Bull markets tend to climb a wall of worry, and the 100% gain in the MSCI World Index of stocks since bottoming in early March 2009 has come amidst a plethora of very credible worries. In this cycle, defined as March 2009 through February 2013, the typical economic concerns were compounded by troubles uniquely tied to the depth of the global financial crisis. In addition, coming within eight years of the technology stock bubble bursting, investor risk appetite has been shattered. Skepticism abounds about the risk of central bank money printing and market manipulation this cycle, a risk that we take seriously. There has also been widespread criticism of governmental competency, from the structure of the European Monetary Union to the effectiveness of U.S. policy making. Frequently lost in the discussion, however, is the significant rebound in the performance of the corporate sector whether in managing their balance sheets or increasing their profitability. In fact, we believe that the strong advance we have seen in equity prices over the last several years is justified by corporate profits (as seen in Exhibit 1 below) and we believe that stocks today still present a reasonable absolute value. The attractiveness of the valuations only increase when you compare them against the counterpart valuations in the fixed income markets. We also see any upward movement in interest rates as being gradual in timing with overall levels of interest rates staying relatively low for several years to come.
EXHIBIT 1: PROFITS HAVE UNDERPINNED STOCKS
Source: Northern Trust, Bloomberg. Data through 12/31/2012.
Much of the liquidity created in recent years has sought out lower risk, yield-oriented returns through investing in fixed income suppressing risk-free yields and credit spreads. This has reduced the cushion around future rate hikes, but the talk of a bond bubble is mischaracterized in our minds as the bonds of credit worthy issuers are still likely to pay par value at maturity. This limits the downside risk to bond investing, in contrast to the possibility of complete capital loss in technology stocks in 2000 to 2001 or housing/mortgage related investments in 2008 to 2009. Central banks, likely led by the Federal Reserve (the Fed), will set the pace of interest rate normalization, as long as inflation is quiescent. This is another distinction from the last two cycles, where the risk of adjustment was much greater as the owners of the inflated assets were diffused. A slow but steady increase in rates (such as those made by the Fed in the 2004 to 2005 time period) will be much less disruptive than a heavy-handed approach (circa the Feds approach between 1994 and 1995). We think this does change the reason to own bonds going forward, moving them from a yield-producing asset to one primarily providing liquidity.
One of the markets concerns about over-abundant central bank liquidity is the resulting impact on asset prices and valuations. This is somewhat ironic in that the increasing asset prices have been an overt policy objective of central bankers, as a re-inflation of asset prices has been used to generate a positive wealth effect to stimulate confidence and growth. Due to the lingering concerns about the health of the economy and financial markets, investor flows and investor interest have been more directed toward fixed income assets than equities over the last several years. Despite the strong run in equity prices from 2009, the major markets are still trading below long-term median valuations as shown in Exhibit 2.
EXHIBIT 2: LEAVING EQUITY VALUATIONS IN GOOD SHAPE
Source: Northern Trust, Sanford Bernstein. Based on trailing net income before extraordinary items. Data through 2/28/2013.
U.S. valuations have climbed from 12 times earnings in September 2011 to 15 times today, still a discount to the historical median (since 1955) of 16.8 times. In Europe, valuations have climbed from 10.5 times earnings over the same time period to 13.7 times today, approaching the median of 13.9 times since 1970. Finally, Asian valuations have increased from just more than 12 times to current levels of 16.9 times, below the median of 22 times (inflated by Japan). While these data support current equity market valuations, market skeptics will cite the Schiller Cyclically Adjusted P/E level of 22.9 times on U.S. equities as evidence of market overvaluation. Using a measure of trailing 10-years of earnings captures two of the worst earnings recessions in the last century and assumes they are representative of what may occur in the future. We believe this significantly overstates the risk to the U.S. economy, inflates the valuation levels of stocks today and we conclude that current valuations are reasonable. The current discount to historical medians is justified by the continued intervention of global central banks, raising the risk of market dislocations when they start to change policy.
Turning to fixed income valuations, the distorting impact of the financial crisis (slow growth, low inflation and investors risk aversion) and central bank intervention remain high. Historically, 10-year government bond yields have generally tracked nominal Gross Domestic Product (GDP), as can be seen in Exhibit 3 showing U.S. bond yields. Because of the unusual post-financial crisis environment and Federal Reserve policy actions, bond yields are currently below this rule of thumb as the 10-Year U.S. Treasury sits at 1.9% while nominal GDP in 2012 was 3.3%. As a result of the search for yield, spreads across various asset classes have contracted with investment grade spreads at 1.27%, high yield at 4.73% and emerging markets at 3.07% (as of February 28, 2013). While nominal yields for these bonds are at historically low levels, this is mostly due to the effect of low risk-free rates. Investment-grade bonds spreads, at nearly 1.3%, are well above levels that persisted for most of the non-recessionary periods of the past two decades. High yield spreads of 4.73% are also above much of the non-recessionary levels of the last two decades, but high yield has tended to widen as a precursor to recessions and has taken longer to tighten afterward. Finally, emerging market bond spreads are also being heavily affected by the global flight toward yield, reducing those spreads to just over 3%. While credit spreads have contracted significantly over the last several years, we dont expect much widening over the next year or two as we see steady economic growth, strong corporate profitability and continued investor appetite for yield.
EXHIBIT 3: BOND YIELDS MORE DISTORTED
Source: Northern Trust, Bloomberg, Haver Analytics. Data through 12/31/2012.
The hunt for yield has left bonds with a smaller cushion for rising interest rates going forward, however, as shown in Exhibit 4. In 1993, before 1994s jump in interest rates, bonds could absorb a much larger increase in interest rates (and the negative impact on price) because of the higher level of income generated. As such, high yield bonds for instance could see an increase in rates of more than 200 basis points before the price reduction ate through the entire annual yield; investment-grade bonds had a shorter leash, ranging anywhere between 30 and 120 basis points depending on duration.
The risks are more asymmetric for investment-grade bonds today. While high yield can still withstand a 100-plus basis point back up in yields, a quarter-point increase in rates puts investment-grade bonds into the red (at least temporarily). It is important to note that bonds held to maturity will not suffer an outright loss unless those securities default (traditionally only a concern for high yield, where default rates have historically averaged 2.75%). Even during the bond bear market of 1993 and 1994, the total return of investment-grade bonds was -5.0% while high-yield bonds generated a positive return of 2.0%.
While the compression of yields has been tied to risk aversion and surplus global savings, the worries about bonds being held in weak hands that are ready to sell at the first sign of trouble seems overstated. The U.S. household sectors holdings of credit have been falling as a percentage of assets over the last several years (as stocks have rallied) and the overall commitment is comparable to levels of two decades ago. Conversely, the percentage of U.S. Treasuries in strong hands (defined as owned by government players driven more by policy than economics) has risen from 20% to 50% over the same two-decade period.
EXHIBIT 4: LEAVING LESS CUSHION
Source: Northern Trust, Barclays Live. Note: 10/31/1993 date selected given the subsequent increase in interest rates.
So what are the likely drivers for a back-up in rates that could lead to falling bond prices? Developed central banks clearly control the short end of the yield curve, while they influence the long end through open market (such as quantitative easing) and open mouth (setting expectations) policy. The most likely catalyst for a noticeable change in policy is a positive upward surprise in growth, which could rekindle concerns over inflation. With the Federal Reserve likely at the front of the line for an eventual tightening in policy, due to the potential for the U.S. to be the strongest recovering major developed economy, we take the Fed at face value when they espouse their focus on improvement in the labor markets as the key metric.
Our base case scenario remains one of subpar global growth, resulting in restrained inflation and only moderate upward pressure on interest rates. Of the major developed economies, U.S. growth looks best to us, but we only expect growth of 2.1% over the next five years. We forecast Japanese growth at 1.5% and European Union growth at just 1.0%. While economic growth likely will not pressure inflation, continued easy monetary policy could lead to increasing inflation expectations. We expect inflation to average just under 2% in developed markets outside the United States, while registering a rate of around 2.3% in the United States. In this environment, we dont expect developed economy central banks to lose control of inflation, and subsequently have to suddenly reverse course on monetary policy. Our forecast for unemployment in the United States calls for a rate of 7.3% at the end of 2013, which might be sufficient progress for the Fed to begin slowing its monthly purchases of $85 billion in Treasuries and mortgage backed securities. However, the Congressional Budget Office (CBO) doesnt expect the unemployment rate, under its baseline forecast, to reach 7.6% until the end of 2014. This makes the Feds threshold of a 6.5% unemployment rate for changing its ultra-low interest rate policy a seemingly distant goal one that the CBO pessimistically doesnt expect to cross until the first quarter of 2016. So if the Fed moves off its pledge to keep interest rates at current low levels until mid-2015, the probabilities may lean toward extending the deadline further in the future.
EXHIBIT 5: CENTRAL BANKS CONTROL THE DESTINY
Source: Northern Trust, Bloomberg.
How have the stock and bond markets historically dealt with changes in interest rates? On a short-term basis, it really depends on the speed of action. As shown in Exhibit 6, the Fed implemented an incremental tightening policy in mid-2004. The S&P 500 churned at first but then began a steady rise upward, appreciating by an annual rate of 7.9% during this period positive performance that continued until the financial crisis. The gradual approach also allowed for steady returns in the bond market as the 10-year U.S. Treasury generated a positive return after some consolidation. In the early 1990s, in contrast, the Fed was more strong-handed, hiking interest rates more than the customary 25 basis points four times. The S&P 500 responded with a more sporadic return profile. During this spate of rate hikes, the S&P 500 remained barely positive until the markets sensed the end of the tightening campaign and shot upward starting in September 1994 starting the technology stock rally. Bond market returns during this period were less constructive, with increasing interest rates pushing the 10-year U.S. Treasury cumulative total return negative for most of the period.
EXHIBIT 6: AND DICTATE MARKET RESPONSE
Source: Northern Trust, Bloomberg.
If one steps back and looks at a longer-term picture, as shown in Exhibit 7, changes in long-term interest rate cycles can be disruptive to financial markets as investors grapple with the implications of a change in the interest rate environment. Over the course of the last 150 years in the U.S. economy, there have been five major interest rate cycles. With three of these long cycles occurring before World War II, where the economies were much more volatile and less globalized, the predictive value of these cycles is unclear. It is our base case assumption that the end of this long downward cycle in interest rates (the fifth cycle), which began in 1981, will result in a slow upward move in interest rates that is probably not going to generate a shock comparable to the earliest cycle changes.
EXHIBIT 7: DISRUPTIVE ENDS TO LONG RATE CYCLES
Note: Dow Jones Industrials used for cycles I/II; S&P 500 for cycles III/IV. CAGR: Cumulative Annual Growth Rate.
Source: Northern Trust, BCA Research.
What will be critical is the environment in which rates are increasing. If the Fed and other developed central banks are able to start normalizing interest rates because the economy is healing and the monetary policy medicine is less necessary, equities will likely be less affected. However, if rates start rising abruptly because developed market central banks have lost control of inflation, equities will likely sell off sharply as investors worry that the central bankers have gotten behind the curve and that the solution to the inflation problem will be to push the economy into recession.
We would like to end on a final note regarding one of the major overhangs on the financial markets over the last five years the issue of government deficits and debt levels. A debate has raged amongst economists, politicians and investors about the right approach to fiscal austerity in the post-financial crisis world. We have taken the position that Europes rapid austerity was suboptimal, and that the more tempered cuts in deficits being realized in the United States were better policy. We made this judgment assessing the potential economic impact of the policies, as opposed to the political and market realities facing the respective regions. While economists have argued about the potential fiscal multiplier (what is the actual impact to economic growth from a change in government spending), the International Monetary Fund (IMF) has led both the discussion ahead of the austerity and the revisions afterward. As reported in a January 2013 IMF Working Paper, the IMF believes that most forecasters assumed a fiscal multiplier of 0.5 times before the crisis (i.e. a $1 reduction in government spending would only reduce economic growth by $0.50, as resources were freed up for the private sector). However, as several European economies struggled mightily under their austerity plans, it became clear that the fiscal multiplier forecasts might be erroneous in the current environment of near-zero interest rates and high slack. The IMFs Working Paper posits that the multipliers may have been too low by about 1 times, meaning that instead of a $0.50 reduction in growth the actual impact was a $1.50 drop in growth.
EXHIBIT 8: KEYNES MAY BE VINDICATED YET
Source: Northern Trust, International Monetary Fund.
As a result of this underestimation of the fiscal multiplier, forecasters significantly underestimated the negative impact of fiscal cutbacks on growth and unemployment. To the extent that this lengthens the adjustment period due to high unemployment and atrophying job skills, the scale of the austerity is counterproductive. As shown in Exhibit 8, fiscal deficits across Europe have fallen significantly over the last several years. Unfortunately, growth has fallen in many instances just as rapidly. In contrast, the more measured pace of austerity in the United States has been partly contributory to growth holding above 2% over the last three years. After hitting a deficit level of 10.3% of GDP in 2010, it is forecast by the IMF to hit 7.3% in 2013. A more recent forecast by the CBO, reflecting current tax policy and economic growth trends, targets a 5.3% deficit in 2013. Our conclusions are that the near-term fiscal deficit risk in the United States has been taken off the table, despite the bi-monthly debates and deadlines ongoing in Washington. The risk has now been morphed into a longer-term entitlement risk, as the cyclical improvement in deficit reduction we are currently experiencing will run out of gas around 2018. For Europe, the focus needs to be on restoring growth through increased competitiveness. A meaningful easing of austerity measures seems unlikely, especially ahead of the September 2013 federal elections in Germany.
Special thanks to Mark Sodergren for valuation data research and to Michael Baer for report assistance.