One upon a time, back in early 2012, David Rosenberg was a prominent bear and deflationist, while his nemesis, Wells Capital's Jim Paulsen, was one of Wall Street's biggest equity bulls. The confrontation between the two culminated with a January 2012 article explaining "What (If Anything) The Bulls Are Seeing."
Since then, Rsoenberg infamously flip-flopped to bullish (predicting incorrectly that wages would rise and TSY prices would tumble), while Jim Paulsen, almost exactly a year ago, threw in the permabull towel and warned on January 10 of 2015 that "stocks are massively overvalued."
Paulsen laid out his skepticism in the following 8 bullet points:
- First, the valuations of U.S. stocks are much higher today than widely perceived or as suggested by the valuation of the popular S&P 500 Index. Moreover, today’s valuation extreme is not limited only to a subset of stock market sectors but rather is very widespread whereby nearly all P/E multiple percentiles are at or close to post-war records. Finally, the current valuation extreme is not the result of poor performance from a single valuation metric. U.S. stocks are broadly and richly priced compared to earnings, cash flows, and book values.
- Second, because valuation dispersion is relatively low today, there are not many areas to hide from overvaluation. In 1973 or 2000, investors could reduce extraordinary valuation risk by simply diversifying away from the Nifty Fifty or new era tech stocks. Today, because values are both high and tight, lessening valuation risk may not be possible except by allocating away from U.S. stocks.
- Third, this valuation extreme has only recently materialized. Charts 2, 3, and 4 show that until 2014, although median stock valuations were relatively high, they were not at the acute or record highs they are at today. Indeed, the current excessive valuation profile is a product of this recovery cycle. The median U.S. stock began this bull market below 12 times earnings in 2009. In the last five years, however, the median P/E multiple has risen by about two-thirds to slightly more than 20 times earnings. It is important for investors to fully appreciate just how much this bull market has already elevated the valuation landscape.
- Fourth, is the current widespread valuation extreme more dangerous than a concentrated extreme simply because concentrated extremes tend to be more obvious and eye-catching? When the Nifty Fifty or dot-com stocks exploded to outlandish P/E multiples, most investors realized the stock market was getting a bit frothy. Today, even though a larger portion of the overall stock market is aggressively priced, it has not garnered nearly as much attention. A concentrated valuation extreme tends to loudly announce itself whereas a broad-based valuation extreme seems more stealth and, therefore, perhaps more dangerous.
- Fifth, how important have record low bond yields and a zero short-term interest rate throughout this recovery been in producing the contemporary broad-based stock market valuation extreme? And, how will this highly valued stock market react should U.S. interest rates soon finally start to rise? Many believe since interest rates are so low today, they could rise for some time before negatively impacting the stock market. However, what if today’s widespread extraordinary valuations actually make the stock market much more sensitive to interest rates?
- Sixth, historically when the valuation of the median NYSE stock has been as high as it is today (e.g., from Chart 2 consider 1962, 1969, 1998, 2000, 2005, and 2007), the overall stock market has usually either suffered an outright bear market (i.e., in 1962, 1969, 2000-2001, and 2007-2008) or a correction (i.e., in 1998). Only in 2005, from a similar median stock valuation, did the overall stock market avoid a correction or bear market until 2008. At a minimum, this historic record suggests investors should proceed with greater caution.
- Seventh, the current valuation profile of the U.S. stock market argues in favor of S&P-like indexation. When the stock market is characterized by a concentrated valuation extreme (like during the early 1970s Nifty Fifty or the late 1990s dotcom eras), investors are best served by avoiding indexation. Often, however, during such stock market manias, investors become frustrated by being unable to match the strong advance in the S&P 500 Index and ultimately are enticed to simply index. For example, during the late 1990s, just as valuation risk became acute among the S&P 500 stocks, more and more investors piled into S&P 500 Index funds. Today, by contrast, some exposure to indexation seems reasonable. The S&P 500 Index may possess less valuation risk than does the median U.S. stock. While the median P/E in Chart 2 is at a post-war high, the S&P 500 market-cap weighted P/E multiple is still far from an extreme.
- Eighth, overweighting international stocks may be an approach to diversify away from the widespread valuation extreme evident in the U.S. stock market. Perhaps international stock markets also are highly valued today. However, since most have significantly underperformed the U.S. stock market in recent years, international markets are far less extended on a valuation basis.
To his credit, Paulson was right in 2012 (even if for all the wrong centrally-planned reasons), and he was right again this past January: as of this moment the S&P is lower then where it was when Paulsen decided to no longer be an unconditional permabull.
However, if valuations were stretched at the beginning of 2015, they are literally off the charts now that both revenues and earnings have posted annual declines, suggesting the median market multiple is now even more stretched than it was a year ago.
So now that the 1 year anniversary of Paulsen's cautious - and accurate - forecast is almost upon us, is the Wells strategist back to bullish mode, or is he even more cautious?
The answer, according to this latest letter, depends on just one number: 5%, the so-called "full employment" threshold. Paulsen says that "not only have stock returns proved much less robust once the unemployment rate has reached 5%, but sector leadership has also undergone a fairly radical change."
In the world of fiction, the most famous threshold may be that of 88 miles per hour. In the non-fictional world of economics and finance, however, an even more important threshold is that of 5% unemployment. At that moment everything changes.
Jim Paulsen explains.
During most of the post-war era, the economy was considered to be near full employment once the unemployment rate declined to between 4% to 5%. Reaching full employment marks an important shift in the economic cycle. Once slack in the economy is no longer excessive, further economic growth begins to pressure resource prices and other business costs, inflation risk increases, interest rates typically rise, and policy offi cials begin to lessen or reverse accommodation.
Chart 1 illustrates the U.S. unemployment rate. Since 1948, the unemployment rate has been at or below 5% (i.e., at full employment) only about one-third of the time. Most of the sub-5% labor markets were either prior to the 1970s or since the mid-1990s. Many compare the contemporary era of low inflation and near-zero interest rates with the 1950s and as this Chart illustrates, a sub-5% unemployment rate during the balance of this recovery seems likely to be yet another similarity.
Since the unemployment rate recently neared 5%, wage inflation has shown signs of quickening providing the underpinnings last week for the first Fed funds rate hike in this recovery. The contemporary recovery has seemingly reached full employment and if history is any guide, stock investors should be prepared for some key changes during the rest of this bull market.
Full employment changes the stock market
Exhibit 1 illustrates the performance of the overall U.S. stock market and its individual sectors during the post-war era when the unemployment rate was above 5% compared to when full employment was reached. The data for this comparison and the sector indexes (the defi nitions of the sectors are shown in Exhibit 2) are from the extensive Kenneth R. French data library. It has market capitalization weighted indexes which include all U.S. stocks on the NYSE, AMEX, and NASDAQ since 1948.
The charts in Exhibit 1 are based on average annualized total returns derived from all monthly data since 1948 comparing returns when the unemployment rate was above 5% to returns when the unemployment rate was 5% or less. Chart 1 shows that nearly every stock sector (except energy stocks) posts significantly lower returns once full employment is reached compared to when the unemployment rate is above 5%. Indeed, the annualized return of the overall stock market is 50% lower once full employment is reached and many sectors (e.g., Chems, Shops, Durable, Other, and NoDur) produce returns which are only about one-third of what they are when the unemployment rate is above 5%.
Chart 2 illustrates that full employment brings new challenges for the stock market. Continued economic growth typically produces cost-push pressures eroding profit margins and infl ation and interest rates usually begin to rise. Although full employment does not necessarily end a bull market, as Chart 2 shows, it does tend to signifi cantly lower future stock returns.
Charts 3 and 4 illustrate that full employment also typically brings a leadership change in the stock market. Chart 4 ranks the annualized sector returns in the post-war era when the economy was at less than full employment while Chart 3 shows the ranked sector returns once full employment is reached. Several points are noteworthy. First, before the economy reaches full employment, returns across the stock market are much higher compared to when the economy is at full employment. Overall, the annualized U.S. stock market total return averaged about 15% when the economy was at less than full employment compared to only about 7% once full employment is reached.
Second, the dispersion of sector returns within the stock market widens considerably once the economy reaches full employment. Before full employment, the stock market sector return diff erential is only 4.27% (i.e., from Chart 4 the difference in annualized returns between the best performing sector NoDur and the worst sector Utils is 4.27%). However, the sector return differential widens considerably to 9.31% once the economy operates at full employment. When the unemployment rate is above 5%, the most important investment theme is to be overweight the stock market. Individual sector or stock picking is less important (because sector/individual stock return dispersion is low) compared to the overall asset allocation decision. However, once at full employment, sector or stock picking prowess becomes much more important. Return dispersion tends to widen improving the potential to add signifi cant alpha from sector/individual stock selections.
Finally, the stock market has usually experienced a significant leadership shift once the economy reaches full employment. Chart 4 shows when the economy is at less than full employment it has been led mostly by consumer sectors (e.g., NoDur, Shops, Durbl, and Hlth). Conversely, at full employment, the stock market has been led more by industrial sectors (e.g., Enrgy, BusEq, and Manuf). While the leadership change is not perfect (e.g., historically, Hlth does well in both environments), Nondur and Durbl are among the best performing at less than full employment and the worst performing at full employment. Likewise, while Energy is among the worst performing at less than full employment it is the best performer once full employment is reached.
Paulsen's conclusion: "Historically, reaching full employment has proved highly significant for stock investors. In the post-war era, once a 5% unemployment rate is reached, both the character and the performance of the U.S. stock market have been altered. While reaching full employment does not necessarily suggest an imminent end to the current bull market, it does suggest investors should anticipate significantly smaller future returns. Moreover, it has often led to much wider return dispersions favoring stock pickers over asset allocators, and finally, it has typically produced a major leadership change away from consumer toward industrial sectors."
Will he be right? The only way to be sure is if Rosenberg issues another lenghty missive dismissing everything said above.