchases, etc…. This is perhaps a kind of dour bubble, where asset ownership at any price seems prudent in an economy that is becoming less and less hospitable to ordinary workers. Or, as Montier wrote in a more recent paper, this may be a “cynical” bubble where investors know that shares are overpriced, but think they can be the first ones out when the inevitable decline begins.

The Dumbest Bet in Finance

In this past weekend’s Real Investment Advice Newsletter, I wrote about financial advisor Larry Swedroe’s excellent article on the “Four Horsemen of the Retirement Apocalypse:”

  • low stock returns,
  • low bond yields.
  • increased longevity; and,
  • higher healthcare expenses.

In his article, Swedroe mentions that high yield (junk) bonds won’t save investors, who haven’t historically been rewarded well for taking on their risk.  Swedroe also says high yield bonds correlate well with stocks, which means they don’t provide much diversification.  Swedroe writes from the point of view of modern portfolio theory, which looks for ways to increase volatility-adjusted returns in a portfolio. In this post, I’ll treat junk bonds a little differently, showing why now is a terrible time to own them. My analysis doesn’t completely contradict Swedroe’s though; it supports his thesis that stocks and junk bonds are highly correlated.

Unlike Swedroe, I don’t dislike junk bonds per se. These loans to decidedly less-than-blue-chip companies are just like any other asset class. They can be priced to deliver good returns, as they were in early 2009, or not.

Right now, they’re not.

Everyone looks at junk bonds initially by observing the starting yield or yield-to-maturity. Right now, the iShares High Yield Corporate Bond ETF (HYG) is yielding 5.53%. That can look attractive to some investors. After all, where else can you get over 5%?

Other people look at the spread to the 10-Year U.S. Treasury. 5.53% is around 2.7 percentage points more than the 2.8% yield of the 10-year U.S. Treasury. That might look find to some too. Of course, a little bit of research shows that spread is lower than the historical average of around 5.7 percentage points.

Still, investors seeking higher yield may be undisturbed by a historically low spread. Some people need the extra yield pick-up over Treasuries, however small it might be by historical standards, and that’s enough for them to make the investment.

Yield Isn’t Total Return

There’s one extra bit of analysis, however, that should make investors think again about owning junk bonds – a loss-adjusted spread. The problem high yield investors often fail to consider is that junk bonds default. And that means the yield spread over Treasuries isn’t an accurate representation of what high yield investors will make in total return over Treasuries. It’s easy to forget about defaults and total return because defaults don’t occur regularly. They tend to happen all at once, giving junk bonds a kind of cycle and encouraging complacency among yield-starved investors during calm parts of the cycle.

Default rates for junk average about 4.2% annually, according to research from Standard & Poor’s. And investors have typically recovered 41% (or lost a total of 59%) of those defaults, according to this Moody’s study from 1981 through 2008. That results in an annual loss rate for an entire portfolio of around 2.5%. So the iShares fund’s 5.53% yield isn’t quite what it seems to be. In fact, if we subtract 2.5 from 5.53, the result is 3.03, meaning investors in junk bonds are likely to make only 20 basis points more than the 2.8% they could capture in a 10-Year U.S Treasury currently.

Now, a more careful analysis should consider an “option-adjusted” spread, which accounts for the fact that issuers can call bonds prior to maturity and lenders or bondholders can sell bonds back to the issuer at prearranged dates. This adjustment usually adds something to the spread, making higher yielding bonds slightly more attractive. So we took the options adjusted spread data, and adjusted it for an annual loss rate of 2.5 percentage points. Remarkably, there have been times such as immediately before the financial crisis when investors weren’t making anything on an options-adjusted basis above Treasuries to own junk bonds. Now at least it’s around 1 percentage point.

 

Still, even with the option adjustment, one percentage point over Treasuries is still very little, especially considering that the option-adjusted spread we used compares a junk bond index with Treasuries. In other words, the 0.50% expense ratio of most junk bond ETFs isn’t factored into the equation. At a 0.50% or so yield pickup over Treasuries, investors just aren’t making enough from junk bonds to justify owning them. Also, advisors pushing junk bonds on yield-hungry clients aren’t doing much due diligence. The mark of a good advisor is one who can say “No” to a client and bear the risk that the client will go to another advisor doing less due diligence.