2017 was the year of the structural winner, with the market led higher by the digital platform “FAAMNGS” (the acronym keeps extending as the list of perceived winners broadens) and the stylistic domination of Growth. Disruption was never more topical, with nothing appearing more disruptive than crypto-currency euphoria. 2018 may be the year when central banks finally reconsider the wisdom of “inflation targeting” and savers revolt.
More prosaically, the year continued a period of sustained and elevated global growth with Emerging Markets (EM) and Europe taking the leadership mantle from the U.S. In Germany, where inflation is running at a steady 1.8%, Bunds yield ~0.4% resulting in around 1.4% p.a. wealth erosion to the holder. The situation is similar in the United Kingdom, where CPI is likely to remain stubbornly high due to a weaker currency and structurally higher fiscal obligations. UK Gilts yield 1.3% against 3% inflation. But despite this regime of strong growth and moderate inflation, approximately US$11 trillion in bonds remain negative yielding and over US$2 trillion is parked in strategies dependent on low volatility and stable asset price correlations.
The actions taken by the ECB in 2012 to prevent the Euro area sovereign crisis from morphing into a full blown banking solvency crisis were warranted. However, today the ECB is maintaining emergency policy settings even though the European Union (EU) is set to grow faster than the U.S. for the second year running. At present, the ECB purchases seven times the net issuance of EU sovereign debt, suppressing yields across the world. Additionally, the Bank of Japan (BoJ) continues to purchase the equivalent of 16% of GDP per annum (approximately twice the net fiscal issuance), all in the name of inflation targeting, though ignoring the strength of 2.2% real GDP growth. With European corporate high yield or “junk” bonds yielding less than the U.S. 10 Year Treasury, it would seem that investors are extrapolating the ECB’s and BoJ’s emergency settings into perpetuity.
Globally, central banks have accumulated approximately US$20 trillion of sovereign bond assets via QE programs against total global GDP of US$90 trillion. If central bank targeting of long dated low real yields has been the lever lifting long duration assets, then the lever is now slowly reversing. Over the quarter, the ECB announced a reduction in its bond purchase program, the BoE hiked interest rates and the Fed hiked in December. Asian central banks are also on the move with the BoK hiking interest rates and robust economic growth in Malaysia, Thailand, and India pushing those central banks towards normalisation.
The potential distortions resulting from this enormous monetary experiment need to be considered. Examining the mid-2000s, a virtuous cycle of cheap funding via surplus EM savings combined with a search for yield resulted in low volatility which fuelled leverage in the developed world private sector. Today, we have a different set of issues:
- Central banks are the key actors aggressively targeting the longer end of the yield curve, with the weighted average maturity of central bank government bond holdings at 8-9 years for the Fed, ECB and BOJ. Volatility targeting investment strategies, such as risk parity and minimum variance are an OUTCOME of large scale bond buying programs, rather than the CAUSE of persistently low volatility.
- Whilst equity investors see themselves as central, it is higher up in the corporate capital structure in credit where the impact of QE is much more direct. As central banks have crowded private investors into high yield debt, spreads have compressed across the risk spectrum.
- Now the cycle is in a virtuous period where low yields and low volatility reinforce the greater use of leverage to manufacture returns.
By kicking the proverbial can down the road, central banks have somewhat cornered themselves. Increasingly, political and economic pressure to normalise interest rates or withdraw stimulus is likely to trigger volatility and widen credit spreads. Our analysis suggests that U.S. high yield, or junk bond issuers are most vulnerable to this risk – see our research paper titled “The Global Corporate Debt Unwind”. Whilst the low-volatility regime may endure, investors have grown too comfortable with the central bank reaction function.
Regional & Sector Return Expectations
Note: Cyclically Adjusted PE, Sector and Factor definitions can be found in the Glossary
We find it useful to examine long-term empirical data in terms of what starting multiples imply for expected future returns. Given broad differences in the timing of earnings cycles across both regions and sectors, we prefer to measure broader expectations for future returns based on “Cyclically Adjusted PE” (CAPE) valuations.
When a market commentator makes a pronouncement regarding the valuation of a particular country or region, investors should inquire as to what extent this is driven by a sectoral bias within the market. For an extreme example, the Korean market valuation is heavily tied to the valuation of a single stock, Samsung Electronics (as is Taiwan to Taiwan Semiconductor Manufacturing) and EM valuations are generally sensitive to global hardware, commodity and domestic banking valuations – the three key EM industry “over weights” relative to DM. For this reason, Antipodes Partners’ structures its analytical efforts to get the most out of BOTH global sector (globally oriented businesses) and geographic (domestically oriented businesses) contexts.
However, for the purposes of assessing the valuation of equities more broadly versus other assets, the best data-set is country based, as the pricing of cash and bonds will be heavily driven by domestic considerations. With data back to 1947 and the greatest longevity of record, we can observe a strong relationship between the S&P 500 CAPE, ten year forward S&P 500 returns, and U.S. corporate profit share as a % of GDP (as a proxy for U.S. profit margins, which effectively applies a “structural” adjustment to the CAE). A high CAPE is typically associated with lower forward returns, though this relationship is tempered when profit margins are low and the probability of the market growing into its multiple increases.
CHART 4: RELATIONSHIP BETWEEN THE S&P 500 CAPE, U.S. CORPORATE PROFIT SHARE AND 10 YEAR FORWARD RETURNS
Source: Robert Shiller, Federal Reserve Economic Data, Antipodes Partners
Chart 4 captures this thought process. Given the timeframe of the forecast, it has little relevance to timing shorter-term market moves. However, taking a long-term view, with U.S. corporate profit share at a near peak of ~9.5%, and the CAPE approaching a 15 year high of 30x, the historical pattern implies relatively modest ~1.5% p.a. forward returns, significantly lower than the long-term average.
TABLE 4: FUTURE EXPECTED SECTOR AND REGIONAL RETURNS (2017)
See Glossary for CAPE, EMR and Sector definitions. Source: Antipodes Partners
The question then becomes, if U.S. equities empirically appear expensive, what does the global, regional and sector picture look like? Similar to the above approach, this requires the formulation of both a cyclical and structural view of return expectations. Our CAPE based view of returns (Table 4) captures through-cycle earnings expectations (in the context of the long-term CAPE) without an adjustment for the longer-term profit margin potential. For this reason, the expected returns produced by the Sector/Regional CAPE analysis are not directly comparable to the U.S. expected return analysis (Chart 4) Rather, they reflect a relative ranking of opportunities that then need to be adjusted by a longer-term structural assessment of profitability and valuations (Figure 1).
In this sense, with the U.S. as the benchmark for absolute return expectations, broadly both North American and Developed European equities look expensive. Given that these regions represent ~76% of the MSCI ACWI, investing in the global index is unlikely to lead to a great long-term return outcome. Comparatively, both Developed Asia (Japan, Korea and Taiwan) and EM stand out as regions with great return potential.
FIGURE 1: REGION-SECTOR VALUATION HEAT-MAP – EV/SALES RELATIVE TO WORLD – Z-SCORE (1995-2017)
See Figure 2 for a more detailed explanation. Source: Antipodes Partners
We would stress that much of the difference in regional return expectations is driven by compositional differences in industry exposures. From an industry perspective, in spite of growing nationalism/populism, globalisation has resulted in valuation multiples that are relatively similar within industry sectors across regions. In this sense, we broadly observe:
- As the rally in Chinese growth sensitive equities has played out over the past two years, Materials and Industrials outperformed and, though Energy rallied late in the year, it remains very cheap by historical standards.
- Though Financials significantly outperformed in the first half of last year, the sector remains cheap by historical standards with sentiment and profitability expectations weighed down by macro-concerns, low rates and yield curve compression.
- Domestic Cyclicals are also cheap by historical standards, especially strong incumbent retailers where the market is potentially underappreciating the brand equity and/or a successful adaptation to online reality.
- Interestingly, in a market that until recently has paid up for Yield -most intensely reflected in the North American Infrastructure sector- traditional yield sectors such as Telecommunications have de-rated (Feature, page 12), coinciding with the apparent value of Good Yield (funded through cash flow, see Figure 3).
- Whilst Technology appears expensive on a more structural view of valuations, we guard against comparisons to the 1999/2000 tech bubble – Growth as a style is currently pervasively expensive across the global market, not just in Technology. Whilst real structural change (cloud, social, virtual reality, media streaming, big data, autonomous driving, etc.) has underwritten the outperformance of the sector, conversely due to their sheer size, risks are building for the Titans of Tech, including:
- Tax and regulatory risk arising from increased scrutiny from governments around the world due to their influence and role in society, particularly around managing sensitive information.
- Intensifying competition, whether it’s Amazon and Netflix competing on content streaming or Amazon’s Alexa threatening Google’s core search business with its voice search capabilities, the titans are bumping heads. Likewise Google is encroaching on Apple by focusing on its own smartphone and Microsoft is attacking Amazon’s AWS cloud business with its successful Azure platform. Related to this is the growing capital intensity of many of these businesses, as they make heavy infrastructure investments in compute capacity required to handle increasing artificial intelligence workloads and video streaming demands.
- Style/Macro risk represented by excessive crowding by investors that have paid up for structural growth. If the current cyclical rebound in global growth continues, long-term global interest rates will be forced higher, triggering a rotation out of expensive longer-duration exposures into out of favour cyclical stocks.
- Consumer Staples were the expensive defensives, once enamored for their perceived Profitability and Growth characteristics (Figure 2), that have now underperformed since the beginning of 2016. However, they still remain one of the most expensive DM sectors and in many cases structural pressures, such as substitution by private label, are intensifying.
- Healthcare has underperformed to the point that relative value has appeared. However, broadly healthcare related businesses will continue to be pressured by the public and governments grappling with affordability given decades of high cost inflation.
Most quantitative strategies would measure the attraction of a certain “factor” exposure on the basis of its price momentum. These same systematic strategies see Value as a separate factor, by which they mean low multiple stocks, but will only buy Value if it exhibits momentum. At Antipodes Partners, we worry that factors favoured by systematic strategies will eventually become overvalued (a symptom of momentum based crowding), offering little margin of safety at the stock level and exposure to “regime change” style draw-down risk at the portfolio level. Hence, we value a range of factors (e.g. Profitability, Growth, Resilience, Multiple Dispersion, Good Yield, Volatility and Momentum, etc., Figures 2, 3 and 5) rather than Value in isolation (we prefer to label Value as Multiple Dispersion to make a clear statement that the starting multiple is meaningless without the context of growth).
Keeping with our philosophy of finding investments with multiple ways of winning, another way we can win is by gaining:
- Cheap exposure to an expensive factor, e.g. Cisco Systems, a very cheap “Profitability” exposure at a PE of 13x where an equivalent exposure would cost on average +20x PE.
- An exposure to an out of favour factor based on the view that the market view may change, e.g. “Good Yield” (cash-flow funded) is offered cheaply by many Utility and Telecommunication stocks, as opposed to the expensive “Bad Yield” (capital market funded) offered by many Infrastructure stocks.
Accordingly, Figures 2, 3 and 5 apply our proprietary quantitative tools to determine how expensive various factors have become relative to the last 30 years (expressed as a Z-Score) by comparing the valuation of the most profitable (highest growth/momentum or lowest multiple) stocks to the least profitable (lowest growth/momentum or highest multiple) stocks.
CHART 5: Z-SCORE (MEDIAN EV/CE OF UPPER QUINTILE RELATIVE TO LOWER QUINTILE 1987 – 2017)
See Glossary for factor definitions. Source: Antipodes Partners
Chart 6: Z-Score (median EV/CE of upper quintile relative to lower quintile 1987-2017)
See Glossary for factor definitions. Source: Antipodes Partners
With reference to Figures 2, 3 and 5:
- The market is celebrating stocks that display high Growth, Profitability and Momentum independently of starting multiple. Further, it’s noteworthy that the market’s willingness to pay up for Growth, Profitability and Momentum is approaching the heady days of the late 1990’s tech bubble. More specifically, extreme policy settings in developed markets have led to severe investor herding, evidenced by the extreme overvaluation of Growth and Profitability in these regions. One can also observe the subsequent derating that occurred as high Growth and Profitability attracted competition and these stocks lost their allure, a clear example of how a high starting multiple was predictive of future sub-par returns.
- Extreme thirst for equities with bond like characteristics, i.e. Yield and Low Volatility, without concern as to the inherent risk equities represent. As Minsky alluded to, prolonged periods of stability or low volatility do not necessarily equate with low potential risk. In fact, today many investment strategies are increasingly betting on stability and low volatility to generate returns. Interestingly, whilst equity index and cross asset index volatility remains low globally (Figure 7), at the individual security level, the market’s love affair with Yield and Low Volatility appears to be waning with a notable retreat in the valuation of both these factors from the extremes of early last year.
- Encouragingly, Multiple Dispersion is evident across all regions, however, low multiple stocks continue to underperform high multiple stocks.
The general uptrend in the broader equity market seems set to continue given economic data globally remains robust and central banks very accommodating. However, the blind assumption of unendingly low rates is a dangerous one. As a result, we simplistically see two likely scenarios:
- Growth continues to surprise to the upside driving greater urgency from central banks to normalise policy. To minimise disruption to short-term funding markets, tapering would likely focus on the long-end of the yield curve leading to a potentially self-reinforcing pro-growth steepening, resulting in a significant increase in bond volatility and headwinds for the crowded/expensive low volatility, bond proxy and growth/quality equity exposures.
- Growth disappoints due to policy tightening by China or the U.S. In this scenario, credit volatility would spike triggering a major sell-off in credit sensitive equities regardless of their duration, i.e., a repeat of the 2015/16 commodity high yield melt-down which ended up spilling over into non-commodity exposures. Conversely, the inevitable central bank response would extend the illusion of stability and amplify the imbalances with a continued melt-up in the low volatility, bond proxy and growth/quality equity exposures.
Antipodes Partners’ investment goal is to build portfolios with a capital preservation focus from non-correlated clusters of opportunity. In our long investments we seek both attractively priced businesses (margin of safety) and investment resilience (characterised by multiple ways of winning), with the opposite logic applying to our shorts, i.e. no margin of safety and multiple ways of losing. Whilst the investment case will always be predicated on idiosyncratic stock factors such as competitive dynamics, product cycles, management and regulatory outcomes, we seek to amplify the investment case by taking advantage of style biases and macroeconomic risks/opportunities.
Given the divergent risks that the above two scenarios represent, investors should focus more than ever on uncovering sources of idiosyncratic alpha, rather than relying on momentum or passive beta. In this sense, we’re encouraged by the high level of valuation dispersion within and across markets (region/sector/factor) as indicative of broad pragmatic value opportunities, both long and short.
To find out how Antipodes has positioned itself in global markets, read our December quarterly report