What our House View means for asset allocation and portfolio construction.
15 minute read
- Favour capital preservation
- Slight underweight in equities to reduce the economic sensitivity of portfolios
- Overweight government bonds with preference for US, Australia and UK
- Neutral across credit
The start of 2020 feels like a lifetime ago. In our 2020 Outlook (published in December last year) our expectations were for a moderate growth recovery this year and a preference to be overweight risk assets across our portfolios.
The COVID-19 pandemic has dramatically impacted the global economy and financial markets, and caused us to materially change our own outlook. The impact on the well-being of those impacted, as well as the policy measures taken to curtail the spread of the disease has led to a sudden halt in economic activity.
Overweight sovereign bonds and underweight equity
As a result, we have changed our asset allocation views to reflect both the immediate impact and further disruption over 2020. Specifically, we have increased our overweight in government bonds, expressed through overweight positions in the US, Australia and the UK. We have also reduced our equity allocation to be overall underweight.
The equity allocation reflects our concern about further severe economic weakness putting material downward pressure on corporate earnings over the course of 2020. Our duration view is a reflection of both the demand for safe haven assets and the powerful response from central banks to ease monetary and financial conditions.
Further, we moved to a more cautious stance in our currency allocation, introducing a short sterling view and retaining a preference to be long Japanese yen and short Australian dollar. We have a neutral view across credit, where corporate bond spreads have widened dramatically.
Overall, these changes imply a significant reduction in the economic sensitivity of the portfolio.
Huge uncertainties favour capital preservation strategies
We acknowledge that there is enormous uncertainty regarding the outlook. That uncertainty pertains to the likely further spread of the virus, the measures governments might take to contain the spread, the economic impact of the both of those factors, and finally the monetary and fiscal policy support for households, businesses and asset prices.
With so much uncertainty, our confidence in market developments from here is low. As such, at this time we regard it as imperative to act in favour of capital preservation. Once there is more clarity, and a recovery is within sight, there will undoubtedly be more investment opportunities. Our asset allocation table provides a more detailed view. The remainder of this section looks at the seizing up of global liquidity in March, followed by the outlook for the major asset classes in more detail.
Global funding market stress
Severe liquidity stresses materialized in many markets, particularly funding problems and leveraged fixed income securitie
Liquidity and market function (sometimes referred to as “plumbing”) problems, particularly a global dollar shortage, became a key concern in global markets in mid-March.
This breakdown was reminiscent of the Global Financial Crisis (GFC) and was causing the prices of certain assets to be dislocated away from fundamentals, as liquidity risk rose sharply. Without a circuitbreaker, the rising liquidity premium can feed back into fundamentals and create further problems – at the extreme precipitating a fire sale of assets.
Central banks have stepped in to address these issues, launching a range of facilities, which have already helped to stabilise and, in some cases reverse, the illiquidity problem. But challenges remain. A combination of embedded leverage in some investment structures and illiquidity in the underlying assets could continue to make it difficult to ascertain credit issues versus liquidity risk in some assets. The main aspects of these liquidity risks – a longstanding risk to our past House Views that has now materialized – are as follows.
Commercial paper (CP), credit lines, LIBOR
Problem: Many banks and companies fund themselves via short-term CP or repurchase agreements, which in turn form multi-trillion dollar pools of assets for money market investors.
This has become illiquid and issuance has stopped, while funds have faced redemptions; prime money-market funds lost nearly ten per cent of assets. Companies have drawn down committed credit lines from banks to increase their cash buffers, or substitute for other funding and/or lost revenues. Banks, faced with this sudden demand for cash, are hoarding liquidity or need to borrow from others via interbank markets, which elevates repo rates, LIBOR and unsecured lending rates.
Solution: Central bank buying and funding programmes for CP, opening the discounting facility/LTRO to a wider range of assets at less punitive rates, without stigma.
Deleveraging, forced selling
Many investors use derivatives to optimise portfolios or lower costs, from options, futures, and swaps to more complex structured products.
As markets became volatile, counterparties and clearing houses needed to raise margin requirements. Meanwhile, other investors (e.g. risk parity funds, or CLOs) became forced sellers to reduce risk or protect capital, and other managers got redemptions from their investors, and had to raise cash.
Leveraged hedge funds, who had been arbitraging treasuries against futures, also hit stop-losses or just tried to rush for the exit door before everyone else did. Credit funds who hedged interest rate risk sold their risky assets, but then had to cover their hedge by receiving swaps.
All these factors – and others – caused volatility to skyrocket, and cash products like treasuries or corporate bonds to underperform; this can be seen in the extreme negative basis seen especially in 30-year treasuries, where yields were as much as 104bps above swaps (Figure 1).
Figure 1. 30-year UST-Swap Spreads have dislocated
Solution: Central bank corporate bond buying and QE helps to a degree, but sometimes the fire just needs to burn itself out; eventually the selling will have run its course, and new funds will be deployed to take advantage of the dislocation. As the saying goes in commodity markets: “the cure for low prices is low prices”.
Internationally, the dollar is the one global currency, the safe asset and common medium of exchange par excellence. This may or may not be optimal or desirable, but that’s the way it is in the global financial system of 2020.
The sudden halt in much global activity and trade, combined with the drop in oil revenues to producers, has caused a global dearth of dollars. This impacts the global financial system, as it did in the 2008-9 financial crisis: everything from currencies to trade finance to collateral on futures and ISDA agreements with counterparties.
There is real risk of a systemic shock in some part of this complex system. One measure of this is cross-currency basis, which measures the additional cost of currency hedging beyond that implied by interest rates (Figure 2).
Figure 2. Intense USD funding stress partly assuaged
Solution: Once again, central banks needed to step up to the plate, providing dollars to the market to assuage the drought.
Global central banks, particularly the Federal Reserve, have stepped in to prevent a worsening crisis
Many central banks have amassed plentiful reserves; hoarding them instead of using them is a policy error. The key actor here is the Fed, which, not before time, redeployed its dollar swap lines for foreign central banks, made the facility accessible daily rather than weekly, and expanded the lending to important central banks of friendly countries – though leaving some, like Russia, Saudi Arabia, China, and Turkey out in the cold. The recent introduction of a FIMA repo facility for a broader range of central banks has plugged this gap.
Equity market returns are heavily dependent on realised and expected earnings (Figure 3). Earnings, in turn, are closely tied to the economic growth environment.
Figure 3. Tight relationship between equity prices and earnings
A sharp deterioration in global economic growth – we expect global demand to decline in the first half of 2020 by the most in the post-war period – thus translates into a deep contraction in earnings, increased probability of corporate default and therefore materially lower equity prices.
The cycle ends and equity markets enter bear market territory
As such, it comes as no surprise that global equity markets have sold off sharply in light of the global COVID-19 outbreak and the containment measures adopted, both of which are expected to substantially reduce revenues. Global equities quickly entered a bear market and as at the time of writing had fallen by around 25 per cent from peak levels in mid-February.
The emergence of value critically depends on the magnitude of earnings decline ahead
The selloff has spared no region or sector; however, broadly speaking, those areas of the market that are most economically sensitive, where balance sheets are weakest, where policy tools to combat the crisis are most limited and containment measures have been applied particularly extensively, have sold off most.
One area where we regard economic sensitivities to global demand as currently mispriced is between European and emerging market equities. While both are heavily geared to the global cycle, emerging market equities have so far fared much better than their European counterparts – a development we expect to reverse once the ramifications of slower global demand start affecting emerging market profits.
The sell-off in global equities led to steep valuation adjustments, as shown by the current cross-regional trailing P/Es vs. the multiple that prevailed when global markets peaked in mid- February (Figure 4).
Figure 4. Taking stock of equity valuations
There haven’t been many instances over the past 30 years or so during which the P/E ratio for global equities reached levels as low as those prevailing in mid March (as indicated by the red bars in Figure 5). The few times they did, average returns over the subsequent three months tended to be negative, but positive over the subsequent 12 months.
Figure 5. Sell-off has led to valuation adjustments
However, whether returns are positive over either time period critically depends on the severity of the economic downturn, both in terms of magnitude and duration. For example, in the Global Financial Crisis, a deep and prolonged crisis, three and 12-month returns were negative after multiples reached this year’s mid-March valuation lows.
Whether value has emerged or, asked differently, how much more downside there is to equities, critically depends on the expectation for the denominator, i.e. for the low point in the level of earnings. Consensus expectations for 2020 earnings growth are only slowly starting to adjust to economic reality. Having started the year with an expectation for around ten per cent year-on-year EPS growth for 2020, consensus has adjusted expectations down to low single-digit growth. We view this number as utterly unrealistic and expect further downgrades to come.
Rather than attempting to estimate P/E and earnings separately, we are mapping our global growth forecast directly onto global equities, thereby gauging what the historic relationship between the two could imply for near-term performance in the AC World Index.
Global growth outlook presents material downside risk to equity markets
The results of the analysis leads us to assume that in a scenario where policy is partially successful in offsetting the negative second round implications on growth, but where the healthcare and economic crisis continues for the next 3-6 months, we could see an additional double-digit drop in global equity prices, potentially going as far as reversing the largest part of the gains of the 2009-2020 bull market. It is worth remembering that the last two bull cycles in global equities prior to the one that started in 2009 ended by giving up 70-90 per cent of their gains.
These estimates by no means aim to give precise forecasts but rather attempt to offer a more comprehensive approach to assessing value in equities as opposed to merely looking at valuation metrices.
As equities tend to over- and under-shoot “fair” and/or long-term valuations and prices frequently and for extended period of times, we draw on additional tools to assess the attractiveness of being invested in the asset class. Owed to the well-researched tendency of humans to herd and for feedback loops to be reflexive (e.g. Soros, 2009), the speed and magnitude of declining prices by themselves help assess the risk-taking environment. The speedy declines seen over the past weeks have typically been associated with non-linear negative return territory.
Combining both, our fundamental framework with the observation of sharply negative equity market momentum, we regard the risk/return trade-off for broad equity indices as challenging, even after the recent steep drop.
However, there is a silver lining. As quick as the sudden halt to the economy has been brought upon us, once we move beyond the virus there is the potential for the steep decline be reversed just as quickly.
Contrary to most past recessions, this is not one driven by imbalances or excesses, but rather one caused by an external shock and the policy response to it. Once there is greater visibility on where economic growth troughs, equities might be quick to focus on the equally steep subsequent increase in economic and earnings growth. Remedy could also come from the implementation of further exceptional policy tools, such as supporting companies’ income directly and without insisting on the creation of future obligations, i.e. handing out money to companies.
Rapid revenue deterioration and funding problems have widened credit spreads, but not to usual recession wides. Defaults and more downgrades loom.
Corporate credit spreads widened violently in March due, in part, to a fall in oil prices of over 60 per cent from January to March. This was caused both by OPEC and Russia increasing supply, and because of the large demand destruction for oil and other commodities from the COVID-19 crisis.
In addition, disruptions to manufacturing supply chains alongside falling demand as a result of the outbreak will impair firms’ revenues, and this in turn will hamper their ability to access liquidity. Much of this widening in spreads is justified by worsened fundamentals, but some of it is already building in liquidity premium, in our view.
Specific sectors, particularly shale producers, leisure, hotels, and restaurants, are taking a huge direct hit on the demand side. Globally, governments are responding to help corporations, but such help may involve nationalisation and/or debt restructuring, as for example with automaker bailouts in the GFC.
A default cycle is beginning that we expect to be concentrated in these sectors; this is priced into high yield energy spreads which were in the 2000s in late March, but less so elsewhere where spreads remain well below previous recession peaks.
Central banks slashing rates and buying sovereign and, in many cases, corporate debt through massive asset purchase programmes will help governments to finance their fiscal packages at a lower cost, and provide aid directly to businesses.
Banks, companies, and households with mortgages should also benefit from these measures, so long as markets are functional. If not, they may be able to access central banks’ commercial paper or corporate bond purchase programmes, or take advantage of banks’ ability to fund these assets cheaply with the monetary authorities. However, some business models may be unviable, regardless of debt provision.
The current liquidity shock (see earlier section) means that credit lines may be cut off, just as corporate bond funds face redemptions, the commercial paper market has issuance problems, and CLOs or other “rules-based” managers become forced sellers. This can cause spreads to widen due to supply (Figure 6), but also can cause firms themselves to have serious problems, with illiquidity potentially translating into insolvency. Again, government and central bank actions are critical, but it is unknown how the capital structure might evolve, and that uncertainty also requires elevated risk premia.
Figure 6. CP and High Grade spreads under pressure
Finally, the global downturn, even if it only lasts several months, is a negative event for the cash flows of companies and many sovereigns, especially those that are dependent on commodities or tourism (the positive oil supply shock will be a boon for many, in the long run). The downgrades have only just begun, and the large number of triple-B rated companies means that the fallen angel risk – forced selling by investment grade funds, and inability to absorb billions in “new supply” by high yield managers – is expected to cause some dislocations in pricing.
There are – and will be – bargains here, but substantial real credit deterioration is occurring.
With both investment grade and high yield spreads widening so dramatically, credit as an asset class has underperformed equities on a risk-adjusted metric, and so is pricing in more of the recession which is just beginning. Should spreads stay elevated, not only will defaults ensue, but the equity risk premium will be impacted as well (Figure 7).
Figure 7. HY spreads price in a more dire picture
Two opposite forces will drive sovereign yields, on the one hand massive purchases by central banks and on the other hand higher issuances related to contingency plans
The shock imposed to the global economic activity by the outbreak of COVID-19, plus the tensions related to the oil war between Russia and OPEC countries, has driven bonds yields to new historical lows in many developed countries. The yield of the global developed market sovereign index fell to just 0.2 per cent in early March (Figure 8).
Figure 8. Global 10Y yield Aggregate
Our bond volatility spillover index, which accounts for the way in which G4 sovereign bond markets impact each other, reached a peak of 60 per cent, higher than what was recorded during GFC or the euro zone debt crisis (Figure 9), with the largest contribution coming from US Treasuries.
Figure 9. G4 bond volatility spillover index
Taking the US yields as a reference, we observe that the decline of bond yields during this crisis has been characterised by two phases: from the beginning of January to the third week of February, yields were pushed down by a higher risk aversion, which triggered lower term premia; afterwards, with the spread of the virus in Europe, the dominant driver of lower bond yields has been monetary policy and expectations of lower risk-free rates.
Central banks’ response was timely and impressive. Government bond markets started showing signs of dysfunction in mid-March, as forced liquidations and margin calls triggered selloffs for even the safest sovereign markets.
However, large scale purchases by central banks subsequently saw yields stabilise and fall back in late March, with expectation of central banks moving to their effective lower bound (ELB) and ramping up asset purchases (Figure 10). The Federal Reserve (Fed) and Bank of England began to buy government bonds at a pace and size many times greater than in the GFC. The ECB has similarly committed to rapid and large-scale purchases. Meanwhile, the BoC, RBA and RNBZ have joined the club of quantitative easing (QE). The aim of QE is precisely to keep the bond term premia (TP) low so that global financial conditions remain easy (Figure 11).
Figure 10. Central bank asset purchases (USD)
Figure 11. Bond term premia (10Y maturity) and CB holdings
As a reminder, past Fed QE programmes’ cumulative effect in the US TP are in the range of 100-120 basis points on 10-year maturity. Estimates of the impact of ECB QE on TP are similar.
The fiscal support that is also coming will work out as an opposite force to asset purchases via supply of new bonds. A recent IMF paper estimates a baseline increase of approximately 20bps in yields for a one-percentage point deterioration in the fiscal deficit, which could rise to over 50bps if exacerbated by additional adverse factors, such as unfavorable initial fiscal conditions or weak institutions.
But the debt management policy will also matter and, so far, most countries have funded the COVID-19 fiscal spending mostly through the issuance of bills rather than bonds, probably due to the fact these measures are perceived as temporary. This should mitigate downside pressures on bonds for now. However, ultimately this issuance will need to be extended.
There are concerns that the announced fiscal stimulus plans will create an inflationary world. The effect on inflation will depend on which of the two shocks (supply-side or demand-side) prevails.
However, there are two important elements. First, the bulk of fiscal measures aims at addressing socio-economic impacts of the COVID-19 pandemic, essentially replacing lost income (via unemployment benefits) or ensuring there is no lasting decline in production capacity via public guaranteed loans. In principle, these measures should not be inflationary, and even more in a context of global downturn. Second, lower oil prices will exert downward pressures on inflation in coming months. We therefore consider that the risk of upside inflation surprises is limited, at least for the rest of this year.
A low yield environment should prevail, namely in a context of non negligible uncertainty and downside risks
As for euro government bonds, the €870bn of additional securities’ purchases announced by the ECB on top of regular QE will absorb the extra issuances even for peripheral countries in 2020.
However, we believe that a wider coordinated fiscal response is crucial to increase the effectiveness of the fiscal measures, in particular for weaker countries, starting with Italy. ESM Precautionary Conditioned Credit Line (PCCL) or “Corona bonds” are some of the measures under discussion. As always the devil is in the details, and the conditionality of credit lines might leave long-term marks on the Italian domestic political debate, and/or a relatively small size of the proposed instrument might be considered not enough by the market. But don’t forget that usually Europe advances in crisis time!
All these factors support our view of a prolonged low-yield environment. We believe that an additional deterioration in risk appetite, due to the materialisation of a more severe and prolonged global recession (our downside scenario), will boost demand for top-rated fixed income securities, triggering still lower yields. Expectations of Yield Curve Control from the Fed might push in the same direction. We therefore remain overweight bonds, more pronounced in US Treasuries which should offer more protection relative to others.
Emerging market debt
EM has been hit hard by the significant downgrade to the global growth outlook and central banks have responded quickly
Emerging markets – economies and assets – have been hit hard by the onset of the COVID-19 crisis and the rapid deterioration in the global growth outlook. For those with significant exposure to commodity (particularly energy) exports, the oil price shock has added to the pressures.
Even prior to the outbreak of the coronavirus’ growth in a few of the core EM countries was already low. Following global developments, these already struggling economies have seen significant downgrades to their growth outlooks.
Nonetheless, major emerging central banks have reacted and are expected to continue to provide ongoing monetary support. Figure 12 shows that emerging market GDP-weighted policy rates are 60bp lower so far this year despite having already fallen significantly throughout 2019, bringing emerging market policy rates to all-time lows.
Figure 12. Emerging market monetary policy (GDP-weighted; ex-China)
The willingness of emerging market central banks to cut rates highlights their desire to prioritise the real economy over potential currency implications. This is a notable shift in priorities relative to the global financial crisis (GFC) where currency concerns were more dominant. However, the extent to which they are able to pursue this approach may well be tested over the coming months, particularly as they see an increasing number of COVID-19 cases.
Indeed, emerging economies that rely on external funding will need to carefully judge their fiscal and monetary response against their perceived institutional robustness. Otherwise, the currency could move from being a pressure release valve to a destabilising force. Given the backdrop, our appetite for currency is limited here.
Liquidity continues to be a challenge and therefore any opportunities must be longer term in nature to be attractive to trade
Liquidity continues to be a challenge within emerging markets and therefore investment opportunities must be considered from a structural rather than tactical basis. At this time, therefore, investors in emerging market assets must focus on those economies that offer longer-term fundamental value and acknowledge that returns are likely to remain volatile over the near term.
Within local currency bonds, currency depreciation has dragged on performance and pulled year-on-year returns negative. However, some high yield rates markets are now approaching their post-GFC highs both in real and currency hedged terms and could start to look attractive.
Elevated yields in parts of the local currency universe, alongside attractive valuations for some currencies, suggest potential for above average returns once the external environment stabilises. Similarly, some investment grade sovereigns in hard currency (HC) look to have widened beyond fundamentals given the balance sheet strength they possess.
In the high yield segment of hard currency, return dispersion should remain elevated as vulnerabilities are exposed by the current situation in certain countries, while others have greater internal resources, and access to external support, to allow them to navigate through the ongoing challenges.
Liquidity continues to be a challenge and therefore any opportunities must be longer term in nature to be attractive to trade
While small pockets of value may be found, the overall outlook for emerging market assets remains challenged given our global economic expectations. Figure 13 and Figure 14 break down local and hard currency into their component parts and show how they relate to other global growth sensitive risk assets, indicated by Aviva’s risk asset growth basket.
Figure 13. EM LC vs Risk asset growth basket (yoy)
Figure 14. EM HC v Risk asset growth basket (yoy)
Emerging market spreads have historically had a positive relationship to the global growth cycle and generally drive total returns in local and hard currency, respectively.
While emerging markets, along with other risk assets, have already priced a material decline in global growth, we do not yet feel it fully reflects our expectations for where global growth will eventually bottom, and how long it takes to return to normal. As with other risk assets, emerging markets will likely rebound strongly when the growth outlook improves and concerns around the coronavirus spread dissipate.
Given the current uncertainty for when this will materialise and the more difficult liquidity situation in emerging market assets, we prefer to hold a neutral exposure at this time.
Like other asset classes the spread of the coronavirus and subsequent repricing of the global growth outlook has shaken currency markets. Volatility has increased dramatically through Q1 and the carry trades that looked promising at the start of the year have been shaken loose.
So far this year, nearly all major currencies have depreciated verses the US dollar, with many high yielding emerging markets depreciating by more than 20 per cent.
Despite the dramatic moves, central banks have rightly focused policy on the real economy, cutting rates and providing liquidity. Of course, it’s not just the economic implications market participants are considering in currency markets but the portfolio implications also.
Cross asset correlations have been nearly as volatile as markets themselves, making portfolio construction increasingly challenging. In crisis, cross asset correlations tend to move towards the extremes. Figures 15 & 16 show the 6m correlation between G10 currencies and global equity markets (MSCI ACWI).
Figure 15. USD, JPY, CHF, EUR & AUD correlation to global equity (rolling 26 weeks)
The yen has long been the risk reducing currency used in portfolios given its low median correlation to global equities. Even with a 5y window, the range of correlation shown as the shaded blue area in Figure 16 has been consistently negative.
Figure 16. G10 correlation to global equity (rolling 26 weeks, 5y window)
However, traditional risk reducing positions such as short USD/JPY have recently failed to protect portfolios. The breakdown in US dollar correlation to risk assets in March was sharp, rapidly changing from near neutral correlation to deeply negative due to the dollar shortage described above.
If US dollar correlation to risk assets persists in being lower than that of the Japanese yen, short dollar positions will be unhelpful to portfolios looking to limit downside as equities fall. Alternatively, investors can look to take a position against those G10 currencies which historically have had a positive correlation to risk assets such as AUD, whose correlation to global equities has strengthened over recent months.
Historically, the USD has been sensitive to significant changes in the global growth outlook, particularly vs EMFX
The year-on-year change in the US dollar has historically had a counter-cyclical relationship to global growth, with the dollar appreciating as the global economy weakens and depreciating as the global economy strengthens.
Figure 17 shows this relationship is particularly strong for emerging market currencies vs the dollar, depicted by the red line in the top pane. In a crisis, foreign direct investment and exports wane, portfolio flows reverse, and dollar debts become difficult to service. Therefore, it is unsurprising that the negative correlation of the dollar to risk assets should strengthen so significantly as markets price a rapid deterioration in the global growth outlook.
Figure 17. USD YoY vs Global Growth
When the global growth outlook and risk sensitive assets decline, the USD appreciates
Beyond the general “safe haven” bid for the dollar, there are a number of potential drivers that underpin the negative correlation to global growth and risk assets.
Globally, international investors hold more dollar assets than domestic US investors hold of foreign assets. When the global economy is hit and risk assets fall, foreign investors with hedged currency positions find themselves overly short of dollars leading to increased demand for them.
Because the size of the imbalance of dollar assets abroad versus foreign currency holdings in the US is so large, dollar buying flows by foreign investors far outweigh the dollar selling flow from US domestic investors.
In addition to this, emerging market central banks often need to sell dollars and buy their domestic currency to defend it in times of market stress; this stabilises markets but damages sovereign creditworthiness. Selling dollars changes the currency composition of their reserve assets and when this imbalance is corrected this can lead to the dollar strengthening against other G10 crosses.
Current YoY appreciation in USD has been small vs historic comparisons to growth shocks
As discussed above, major central banks have initiated swap lines with the Fed to alleviate immediate dollar funding stresses. However, when you look at the current year-on-year change in the dollar, whether that is against emerging currencies or G10, dollar strengthening could have a lot further to run if we are to see a global slow down anything like that of the GFC.
For these reasons we favour a long US dollar bias against more growth/risk sensitive currencies and restrain our exposure to the Japanese yen.