The asset management business involves its fair share of travel. Mechanical delays, cancelled flights, inclement weather, hotel overbookings, and traffic snarls are just a few of the many things that can get in the way of getting to a meeting on time. But every once in a while, we get lucky—security is a breeze, the flight arrives 20 minutes early, there’s no line at the cab stand, traffic is nonexistent, and the hotel gives us a free upgrade. These rare instances are a blessed welcome.
Of course, it is not prudent to rely on good fortune, planning our itinerary on the basis of everything going right. Suppose we’re planning a very important trip—one that will determine the financial well being of our company and our employees, not just for the next few years but the decades ahead. Most of us would be ultra-conservative in building our itineraries, with contingency plans for anything that might go wrong. We’d arrive not just the night before, but the morning before. We’d FedEx our materials and bring some hard copies with us (not even relying on our computers). We would map out the route, research the traffic, and leave an ample cushion of time for all the potential pitfalls of the journey.
Unfortunately, the return assumptions built into pension and retirement plans today are analogous to our traveler assuming that everything will go right. Hope is now the bedrock of financial planning, discount rates and pension return assumptions, allowing for no disappointments along the way. In this issue we attempt to quantify the hoped-for good luck that is needed for today’s retirement assets to fully cover tomorrow’s retirement liabilities. We discover $16 trillion in assets are, in effect, counting on the plane getting to the gate an hour early, followed by a road-clearing motorcade.
We can’t predict the future with complete accuracy. As physicist Neils Bohr once quipped, “Prediction is very difficult, especially if it’s about the future.” But we can build reasonable starting points by looking at the key components of long-term asset class returns. As we outlined in February, the return for almost any asset class can be broken down into income, growth (real growth plus expected inflation), plus changing valuation multiples.1 These are the “building blocks” of return. Using this simple method and today’s yields, we get the long-term expectations (10–20 years) for stocks and bonds shown in Table 1.
Most pension funds and 401(k) calculators assume total returns in the 7–8% range, and sometimes a bit higher. And yet, stocks and bonds—the two pillars for most investor portfolios—are expected to return 5.2% and 2.5%, respectively. Indeed, the return on the classic 60/40 blend of the two is not even 4.5%. With an approximate 3% differential, we have a stark disconnect between these simple “building block” estimates and “required” return rates.2
Are the return estimates wrong? It’s a legitimate question: these return estimates shouldn’t be taken as fact. One client remarked to me many years ago that we know our forecasts are going to be wrong; we just don’t know by how much they are going to be wrong. Can the markets do better than these anemic prospects? Of course! Conversely, can they do worse? Absolutely!
Polly Anna’s Projections
Polly Anna, head of the pension plan for Global Giant Corp., uses “typical” U.S. pension fund assumptions for her required return and asset allocation assumptions. Thus, she uses an 8% required return and the average U.S. pension fund asset allocation, currently 51.9% stocks, 30.3% bonds, and 17.8% in everything else (“alternatives”).3 Because the alternatives are used to seek equity-like returns while diversifying away some of the risk, most of her peers use return assumptions for alternatives that are similar to those of stocks. Using the estimates in Table 1, Polly Anna’s 52/30/18 asset mix has a forward-looking return of only 4.7%.
Viewing a 4.7% return as unacceptable relative to her 8% required return, Polly looks to the return forecast for three principal asset classes to see if she can squeeze more return from her investments.
The only remaining assumption is “changes in valuation.” Changes in the value that the market is willing to pay for a dollar of earnings and dividends can have a huge impact on even long-term returns for equity investors.5 The market paid twice as much for each dollar of earnings or dividends just 10 years ago. Maybe we can return to those valuation levels?
Today’s 10-year cyclically adjusted P/E ratio (so called Shiller P/E) is 20. Polly calculated what stocks had done on a subsequent 10-year basis from similar P/E levels,6 and then took the 75th percentile observation, indicating a top quartile outcome from today’s level, as her optimistic projection. This works out to 9.5%, a nearly 4% percentage point premium above our baseline.7 What a relief! Top quartile stock returns from today’s valuation level can get the returns we need.
Finally, Polly likes to assume that her results will capture an “alpha.” Of course, asset managers don’t all grow up in Lake Wobegon; they’re not all above average. Alpha is a zero sum game; if we’re winning, someone else is losing. Fortunately, Polly doesn’t have to make this added leap. Top quartile outcomes for each of the three asset classes gets us to 8.0%. As Table 3 shows, with the revised assumptions shown in Table 2, she’s got what she needs. Or does she?
Table 3 illustrates that Polly can “get there” only by assuming top quartile results for stocks, bonds, and alternatives. Furthermore, all three must produce these lofty results simultaneously over the same span! What are the odds of that? Assuming these projections are representative, this works out to 25% × 25% × 25%, or about a 1.6% chance. Yikes!
This is exactly what the overwhelming majority of the U.S. retirement market—pension funds, state budgets, IRAs, 401(k)s, etc.—is not only hoping for but depending upon. That’s $16 trillion of assets expecting a decade of sunshine to achieve the 7–8% targeted returns used for planning and budgeting purposes.
Targeted Returns—Expectations or Aspirations?
There is nothing inherently wrong with aspiring to a certain rate of return, the same way there is nothing wrong with anticipating an early flight. The disconnect occurs when we rely upon high returns or an early flight. We’re staking huge financial resources on a high return assumption, despite overwhelming evidence to the contrary. The only way today’s expected returns can match tomorrow’s targeted returns is through remarkable good fortune in the years ahead.
We’re relying on hope. But hope is not a strategy; hope will not fund secure retirements. We’re planning for the best and denying that worse can happen. It makes far more sense to hope for the best, with plans for realistic outcomes—and contingency plans for worse ones.
While we think it folly to depend upon lofty targeted returns, we can still aspire to earn more. Next month, we’ll do exactly that, as we share a sensible roadmap to attaining higher returns.
*In the 1990s, there was a book on marketing by this title. The title can apply, just as aptly, to much of the financial services industry.
Learn More About the Author
1. See “Lessons from the ‘Naughties,’” Fundamentals, February 2010.
2. In a 2005 “Editor’s Corner,” Rob Arnott suggested that pensions should compute their liabilities using the Treasury yield curve because the liabilities could be fully immunized at this rate. This idea was—needless to say—not eagerly embraced, though a few funds now do this informally so that they can be aware of their downside. See “The Pension Problem: On Demographic Time Bombs and Odious Debt,”Financial Analysts Journal, November/December 2005.
3. See “Data & Directories,” Pensions & Investments, December 28, 2009. http://www.pionline.com/article/20091228/CHART2/912239982/1044/DataBook.
4. See “The '3-D' Hurricane Force Headwind,” Fundamentals, November 2009.
5. As an example, P/E expansion and contraction has contributed on a per annum basis between –7% and +9% to S&P 500 Index returns over the past five decades.
6. 18–22 P/E.
7. Incidentally, the only time periods that witnessed double digit stock returns for 10 years where starting valuations were in today’s range was from 1992–1995 to 2002–2005 where the P/E ratio ended about 25% higher in the 25 range. This positive outlier will skew stock market returns for the years to come!
8. We used the Fund of Funds Composite (which only dates back to 1990) as it is less likely to be affected by survivorship bias. But even this is highly optimistic. The best 10-year returns were generated when the “industry” was operating on a much smaller asset base than the current $1.7 trillion.