First Quarter 2014 Investment Commentary
Global markets experienced an up and down quarter but ended mostly positive despite developments that include Russia’s annexation of Crimea, further evidence that China’s growth is slowing amidst government efforts to manage a potential credit bubble, the changeover in Federal Reserve leadership, and a general continuation of the slow economic recovery in the United States and Europe. Our quarter-end commentary focuses on several important changes in our outlook and some resulting shifts in our portfolios. This introduction will briefly cover some of the major market and economic themes that affected first quarter results.
After last year’s blistering pace, U.S. stocks set a more muted tone and eked out a gain for the quarter (after starting the year with a sharp loss in January). Large-cap stocks were up 1.8% while small-caps rose around 1%. Monetary policy continued as a prominent theme as Janet Yellen assumed Fed leadership at the beginning of February and presided over her first policy meeting and post-meeting press conference in March, during which she announced another $10 billion notch in the Fed’s gradual reduction of its monthly bond buying. She also indicated, as was widely anticipated, that the Fed’s decision to eventually raise interest rates would be less tied to a specific unemployment number and based more on a number of economic metrics. And while she briefly roiled markets by indicating that rates could start going up sooner than the market had foreseen (i.e., potentially mid-2015), she has since reiterated that the Fed would not be quick to eliminate its support and that markets could continue to count on accommodative policy for the foreseeable future.
In terms of U.S. economic growth, the quarter’s progress was complicated by severe winter weather that likely depressed some of the short-term indicators of the economy’s health. Overall, though, the picture remains one of modest but steady economic growth with a noteworthy rebound in housing (although some data released during the quarter suggest housing may have cooled a bit amidst its generally upward trend) alongside persistently slow-to-recover employment. In particular, long-term unemployment remains very high, part-time jobs as a percent of overall jobs is high, and wage growth has been stagnant until very recently. Toward month-end, the fourth quarter 2013 GDP number was revised slightly upward to 2.6%, which represents a decline from the very strong third quarter 2013 figure, though it falls within the range of +2% growth expected by many strategists. Overall, the positives of modest growth, only a very gradual reduction in monetary support, and a decent corporate earnings season were enough to overcome concern about the Russian situation and some of the more worrisome economic developments abroad.
Among international markets, Japan suffered a steep first quarter decline after its red-hot 2013 gain. One near-term uncertainty is the impact of the country’s new sales tax, which goes into effect at the start of April and could have potential to slow growth. European stocks rose slightly on marginally positive economic growth, though the region continues to face very low inflation (and deflation fears) and still-high unemployment. While the European Central Bank did not take material policy action in its most recent meeting, it has indicated a willingness to resort to additional measures to boost inflation if needed. For the quarter, the broad developed international benchmark gained 0.06%.
Emerging markets have been beset by ongoing concerns about economic growth alongside macroeconomic instability in countries such as Ukraine and Turkey. These issues have continued to weigh on markets, leading to small first quarter losses for emerging-markets stocks (though our actively managed emerging-markets funds were positive).
Core bonds were among the quarter’s stronger performers with the Vanguard Total Bond Market Index—our proxy for overall bond market performance—gaining 1.9%. Yields on the 10-year Treasury declined from 3% at the end of 2013 to 2.73% at quarter-end. Even as the Fed continued its tapering, bonds likely got some benefit from their relative safe-haven status in the face of geopolitical tensions in Ukraine as well as concerns over slower economic growth (primarily in emerging markets). As corporate fiscal health remained robust and bond defaults low, corporate bonds (across the credit quality spectrum from high-yield to investment-grade) were strong performers, benefiting our portfolio holdings that take on credit risk as part of their strategy. Municipal bonds were another bright spot for the market as they outperformed taxable bonds in the quarter amidst improving economic health for states and municipalities and an absence of the headline-based scares (e.g., Detroit’s default announcement, etc.) that affected last year’s results.
It is important to note distinction between the short-term news and information that seems to drive the markets day to day (and over any given quarter) and the longer-term analysis that underlies our five-year scenarios and asset-class-return models. While we pay attention to the shorter-term flow of information and its effect on markets, we aren’t apt to make adjustments to our longer-term views solely as a result of these developments. We are using this quarter’s investment commentary to discuss the longer-term analysis that informs our view of the broad economic environment and our assessment of risk assets, specifically U.S., developed international, and emerging-markets stocks, as well as emerging-markets local-currency bonds. In short, we have concluded that slightly increasing our portfolios’ equity risk exposure is warranted as both return potential has improved and our assessment of the risk of a catastrophic financial crisis has waned. We spend the bulk of this commentary discussing the changes and the analysis that led us to them.
Reassessing Global Equities
In our recent commentaries we’ve noted the significant improvement in household balance sheets (total net worth), driven by both a strong rebound in housing prices as well as a surging stock market. We have also noted many other improvements in the U.S. economy and the broader macro environment over the past year. These developments served as a catalyst for us to reassess the progress of our deleveraging thesis. In our last formal review, in 2011, we concluded that deleveraging headwinds would likely abate by 2018, having started sometime in 2008. As a result, since June 2011 we have gradually removed the deleveraging-related earnings reset we had factored into the earnings path we use in our U.S. and developed international equity models, and it was completely removed by the end of last year. However, we have continued to apply a relatively conservative valuation multiple of 15x P/E to our five-year forward U.S. earnings estimate because we believed the deleveraging process was likely to keep risk aversion relatively high. These are the key metrics that underlie our return expectations for stocks.
Upon our review over the past few months, we have concluded the following:
1. The deleveraging process is still ongoing, though it is probably progressing at a slightly faster pace than we expected. Our key metrics, household debt to disposable income and household debt to GDP, have continued to improve at a slightly faster pace than we estimated (chart 1). This is good news for the economy.
2. While median incomes have not gone anywhere in real (inflation-adjusted) terms, the key to any smooth deleveraging is that nominal incomes do not decline. On that front, the stimulative monetary and fiscal policies we have seen since the crisis have worked well so far (chart 2). And should the economic recovery continue, we would expect at least some improvement in household income growth. Moreover, household balance sheets have improved significantly as the housing and stock markets have risen. Household net worth is now higher than pre-crisis highs (chart 3). Stock-market gains can be fleeting and a correction could hit households’ net worth hard again. However, a recovery in housing could endure because affordability levels remain relatively high and housing supply in relation to demand is not excessive (charts 4 & 5).
3. Largely due to the Federal Reserve’s QE policies, which make holding cash or safe assets with very low yields extremely painful for investors, we have not seen the high levels of risk aversion we expected would occur during a deleveraging process. We did see risk aversion and its impact in 2011 and 2012 when risk-off periods occurred relatively regularly, but not as much over the past year or so: as such, valuation multiples have expanded significantly over the past two years. With the Fed’s commitment to gradually tapering QE and keeping rates low until unemployment and inflation reach levels they believe reflect a healthy economy, and most of the private sector deleveraging headwinds now behind us, we believe continued use of a 15x P/E multiple is too conservative (we discuss why in the “Revisiting Valuations” section below) and are increasing it to 17x in our base case and from 18x to 20x in our bull case.
The result of these changes is that our base case expected return, annualized, over the next five years for U.S. equities increases from less than 1% to over 3%; higher than before but still relatively unattractive. Correspondingly, our returns for developed international markets increase as well since our Europe model return is in part driven by our fundamental assumptions and return expectations for U.S. equities. These higher global equity return expectations are leading us to add to both U.S. and developed international equities (with the majority of the increase going to international). To fund this increase, we are selling our position in emerging-markets local-currency bonds (we discuss why later in the commentary). The net result of these changes is that we are slightly increasing our portfolios’ level of equity risk, though we remain below our long-term strategic target (in our balanced models). The following commentary walks through our research process and current asset class views in detail.
progression of our deleveraging thesis
Before the 2008 financial crisis, our fundamental assumptions for key asset classes, especially for equities, were dictated by the post-1950s financial-market history, in particular post-1980s. This was for a few key reasons:
- The Great Depression had a huge impact on the American psyche and made a whole generation risk averse. As a result, valuation multiples were relatively depressed for a long time, so we felt it made sense to underemphasize that earlier period.
- We believe the modern stock market era (post-1980s) reflected more rational pricing stemming from broad trends such as increased demand for stocks from pensions (following ERISA’s passage in 1970) and the rise of mutual funds.
However, when the housing bubble burst and the financial crisis hit, we believed we were dealt a different set of cards. We believed that the private sector—households and financial institutions—had reached their respective capacities to borrow and lend and that we were likely facing either an acute or chronic period of deleveraging. This led us to study past deleveraging episodes, the 1930s for the United States in particular, and we concluded that deleveraging would lead to a subpar economic recovery, remain a headwind for corporate earnings, depress sentiment, and result in relatively higher aversion to risky assets in the form of lower valuation multiples. As a result, in our base case model for equity returns, we used a 15x P/E multiple on our estimate of normalized earnings, as opposed to the 18x–20x multiples typically seen since the 1980s.
In 2011, we revisited the progression of our deleveraging thesis and came away with a few key conclusions. First, the financial sector had largely delevered, thanks in large part to the Fed’s zero interest-rate policy, which resulted in a steep yield curve and rapidly recapitalized banks, though at the expense of savers (and their consumption). Second, the household sector was healing, also in large part due to low interest rates, which eased debt-servicing burdens. But we believed households’ absolute debt levels in relation to income and GDP to be high, which exposed this sector to the risk of rising interest rates—a risk we considered prudent to consider given the U.S. government’s large fiscal deficits and rising public debt. Based on our study of prior deleveraging episodes, we estimated the headwinds from household deleveraging were likely to remain with us until 2018, by which time the deleveraging effects would wash out of the system and we’d return to normal earnings levels. We continued to apply a historically conservative 15x P/E multiple, though, because we believed a good chunk of deleveraging was yet to be completed.
Fast forward to today, where we find that the household deleveraging has progressed at a slightly faster pace than we expected. In our year-end commentary, we noted improvement in some of the key household debt metrics: the household debt to income ratio, a measure of the willingness and ability of consumers to increase their borrowing, has dropped 20% from its peak in 2007, and is now back where it was in 2003, in line with its long-term historical trend. Meanwhile, household debt service and financial obligations ratios remain at historically low levels thanks to extraordinarily low interest rates engineered by the Fed, along with modest income growth. The bottom line is we believe most of the household deleveraging is behind us. Moreover, with the recovery of the stock and housing markets, the latter carrying greater significance in our minds, household balance sheets have improved considerably. This improves households’ ability to borrow and banks’ willingness to lend, which we feel is essential for the resumption of a more “normal” economic cycle and growth.
but we have not yet achieved normal status
In investing, there is always something to worry about. However, with a long investment horizon, which we have, things generally turn out fine. So, in normal times we’d generally be optimistic about the economy and the markets. But we are living in a period without historical precedent. We are still in the midst of an unprecedented monetary policy experiment, which has resulted in the Fed having a huge balance sheet. How that unwinds, or whether or not it will even fully unwind, and how the Fed will normalize interest rates such that lenders once again have an incentive to lend to productive investments are big unknowns, and they introduce considerable uncertainty.
Uncertainty does not always have to lead to bad outcomes. The Fed’s creative monetary policy experiment (or QE) thus far has led to a surprisingly benign deleveraging process. For that we feel fortunate and thankful. As advisors, the experiment’s success thus far—and its positive effects on financial markets—has helped us better meet our clients’ return objectives. But we worry that the steps taken to make the deleveraging process benign may have unintended consequences that neither we nor anyone else, including the Fed, can fully understand or anticipate. If the Fed were effective at predicting consequences, then the excesses leading up to the 2008 crisis would have been at least mitigated if not completely avoided.
One key question we struggle with is, can the Fed unwind its huge balance sheet and normalize interest rates without major disruptions in the markets? We observe two main sets of views on this among investors and other market participants. One group, and this seems to be the majority, sees it as a huge uncertainty and an unknowable risk. The other group offers a benign view and suggests the Fed has the tools to accomplish its goals without a major impact on the markets and/or the real economy. They even say the Fed may decide not to unwind its balance sheet, and through its new inventions control the flow of excess reserves into the economy (and hence inflation) without compromising its ability to target short-term interest rates. (There is also a third group that sees the Fed’s actions as leading inevitably to another crisis, such as hyperinflation.) While it’s possible the Fed can have its cake and eat it too, it’s hard for us to gain conviction in this benign view or to base investment decisions on it. We think this is a risk we should factor into our decision making, at least until we better understand the new era of monetary policy the Fed seems to be embarking upon.
As a related point, much of the private sector deleveraging has been accomplished by shifting debt from the private to the public sector. As such, overall U.S. debt levels remain uncomfortably high. There are a few reasons why it may not be as big a concern as private sector indebtedness, at least not in the near term. First, the United States, through its privilege of being the world’s reserve currency, is still a preferred destination for the world’s savings, which are in excess supply. This gives the United States greater flexibility to manage its deleveraging and to grow out of its indebtedness. (It did so post–World War II in a relatively smooth fashion.) Second, the fiscal deficit has turned down and debt is no longer growing faster than nominal GDP growth. So, the U.S. public sector deleveraging may already be underway.
However, a sharp increase in interest rates is a big risk to the still-levered economy, and to housing in particular. The Fed has committed to keeping rates below normal for a considerable period of time and is likely willing to again step in as a buyer of mortgage securities and Treasurys if needed. The risk is that the bond vigilantes, who appear to have slept through record high U.S. fiscal deficits, QE, and the debt-ceiling fiasco, awake and penalize the United States with high interest rates if the Fed botches the unwinding of its balance sheet and/or loses control of inflation. Also, a lot of the housing appreciation has been driven by investors. We don’t know what proportion of these are long-term holders of real estate versus those that are there to make a quick buck. (Reminiscent of the pre-2008 days, we are again hearing radio commercials marketing home flipping strategies.) Certainly, the speed with which housing prices have appreciated and the frothiness we are seeing yet again in parts of California, for example, give us pause. Could we be seeing a bubble in housing so soon after it burst? It would be highly unusual to have another bubble in the same asset class so soon, but in a world of unprecedented monetary policies we cannot completely ignore this risk. Further, while we are saying private sector deleveraging is close to being over, we are not saying it’s completed yet.
We continue to see causes for concern outside the United States as well, foremost among which is how China’s massive credit and infrastructure bubble unwinds. We have been expressing our concerns related to this risk going back to at least July 2011 and it has led us to temper our allocations to emerging-markets assets. In recent months we have seen defaults from Chinese entities unable to generate sufficient cash flows to service their debts. This may just be the beginning of a credit-bubble unwind that would be hugely disruptive for global markets.
A good part of Europe continues to flirt with deflation and a credit crunch. Structural imbalances between creditor and debtor countries are far from resolved, and there remains a meaningful risk of a debt crisis stemming from the weaker peripheral countries. The banking system is undercapitalized and in need of a credible backstop such as a region-wide banking union.
To conclude, significant uncertainties and risks remain and they will continue to impact how we think about valuations.
our revised valuation assumptions
As we note above, it was our concern related to the private sector deleveraging that led us to factor the post-1930s period into our analysis, and to use a relatively conservative P/E multiple. Now that these concerns have largely abated, it makes sense to again focus more heavily on the post-1950s investing environment, in particular the post-1980s.
In making this shift, we evaluated a period in the 1960s when the United States, after embarking on a decade-plus-long period of financial repression (i.e., abnormally low interest rates) and public sector deleveraging, saw a steady increase in interest rates. If this scenario plays out, history suggests it can be a relatively benign environment for equities, at least as long as inflation expectations remain relatively tame. Of particular interest to us was the period since 1985. We somewhat discount the early 1980s because inflation expectations among investors were deeply entrenched following the stagflation period of the 1970s. We don’t expect investors to be concerned about the 1970s type of inflation over the next five years. We excluded the technology and telecom bubble of the late 1990s, and we adjusted upward the 2008-2009 recession’s trough earnings, since we do not believe GAAP earnings at the time were reflective of the true earnings power of corporate America. The results are shown in the table at right.
If we believed we were in an environment with a normal level of uncertainties and risks, then this analysis suggests we should use an 18x–19x P/E multiple in our base case scenario. However, for reasons noted above, we don’t believe we are in a normal investment environment, so we will apply a slightly lower multiple of 17x P/E to our normalized earnings estimate to derive a point estimate for the base case expected return. We think such a multiple will be in line with the interest rate and/or inflation levels we expect over our time horizon. We are aware that relying too heavily on a point estimate or one return scenario is false precision, which is why we consider a range of return expectations and economic scenarios in our portfolio construction.
Our primary framework for estimating earnings is based on estimating earnings growth relative to a long-term average earnings trendline growing at a constant rate of about 6%, in line with long-term nominal U.S. GDP growth. To the extent companies are over- or under-earning relative to this trend line, our estimate of earnings growth five years out would be lower or higher than the 6% annual growth. Currently, we believe U.S. companies are over-earning relative to their long-term potential or normalized earnings power by 10%–15%. In our base case scenario, we assume earnings will decline and revert to the long-term trend over the next five years.
In our bull case scenario, we continue to assume earnings overshoot their normalized level by 20%, which they have done on a regular basis historically. This scenario is more probable now than it was when deleveraging headwinds were stronger.
Last year, we began revisiting our bear case earnings scenario. Our bear case scenario assumed another deep recession and/or the risk of entering into debt deflation. This scenario has steadily declined in importance as the private sector has deleveraged. While we see the risk of debt deflation as unlikely at present, a typical business-cycle recession would be quite normal if it were to happen over our five-year time horizon, and we would expect to come out of it as we have in the past. In our bear case scenario, we derive earnings by assuming profit margins revert to 6%, close to post-1980s averages, and sales grow at an annual rate of 4%, which is at the lower end of the range of sales growth S&P 500 companies have achieved over longer periods. To this earnings estimate we apply a P/E multiple of 16x, the same multiple we have applied in the past to what are likely to be temporarily depressed earnings.
Our expected annualized returns for equities over the next five years now range from negative 4% to positive 11%, with base case returns of 3%–4%. Readers may observe that our range of expected returns across our macro scenarios (which we discuss later in the commentary) has narrowed relative to the recent past. This observation makes sense given that the uncertainty from deleveraging has diminished in our minds.
Revisiting Emerging Markets
A confluence of issues troubled emerging markets last year, such as the tapering of monetary stimulus in the United States and slowing economic growth in China. We first invested in emerging-markets local-currency shorter-maturity sovereign bonds in 2005. We sold this position in 2008 as part of our overall portfolio risk-reduction during the financial crisis. In 2009, we established an allocation to PIMCO Emerging Local Bond, which gave us exposure to longer-duration emerging-markets local-currency sovereign bonds, and they performed well for us until last year. In fact, we had a much higher allocation to them until 2011, when we locked in some gains by reducing them by about half in our balanced models and completely selling the position in our equity models. We maintained a relatively small position (3%–5%) in our balanced models to offset some of the risk of a dollar decline, and we funded them from a mix of U.S. and international equities because of the equity-like risk stemming from emerging-markets currencies. In our base case scenario, we continue to expect mid- to upper-single-digit returns from emerging-markets local-currency bonds, and we still view them as a hedge against the risk of a U.S. dollar decline over the very long term. However, with an increase in our U.S. and international equity return expectations, the case to fund emerging-markets local-currency bonds from equities becomes weaker. Moreover, if we continued to hold the local-currency bonds and increase our equity positions, our balanced portfolios would be more at risk of violating their 12-month loss thresholds. For this reason primarily, we are selling our position in emerging-markets local-currency bonds.
As we highlighted earlier, our concerns related to China have increased slightly, and this is also relevant to our decision on emerging-markets local-currency bonds. China’s rapid credit growth and heavy investment in projects has promoted GDP growth, but also generated massive overcapacity, guaranteeing poor economy-wide returns well into the future. Over the past year or more, investors have increasingly acknowledged this risk (so some of the risk related to China has already been priced in), and, importantly, so has the new Chinese government. While the government is trying hard to rein in credit growth, according to GaveKal and BCA Research, the shadow-banking-driven credit growth continues at a pace faster than nominal GDP growth. In addition, China is encouraging market-based, rather than state-sponsored, credit allocation. While this goal, if executed in a disciplined fashion, would be a positive for the long-term health of the Chinese economy, in the near term it stresses the system and exposes past weaknesses that were papered over with credit-driven GDP growth. Already we are seeing signs. In recent months we have seen defaults by entities unable to generate enough cash flows to service their debts. If things get ugly, China’s government could very well decide to write a check (so to speak) to bail out debt holders, and investor concerns about China may recede yet again. Or, this could be the beginning of the unwinding of a major credit bubble where huge amounts of capital went into low-returning projects. Hard to say, but we think it’s prudent to err on the side of caution, in line with our goal of managing to specific 12-month downside loss thresholds.
Finally, the turmoil in emerging markets over the past year has led us to revisit our investment thesis. Many years ago we wrote,
Emerging-markets exports have risen strongly over the past several years, helping many emerging-market countries improve their balance sheets, i.e., their current-account balances look healthier, they have reduced their exposure to more expensive dollar-denominated debt, and they have built large foreign-currency reserves that easily finance their imports and serve to ward off speculative attacks on their currencies. So, relative to the past, many emerging-market countries have less of a need to borrow and, as a result, they are less dependent on foreign capital. This is a significant change because it, in theory, gives them greater control over their monetary policies.
In hindsight, emerging-market countries in aggregate do not have as much control over their monetary policies as we had thought. In the past year, some countries have had to resort to raising interest rates at the wrong time (when their economies were slowing). Granted, most of those countries had current-account deficits, but in the case of at least one major country we didn’t see the current-account deficit as being egregious or in this and other cases we thought their large currency reserves would provide a cushion allowing them to avoid raising rates at inopportune times. Overall, rates haven’t risen a lot. Nevertheless, this dynamic of capital flows forcing perverse monetary decisions on emerging markets has been evidence contrary to our thesis. Overall, the risk that poor investor sentiment leads to capital outflows, which in turn forces emerging markets into policy decisions that contaminate their fundamentals, is higher than we’d thought. (That said, we are mindful that external pressures will ultimately, over the long term, force emerging markets to implement necessary reforms, which will bring to the fore their superior underlying fundamentals.)
The past year has also reinforced the safe-haven status of the United States and the privilege it commands as a result of being the world’s reserve currency. Emerging markets, despite their better balance sheets and better long-term fundamentals, remain susceptible to short-term capital outflows. Because these outflows can contaminate emerging-markets fundamentals, and we cannot know when sentiment will worsen and when outflows will occur, the risks we perceive with emerging-markets investing are slightly higher than they were a year ago. To be clear, we still do not believe emerging-markets fundamentals are as bad as they were in the late 1990s, though this view may not hold true if our worst fears regarding China play out. But we acknowledge that to some extent our thesis has not played out and that warrants a reassessment of our assumptions. This reassessment has not led us to change our long-term return assumptions for emerging-markets local-currency bonds. After all, emerging-markets currencies in aggregate are cheaper now than they were a year ago but we are not raising expected returns because we perceive risks to be higher now. This increase in the risks associated with emerging-markets local-currency bonds influenced our decision to unwind our position at this time.
Our reassessment of emerging-markets-related risks has resulted in changes to our emerging-markets equity assumptions. We are lowering their long-term estimated trend-line growth rate from 7% to 6% annually in nominal terms. We are giving this haircut recognizing that emerging markets are not as in control of their monetary policies or destinies as we’d thought. Secondarily, despite the increase in the size of local corporate bond markets, liquidity remains relatively poor and most emerging-markets companies continue to tap foreign-currency debt, a risk factor we noted several months ago, exposing themselves to currency fluctuations. We believe this warrants a slightly higher discount relative to U.S. equities. As a result, we are not increasing the P/E multiple we assign to our estimate of emerging markets’ normalized earnings power in line with the increase we are making to U.S. (and developed international) P/E multiples in our base case and bull case scenarios. This means we are effectively raising the emerging-markets discount we demand over U.S. equities from 15% to around 25%, which is close to the historical median. This adequately factors the slight evolution in our thinking. After this change, emerging-markets equities remain attractive relative to U.S. equities. Our base case five-year return expectations for emerging markets range from 5% to 12%, with the lower end explicitly factoring in the China-related risk. In our balanced models, we have only a slight overweighting to emerging-markets equities and, considering the return range we have for them, we think this positioning remains appropriate.
A Quick Update on our Broad Economic Scenarios
Our review of our deleveraging thesis also led us to recalibrate our broad economic scenarios. While we don’t make economic forecasts nor does our investment process depend on such forecasts, we do think about broad economic scenarios that may play out and what ranges of GDP growth, interest rates, and inflation we might expect over our investment horizon. The goal is not to be precise, especially since we know we will likely get a point estimate on any macro variable wrong. (Even full-time macroeconomists usually get it wrong.) The goal is to gain a qualitative sense of investment regimes we may experience and how best to position our client portfolios such that the odds are high they meet their risk and return objectives across a range of probable environments.
Below we list three main scenarios we think are probable based on our reassessment. We are constantly evaluating these scenarios and they are subject to change based on new information and analysis. We also consider other scenarios (or sub-scenarios) not listed below as part of our sensitivity analysis and stress testing of portfolios and individual asset class expected returns. (In the past, we have published updates on four scenarios. While we continue to evaluate numerous internal scenarios, we are publishing only three scenarios at present, as these sufficiently capture the range of returns we believe are reasonably likely.)
Base Case: Moderate economic recovery continues with no major crisis, but a normal recession is likely within the five-year time horizon. Assumes GDP growth rates and interest rates start to “normalize” toward the end of our five-year horizon. Key assumptions: inflation is around the Fed’s target of 2%–2.5%, the Fed slowly raises rates, the 10-year Treasury yield is in a range around 4% at the end of year five.
Bull Case: U.S. economic growth is above average and/or earnings end the period above the long-term trendline. Helped by stronger non-U.S. growth, releveraging of the U.S. consumer, and corporate investment spending, a self-reinforcing global growth cycle develops. Key assumptions: inflation increases but remains relatively mild largely because wage pressures remain subdued due to technology and globalization forces (so businesses continue to capture the greater share of economy-wide profits); the Fed exits its accommodative policy without major economic or market disruptions, although a normal recession within the five-year period is still possible; the 10-year Treasury yield is in a range around 6% at the end of year five.
Bear Case: The economy falls into recession for any of various reasons, such as deleveraging/deflation from Europe or China, unexpected systemic shock, Fed policy error, etc. This scenario does not assume another severe financial crisis, i.e., not a repeat of 2008–2009. Key assumptions: corporate earnings are below trend, inflation is around 1.5%, and the 10-year Treasury yield is around 2.5% at the end of year five.
A Test of our Behavioral Biases
The changes to our positions that we’ve described here are not that material in the bigger scheme of things and our balanced portfolios remain underweighted to equities after these adjustments. Still, the idea of adding to an asset that has gone up significantly over the past year (stocks) is uncomfortable for value-oriented investors like us, more so if it results from unwinding a position that has recently underperformed. Like anyone, we want to be right. But an important part of being intellectually honest is having the willingness to look at new information and acknowledge when we are wrong or when fundamentals have changed. While it’s important to be patient and stick to an investment view (so as to not get whipsawed by short-term market volatility), it’s also important to be willing to change our minds if the weight of the evidence collected is pointing in another direction. Sticking to our old views when the facts have changed would be “anchoring” and hoping we are proven to be right. It would not be disciplined investing.
As we started to debate the facts and reassess our assumptions, we were reminded of several behavioral tendencies we were at risk of falling into. We talked about them openly among our research team and doing so helped free our minds to analyze information objectively. Below we list a few behavioral biases we explicitly considered with respect to our investment decisions.
Anchoring: In order to make sense of information, we need a starting point—an anchor. In our case, the anchor was our deleveraging thesis. People have a tendency to attach significance to what may have been an arbitrary or no longer relevant starting point. For example, you purchase a stock at $100 a share. It drops to $50. You believe that the stock’s “real” value is around $100, and based on this expectation you are inclined to hang on since it “should” come back. The real value of the stock is based on its fundamentals and comparable investments, not on what its price was at some point in the past. The fact that the stock was $100 in the past probably has little or nothing to do with its current value or attractiveness as an investment.
Loss Aversion and Risk Taking: Studies show that while investors are risk averse when it comes to gains (they don’t want to give them up), they are risk seekers when it comes to losses (they’ll take big risks to avoid realizing them). In our case it would have been an aversion to locking in recent losses from our small position in emerging-markets local-currency bonds even when a few elements of our thesis changed. Researchers have documented that given a choice between a sure gain and a chance to win more, people usually opt for the sure gain. But given a choice between paying a fine and gambling a chance to avoid the fine against a chance at a greater fine, they opt for the gamble. People find losses more painful than gains are pleasurable. To avoid the emotional pain of a loss, investors have a strong tendency to hang on to losers (they treat a loss as not real until it’s realized). In some cases that may turn out to be a good decision and in some cases it may turn out to be bad. The problem is that the investor faced with a loss is largely leaving it now to chance and hope rather than to a sound and rationally informed decision based on a reassessment of the fundamentals. Emotion has clouded the decision.
Hindsight Bias: Many events, innovations, or concepts seem obvious after the fact. Individuals regularly forget that something that now seems obvious was not nearly so obvious before it happened; we can’t imagine not knowing something once we know it. In our case it would have been: how did we miss that the Fed’s unprecedented monetary policies would be very successful and that we need not worry about the risks associated with deleveraging? The danger is that hindsight can twist a rationale; a decision can appear like a mistake, or it can lead investors to overestimate their ability when they have been successful. As such, hindsight can falsely create the belief that the world is a knowable and controllable place (which is often not the case). This unrealistic overconfidence can lead to excessive risk taking when the hindsight creates an illusion of predictive power, and to paralysis when an investor becomes wracked with doubt on account of his mistakes.
Five years after the worst financial crisis since the Great Depression, we feel fortunate to be where we are. The overall economy and our clients would be in far worse financial shape if our concerns related to deleveraging were realized.
Yes, there are still lots of problems. Unemployment is much higher than stated in reported numbers given historically low labor-participation rates (the long-term unemployed). Real incomes for average Americans are not growing. Our policies seemingly continue to cater to the financial economy rather than the real economy. It is unclear how our large public debt and long-term fiscal obligations, including Social Security and Medicare, will get sorted out and how it will impact financial markets and our clients’ portfolios. Margin debt on the New York Stock Exchange is at record highs (meaning investors are evincing very little risk aversion), raising the risk of a sharp correction in the stock market. We are seeing frothiness in the technology sector, which reminds us of the late 1990s tech bubble. (Again, it’s important to reiterate we remain underweighted to equities even after the changes we are making.)
There’s a whole laundry list of concerns and uncertainties that we always weigh in our investment decisions. The question we always ask is how material are these risks and what’s the likelihood of them playing out. In our minds, a disruptive deleveraging process was a risk scenario we felt was prudent to protect our clients against. Part of that protection came from underweighting equities. By definition, if a risk scenario we were insuring against does not pan out, then there will be a cost, though longer term it can be recovered from other investment decisions we make at the portfolio level, and we have made up quite a bit already. At present the biggest risk in the equity markets is valuation risk. Stocks are not egregiously expensive (yet), but they are definitely not cheap and do not adequately compensate investors for taking on full equity risk.
We have frequently written about our focus on generating absolute returns and mitigating downside risk rather than reaching for relative returns. That philosophy bears out in the 9% nominal return hurdle we use to gauge whether or not we should have a neutral weighting to equities. Many strategists, even some we respect, would overweight stocks at a much lower return hurdle, citing very low interest rates. But that stance does not adequately factor in equities’ higher absolute downside risk in our opinion. Also, rates will eventually normalize: even if we don’t know precisely when, we still have to prepare our portfolios for the possibility of what will happen when they do. Moreover, in the current period of extremely low bond yields where there isn’t a lot of cushion against a recession or an economic shock, the need to insure against downside risks is greater than when yields were higher. Ultimately our asset class weightings—and, specifically, our willingness to take on equity risk—rest on our view of return and risk potential for the asset class as well as the objectives and risk threshold of each portfolio. These are our foremost, and ongoing, considerations as we manage our portfolios and work with our clients to achieve their goals.