The Shift to Reflation: Assessing the Impact to Portfolios
Asset class risk and return characteristics are morphing as markets move to a new reflationary environment. Investors should reassess their portfolio construction to take these new characteristics into consideration and evolve allocations accordingly.
In our earlier research piece entitled From Stall Speed to Reflation: A Turning Point for Investing, we examined a secular turning point as markets moved away from the difficult slow period of 2009 through mid-2016 toward a new reflationary regime. This new regime is characterized by the end of post-crisis private sector deleveraging leading to higher confidence, resulting in an acceleration in consumption, investment, and ultimately an acceleration in prices.
Understanding regime changes can help investors capture higher returns and more effectively manage risk. This paper expands on this topic by offering an actionable framework for portfolio construction under the new reflation regime, with specific portfolio changes investors may wish to consider. In particular, we expect growth assets and actively managed strategies to once again take the lead in generating portfolio outperformance.
Portfolio implications of the new regime
What worked well from an asset allocation perspective during the stall-speed regime is unlikely to offer strong risk-adjusted return moving forward. While academics have demonstrated that asset classes tend to maintain generally consistent Sharpe ratios over long periods, these averages may not hold true under different market regimes, which can, at times, span over several cycles.
As shown in the chart below, our research has identified five types of broad market regimes. Each has its asset class favorites, elevating some Sharpe ratios, while depressing others. The darlings of stall-speed regimes tend to be safety assets; overheating regimes favor real assets, and balanced regimes favor growth assets. During the recent stall-speed regime, investor flows went into the same safety assets that were being pursued by central banks. Accordingly, US Treasuries, German bunds, and Japanese government bonds all produced Sharpe ratios nearly two times that of past long-term averages, while growth asset Sharpe ratios fell by half.
Recent trends point to a reflation shift, prompting changes in asset allocation opportunities.
From Stall Speed to Reflation: A New Market Regime Emerges
Source: PineBridge Investments as of 31 July 2017. This diagram is for illustrative purposes only and does not solicit or recommend any action based on this material. Any opinions, projections, forecasts, or forward-looking statements represent the views of the investment managers, are valid only as of the date of this presentation, and are subject to change.
Since the summer of 2016, however, there has been a changing of the guard, with an expected meaningful rotation back into growth allocations, a trend that should accelerate with somewhat faster unit and pricing growth (which we refer to as reflation). Some investors have already begun reallocating away from safety assets, and the current combined balance sheet intentions of the Fed, European Central Bank, and the Bank of Japan point to a collective shrinking of their sizable exposures sometime after the end of 2018. This represents a long process to reorient portfolios to a more normal mix of capital appreciation and capital conservation.
As cash flow growth strengthens, correlations will naturally evolve as well. Our regime framework points to fewer diversification benefits between growth and capital conservation assets than during the stall-speed period. Pronounced negative correlations between risk and safety assets are stall-speed trademark characteristics. Our research suggests that these correlations will move back to levels more typically seen in reflationary or balanced times. For example, longer-term correlations between global stocks and global bonds have been close to zero, whereas they persisted near -0.4 during the stall-speed years, providing unusually strong diversification benefits. Since last summer, this relationship has already muted back to a -0.15 to -0.2 range, more typical of reflationary regimes.
Along with lower correlations between risk assets, lower pairwise correlations between securities within risk assets also indicate the potential for expanding outperformance between active and passive strategies. High pairwise correlations between stocks tend to signify a market with stocks going up and down in unison. A more normal relationship with lower pairwise correlations implies a market where individual stocks move less in unison and more in line with each company’s successes and failures at the corporate level. Collectively, these changes have significant portfolio ramifications in terms of which assets to own, the best strategies to fill those allocations, and how to manage risk.
Reflation: Good for P/Es
Stall-speed and overheating regimes represent two imbalanced extremes, both characterized by unhealthy correlations, valuations, and inflation levels. After years of low inflation, a common current misperception is that any uptick will hurt asset values. Empirically, this has not been the case, especially when it is a move away from the dangerously low levels experienced in a stall-speed cycle to something more comfortable and balanced (for companies, investors, and central banks).
Logically, higher inflation rates used to discount cash flows should lower valuations; however, this considers only one factor. Equity discount rates actually have two components which create a decoupling of risk-free rates and price-to-earnings (P/E) ratios once interest rates and inflation drop to dangerous levels that can push up risk premiums. As a result, we do not think it makes sense when starting from stall speed to expect a linear relationship between rising inflation, interest rates, and falling P/E ratios as one moves from dangerously low levels of inflation to more balanced levels. Moves beyond balanced conditions toward imbalanced, unhealthy overheating should bring about this inverse relationship, but more balanced inflation can actually be extremely beneficial for stocks.
Using US data back to 1950, P/E ratios have generally risen as inflation moved from a 0 to 1% range to a 2% to 3% range (see chart below). These ratios have then usually started to decline as inflation increased to higher levels. Implicit in this trend is the second component of equity discount rates, the equity risk premium, which tends to rise when growth is fragile, as typified by inflation below 2%. Moves out of a stall-speed environment to more acceptable levels of growth and inflation generally have been accompanied by falling risk premiums that have declined faster than risk-free curves have risen. We expect this trend to hold true this time around as well, particularly given the slowly dissipating glut in savings as companies begin investing again and central banks remove their distortion to the longer end of risk-free curves.
Steady 1-3% inflation levels have been supportive of forward P/E gains for many stocks.
Forward P/E Ratios Versus Inflation
1950 to 2016
Source: Schiller as of 31 December 2016. Data run since January 1950. For illustrative purposes only. We are not soliciting or recommending any action based on this material. Inflation is annualized based on 36 month rolling data, with 18 months prior to the forecast date and 18 months subsequent, and P/E based on future 1 year earnings starting from forecast date. Any opinions, forecasts and forward-looking statements presented above are valid only as of the date indicated and are subject to change. We are not soliciting or recommending any action based on this material.
Shifting to a forward-looking analysis
By applying this knowledge of potential constructive P/E expansion, it becomes clearer how focusing on the rules and conventions of the past regime to construct asset allocations moving forward may actually place portfolios at greater risk. During the stall-speed regime, lower-risk asset classes generally outperformed higher-risk counterparts, as a post-crisis secular flight to safety drove government bond valuations steadily higher. With equity P/E ratios set to strengthen under the new regime, we expect a rotation away from allocations favoring capital preservation assets, which are currently at extremely high levels compared to historical norms across a broad range of investor segments, including individuals, pensions, insurance companies, sovereign wealth funds, and endowments and foundations. Of course, some of this emphasis on conservative assets has been prompted by structural elements, such as risk-based Solvency II regulations and a growing focus on liability-driven investment strategies.
Reallocations should continue to gain speed as investors increasingly follow the favorable returns of growth assets and start thinking of risk in terms of opportunity cost. There is also the very real potential to rediscover the ability of safety assets to lose money when mispriced to the degree they are today. Once that occurs, potentially impinging capital levels, regulators will likely begin to reconsider the degree of safety assets deemed necessary to ensure liquidity. Unlike growth assets, whose cash flows will accelerate during reflation, cash flows for government bonds will not change. With record-long durations, this asset class will likely bear the brunt of revaluation even if rates rise modestly. Consequently, lower nominal portfolio returns will result primarily from the drag from fixed income.
In addition to less favorable returns, government bonds will likely offer diminished negative correlations to risk assets as well, which in turn will mean less favorable diversification benefits. The same fear-based investor mentality that created high correlations between risk assets (which shared a mutual vulnerability to exogenous shocks at a time when the economy was barely moving forward) presented an equal but opposite bonanza for safety assets, resulting in a pronounced, classic stall-speed negative correlation. Yet since private sector deleveraging ended last summer, this relationship has muted in conjunction with declining correlations between risk assets, as well as the pairwise correlations between securities within these risk assets.
Unfortunately, many portfolios utilize backward-looking correlation matrices in their allocation strategies and should consider replacing these with expectations that make more sense given the new environment. One example of how this might translate into portfolio design is the lens of US bank loans. Government bonds should struggle to offer the same levels of defensive characteristics relative to risk assets at the same time that all fixed yield assets face an increasing risk of price deterioration from rising interest rates. This pressure may be gentle at first, but with prolonged low yields comes increased duration, so that it no longer requires much of a rate increase to produce poor return outcomes.
The floating rate construction of bank loans, while still subject to the expected lower protection against equity declines from bond assets in general, should offer a compelling rising-rate defense that can also help satisfy investor need for income. This allocation opportunity could be lost using a backward-looking 10-year trailing correlation matrix that favors safety assets and makes it difficult to differentiate between muted returns across fixed income segments. Using a forward-looking correlation matrix, however, is less prone to have the higher future return potential of bank loans versus fixed maturity credit instruments overshadowed by the past pronounced diversification benefits provided by government bonds that largely drove prior-cycle returns and that are unlikely to recur in the period ahead.
Changing correlations affect cross-asset class behaviors
For other examples of how a reflationary regime may change multi-asset portfolio dynamics, it is critical to assess how the new investment climate may change correlations across the entire spectrum of asset classes. As global growth expands and becomes more sustainable, correlations between growth/ risk assets tend to weaken. This should help markets become less sensitized to the extreme risk-on/risk-off mindset of the stall-speed regime that often pulled all growth assets in one direction and all capital conservation assets in another.
Pronounced allocations to safety inspired by various forms of parity investing are due for a haircut as the new regime cycle unfolds. Once capital conservation offers less short-term downdraft benefits and less return otherwise, and growth/risk asset classes offer greater diversification relative to one another, a focus on risk/return will again steer allocations to higher growth asset levels.
The exhibit below is based on our empirical observations of typical correlation behavior of high-level asset class categories under different regimes. Expected correlations of global bond, commodities, and the US dollar relative to global equities all fall by half in the reflation regime compared to the stall-speed regime. Indeed, as we presented in our last paper, correlations across assets have already started to decline to levels not seen since 2006.
Different Market Regimes Exhibit Different Asset Class Correlations
Cross-asset class correlations tend to move lower in a stall-speed to reflation regime shift
Source: PineBridge Investments. Based on our five-year forward-looking view. Diversification does not ensure against market loss. Data is for illustrative purposes only and should not be considered a solicitation or recommendation of any action based on this material. Past performance is not indicative of future results. There is no assurance that any investment objective will be achieved. Certain statements contained herein may constitute projections, forecasts and other forward-looking statements that do not reflect actual results and are based primarily upon applying retroactively a hypothetical set of assumptions to certain historical financial information. Any opinions, projections, forecasts and forward-looking statements presented herein are valid only as of the date of this document and are subject to change. PineBridge Investments is not soliciting or recommending any action based on any information in this document. While fundamentals driving returns and risk are updated quarterly, estimates for correlation could change that frequently yet often are addressed once per year unless in monitoring patterns in the market we believe a more up-to-date review is required. The last evaluation was based upon historical data as of 28 February 2017 and how that shaped our forward views at the time.
It is also important to evaluate how structural evolutions within an asset class may alter correlations. High yield bonds offer a real-world example of these dynamics. During and after the global financial crisis, high yield bonds behaved more like equities and commodities, with elevated correlations as risk assets sold off in tandem across all three segments. By 2013, the end of the extreme flight to safety saw the asset class’s systemic exposure to equities and commodities declined to more historically normal, bond-like levels. This is similar to the broader trend experienced by other higher-risk fixed income assets during the stall-speed regime.
There were major structural disruptions specific to the high yield bond market that altered correlations. Shale companies had begun to issue significant volumes of high yield debt, taking advantage of low post-crisis yields. This increased the segment’s exposure to the oil industry, and when oil prices crashed beginning in July 2014, high yield bond correlations to commodities once again spiked. In 2016, OPEC and other nations agreed to cap production to stop the steep oil price declines. Volatility fell to more historically normal levels, and high yield bond correlations to commodities subsided.
Despite this stabilization, the asset class remains less attractive in our view, due to tighter spreads and potentially stretched balance sheets that will demand greater selectivity for success. US bank loans in comparison offer a lesser weight to energy and stronger investment potential at this point in the cycle being more senior and secured, a status that means more as companies begin to invest well above depreciation again, a phase that begins to stretch balance sheets and lower credit quality.
Like regime shifts, structural disruptions such as the 2014 oil crash can dramatically alter asset class correlations.
High Yield Bond Correlations to Commodities
Source: PineBridge Investments as of 30 June 2017. Data reflects Bloomberg Barclays US Corporate High Yield. For illustrative purposes only and should not be considered a solicitation or recommendation of any action based on this material. Opinions, forecasts and forward-looking statements presented above are valid only as of the date indicated and are subject to change.
Intra-asset class considerations
The new reflation regime should also affect how sub-asset classes, different management styles, and individual sectors perform relative to one another. For example, we expect equities generally to outperform capital preservation assets but small cap securities, in particular, to benefit more from gradually accelerating and broadening economic growth.
Plus, as correlations weaken, active equity strategies should experience more differentiated returns compared to passive styles, both in terms of skilled active managers adding more meaningful value and weaker active managers lagging the broader market. This should help fuel additional outperformance potential for strong small cap managers, since the segment has historically offered a much wider number of securities to choose from than large cap equity markets, with a much wider dispersion of returns. Therefore, selecting experienced active managers with proven track records should become increasingly important for a return advantage. Greater business disruption also looks likely as new technologies render obsolete former business models, another trend that should support skilled active managers to deliver more differentiated returns.
From an industry sector perspective, we also see greater potential in moving away from bond proxies. Select allocations to those most suitable to the new regime offer return and risk benefits. Consider US financial equities. Gradually rising US interest rates should benefit US financials’ net interest margins, offering a potential hedge to rising rates. This is illustrated below, with US financial equity correlations to US Treasuries maintaining their pronounced negative correlations, even as these correlations dampen between the broader equity market and government bonds.
Decoupling correlations indicate select US financial stocks may offer higher returns and a potential interest-rate hedge.
US Financials Equities and US Equities Correlations to US Treasuries
Source: Bloomberg as of 11 July 2017. Data reflects S&P 500 Index and S&P 500 Financials Index daily correlations (one-year rolling daily data) with US Treasury returns. For illustrative purposes only and should not be considered a solicitation or recommendation of any action based on this material. Opinions, forecasts and forward-looking statements presented above are valid only as of the date indicated and are subject to change.
Additionally, the same rising confidence characteristics bringing about the reflationary regime are also having a pronounced positive effect on financials. Top-line prospects for these institutions should improve with deleveraging and stepped-up investment by corporate clients. Furthermore, if the US regulatory environment does indeed lighten under the Trump administration, significant cost escalation, which was driven by regulation, may itself provide an extra tail wind to income statements.
New areas for portfolio optimization
When correlations change, they tend to do so very quickly and then persist at the new levels. Hence, investors should be well served by re-examining their forward-looking correlations sooner rather than later, as today’s new market regime continues to emerge. In fact, we have revised our proprietary forward-looking outlooks, as highlighted in the exhibit below. Across asset classes, we see a clear move toward declining correlations, indicated by the changes from dark red (high positive correlations) and dark blue (high negative correlations) to more light red, light blue, and white (more independent correlations).
We expect asset classes to move less in tandem, creating opportunity for skilled active managers – evidenced by the shift to 0 correlation.
Asset Class Correlations in the Stall-Speed Regime
Versus Current Asset Class Correlations
Source: PineBridge Investments, based on our five-year forward-looking view. Diversification does not ensure against market loss. Data is for illustrative purposes only and should not be considered a solicitation or recommendation of any action based on this material. Past performance is not indicative of future results. There is no assurance that any investment objective will be achieved. Certain statements contained herein may constitute projections, forecasts and other forward-looking statements that do not reflect actual results and are based primarily upon applying retroactively a hypothetical set of assumptions to certain historical financial information. Any opinions, projections, forecasts and forward-looking statements presented herein are valid only as of the date of this document and are subject to change. While fundamentals driving returns and risk are updated quarterly, estimates for correlation could change that frequently yet often are addressed once per year unless in monitoring patterns in the market we believe a more up-to-date review is required. The last evaluation was based upon historical data as of 31 August 2017 and how that shaped our forward views at the time.
These weaker correlations suggest portfolio success will be dictated by active investors selectively seeking more growth assets. As of this writing, our Capital Market Line (CML), a five-year, forward-looking view into risk and returns, remains inadequately sloped; yet it has steepened since the middle of 2016, in contrast to the flattening continuum of the past few years. This increased slope has also further widened dispersion, increasingly reflecting a widening of nominal fundamentals for growth asset classes whose cash flows can benefit from faster global growth versus fixed maturity assets whose cash flows will remain the same.
With even wider dispersion of returns and less correlation between most asset classes, the theme of selectivity remains intact. In this changing landscape, investors should consider if they may be:
- Risking losses from safe-haven assets that performed well under stall-speed conditions but may be more vulnerable to losses looking ahead, such as government bonds
- Overexposed to conservative assets in general, especially sub-segments that became bond proxies and “couponized” trading ever higher as surrogates for high-quality bonds
- Relying too much on passive assets, given the changing disruptive business environment and lower pairwise correlations between securities that imply greater alpha potential for adept active managers
- Utilizing an allocation strategy that may be optimized for past conditions markets have segued beyond, warranting a new approach
Regime changes can bring rich rewards in the form of returns to prepared investors who understand the potential effects on asset class cash flows. It is equally important to remember that correlations are constantly evolving. Regime shifts are only one input into this equation, and working with skilled allocators who have demonstrated a repeatable ability to effectively manage this multi-faceted aspect of portfolio design can help investors navigate what lies ahead.