‘Stay Defensive’ says Goldman Sachs, macro markets are pricing in a global recession and right now that is appropriate.
The sharp drawdowns over the past couple of weeks have meant that market-based benchmarks of cyclical growth or risk appetite are now more clearly at recessionary levels. Although some assets (the VIX, for instance) have seen moves similar to the depths of the GFC (Global Financial Crisis), most have not.
However, as Goldman writes, given the increasingly harsh quarantines being rolled out across the world, this does not yet seem an excessive market response.
Our latest global forecasts foresee a recession that is likely to end up being worse than not just the more modest global recessions of the early 1990s and 2000s, but also deeper than the early 1980s and the GFC.
Moreover, the uncertainties around the depth and duration of the hit to the global economy remain high and the momentum in our own, and other, economic forecasts continues to be sharply negative with downside risks. The two core messages in our Global Markets Comment three weeks ago were to stay cautious and defensively positioned in risk assets, and to position directly for the policy response, particularly in front-end rates. With the market already pricing policy at the Effective Lower Bound in a wide range of places, we think the opportunities from rate cuts are largely exhausted. But for other, unbounded assets, the question of how long to stay defensively positioned looms large. Given the amount of policy support announced each day – asset purchases, income replacement schemes and fiscal stimulus plans – a temporary bounce could happen at any time.
However, here we lay out key conditions for a more persistent positive view.
Since we do not think those conditions have yet been met, we continue to think a defensive posture makes sense. But it is also easier to imagine how those conditions might be met in the weeks ahead than it was even a week or two ago.
Why markets bottom in a crisis: reducing the deep tail risks.
To think about the conditions for a market trough, we need an explanation for what drives market recoveries in crises.
The core insight is that while conditions are deteriorating rapidly, markets find it hard to be confident in the limits of the damage and so put heavy weight on deep negative tail risks. Inflection points are often, in the first instance, about the market being able to put limits on those tail risks even before true recovery is visible.
The clearest version of this story is the historical evidence that equities, volatility and credit respond more to very early signs that growth is bottoming, while the major shifts in government bond and commodity markets seem to require something closer to a return to trend growth. When the rate of deterioration slows, even before real recovery occurs, the worst tails in the distribution can be priced out and that is usually enough for market recovery to begin.
In practice, in broad crises, there is often more than one kind of tail risk that the market is worried about and more than one kind of trough. In the GFC, for instance, the Fed’s response to funding stresses was mostly in place by the end of November 2008, the Chinese stimulus kicked in hard around the same time, but the rate of decline in US and DM activity did not trough clearly until March. Those assets most sensitive to funding and to China recovered ahead of those more sensitive to US growth as a result. But persistent recovery probably required the tail risks on all of these fronts to subside.
Conditions for reducing deep tail risks now.
Simplistically, for asset markets to recover sustainably from the current crisis, we think the market will need to be able to put limits on the tail risks that are currently centre-stage and for new tail risks not to surface. The broad challenge is that this crisis is unique in its source and speed – rather than starting in the financial system and emanating out to the real economy, it begins with a sudden stop in the real economy and works its way into financial markets – and it is still possible to imagine scenarios where the stop is longer and leads to much larger economic losses than in our own central case. At a high level, we see six conditions that we think would allow the market to define limits to that uncertainty on some key dimension of the problem.
1. A stabilization or flattening out of the infection rate curve in the US and Europe.
After a period of reluctance, the major DM economies have now moved towards much more stringent quarantine and social distancing restrictions. The outperformance of Chinese equities in February suggests that even without clear signs of an end to economic damage, markets may respond to signs that containment measures are beginning to work. Until we see durable progress on that front, it will be hard to anticipate that recent restrictions — and the hit to economies that they bring — might be lifted, which we think is a minimum condition for understanding the depth of the shock. At this point, infection rates are still rising rapidly in large parts of the US and Europe, and there are some worrying signs of renewed infections in parts of Asia.
2. Visibility on the depth and duration of disruptions on the economy.
As many of the major economies now move to more dramatic disruptions, the precise extent and timing of the economic damage is also highly uncertain. The hit to activity in China’s January/February data was extremely large. Will Western economies experience declines of similar depth? We would emphasize that duration is critical in this current episode. If disruptions are as large as we expect them to be, the difference between a 2 month and a 6 month period where they are in force is highly material in economic terms. Our current forecasts assume that the peak impact on US activity from the latest round of disruptions will be in April, with monthly GDP around 10% below where it would otherwise have been and a gradual recovery thereafter. But it will be hard to be confident in either the depth or duration of that trough without seeing signs that we are reaching one, particularly since there may be multipliers to the initial shock. It may be that enough progress on the infection rate or a sufficiently aggressive policy response is enough for the market to move ahead of this point. But we doubt that it will be able to sustain moves without validation from the macro data. With the data not yet reflecting the disruptions themselves, this point may be some way off.
3. Sufficiently large global stimulus.
Market troughs are characterised by both a stabilization in data but also a strong policy response that finally gets ahead of the deterioration in the data. In the GFC, for instance, the dramatic under-writing of credit risk by the Fed, a path towards strengthening bank capital, alongside monetary and fiscal stimulus (importantly from China, not just the US) were enough to stabilize parts of the risky asset complex in late 2008 (three months ahead of the equity bottom). The current policy response is also beginning to become more urgent and more broad-based. DM central banks have cut rates pre-emptively, initiated or expanded asset purchases and started or restarted lending programs to aid the flow of funding and credit. But the role of central banks as lenders of last resort will likely need to be broader and more unconditional. And while fiscal responses are now focusing, appropriately, on income replacement, business disruption and credit risks, with the overall US package now likely worth 5% of GDP, they still generally assume a relatively short period of disruptions to the economy. These are all moves in the right direction, but we doubt that they yet cover all that will be needed. More comprehensive backstops to business lending and perhaps to state and local finances are likely to be needed. In the process, we may find that not all governments can easily expand the public sector balance sheets if the size of the needed bailouts increases sharply.
4. A mitigation of funding and liquidity stresses.
Beyond the economic shock, funding stresses and liquidity problems have also taken centre stage, adding to the economic risks. The two problems are distinct. The funding issues, such as front-end Dollar funding shortages, are reminiscent of those from the GFC and central banks are quickly moving to reinstitute the programs and swap lines that stemmed that damage then. Market dysfunction has been as much about liquidity as funding, however. With post-crisis regulations reducing the ability of financial institutions to intermediate, many fixed income markets are struggling to process the large risk transfers that are required as investors adjust portfolios. In corporate credit, munis and parts of the Treasury market, market illiquidity has become a major issue. ETFs that provide daily liquidity against these markets have also traded at large discounts, reflecting the underlying difficulties of transacting. The Fed’s decision to resume Treasury purchases was aimed in part at alleviating some of these stresses at the heart of the bond market. And the BoE and ECB have resumed or expanded their corporate purchase programs. But unless central banks expand their role as “purchaser of last resort”, as our Credit team has termed it, many fixed income markets may remain dislocated.
5. Deep undervaluation across major assets and position reduction.
Everything has a price. Even without clear signs that the damage is coming to an end, if valuations come close to – or overshoot – plausible worst case scenarios, the risk-reward may be appealing even with high uncertainty. We have seen significant cheapening across assets over the past couple of weeks and pockets of deep value may now be emerging. But on a broad basis, we are not at Global Financial Crisis or overshooting levels yet. For example, in the case of EM FX, the median currency is now through ‘moderate’ levels of undervaluation (corresponding to early 2016 or autumn 2018) but we estimate a further 3% depreciation would be required to reach GFC-type ‘severe’ undervaluation. Similar benchmarking exercises for EUR sovereign spreads suggest that we are not yet in overshooting territory. Our Portfolio Strategy teams draw similar conclusions about equities. Valuations are rarely a sufficient condition for markets to recover. But an unusually large risk premium would allow investors greater comfort in terms of stepping back in if other conditions are fulfilled. The current crisis is forcing a transfer of risk on a number of dimensions, as investors cope not just with rising volatility and credit risk but with the failure of many traditional diversifying assets to offer real protection. Alongside economic conditions, market bottoms often occur only once assets have been transferred from weak to strong holders. That transfer creates the conditions for deep undervaluation, particularly where liquidity is a challenge. So positioning indicators that suggest that this process is nearing completion are also likely to be helpful.
6. No intensification of other tail risks.
The longer this crisis lasts, the higher the probability that other tails risks intensify further. A pandemic induced contraction in economic activity is different from typical economic recessions. For instance, the recovery from the GFC was slow and halting, in part because there was a large overhang of housing investment to work through, balance sheets needed re-building and the credit transmission system needed healing from the trauma of the banking crisis. In theory at least, the recovery from the Covid-19 shock should be more rapid given the absence of these factors – unless the contraction and market reactions causes other things to break. That’s what investors (and policymakers) need to be focussed on – both in order to evaluate the eventual recovery prospects, but also in thinking of the places where market dislocations can extend further and deepen the negative impacts across the economy. We would highlight the following areas, some of which are already in train:
(i) A further Dollar spike: a rush for safe haven assets, dollar funding shortages and reserve recycling have already caused a Dollar surge. An intensification of this move, with the CNY participating as well will cause a new round of tightening financial conditions across the world.
(ii) Even lower oil prices and their implications in EM. The outbreak of the Saudi-Russia price war has already put severe pressure on asset markets across oil exporters. However, in the event of a move to and through $20/bbl oil, especially if it is prolonged, we can expect to see more non-linear price action in EM credit, further market pressure on oil exporting countries, and the eventual adjustments in exchange rate arrangements in places like Nigeria.
(iii) Politics and its spillovers into sovereign spreads and funding. A public health crisis and a prolonged economic shock are likely to put governments under severe pressure, and where there are already pre-existing fragilities, this could create political crises that can exacerbate the already negative price action. This is a key risk in the context of Italy, Spain and peripheral Europe, especially if there are new migration crises; it could complicate the Brexit process further; cause a renewed worsening in US-China relations; and put pressure on EM governments where there are institutional weaknesses. In the US, state and municipal governments may also face stresses if support from federal institutions is viewed as insufficient to contain balance sheet pressures.
A sum of parts.
As in the GFC, sufficient progress on some may substitute for the others, and some assets may be more sensitive to one or other of these conditions. But a broadly favorable mix of these conditions will likely be required for us to turn more decisively positive. And like the GFC it is possible and likely that some issues will be resolved quickly and before others, and parts of the market may experience bouts of relief without all risky assets recovering.
The institutional memory from the 2008-2009 period means that Central banks have been quick to put in place facilities to limit funding stresses, so it is possible that these issues are mitigated and front-end volatility declines before the limits of economic damage are visible.
Or it is possible that a US intervention in the Saudi-Russian oil price war puts a floor under the commodity price selloff, and the pressure on related assets.
But the sequencing may also be different to the GFC because the nature of the shock is different. In the GFC, banks and the financial system were at the epicentre of the shock, and its effects rippled outward onto the real economy. In the current case, the shock is in the real economy with its impact reverberating across markets and financial system, and central banks may be better equipped to protect the financial system (as in the GFC) than replacing lost income across the broader economy in short order. Equally though there are other ‘magic bullets’ that were not available in the GFC – more effective anti-viral drugs or clinical treatments can take the pressure off health systems, and ultimately a credible vaccine will clearly have a dramatic effect on the virus and market sentiment. It is also likely that institutional capacity to respond to the unemployment surge will vary across countries.
The aggressiveness and effectiveness of the policy response, the varying constraints on public sector balance sheets and central bank remits and the political reverberations may ultimately lead to more variation in outcomes than the market is yet able to price.