U,V,W,X,Y,Z…aka shape of recession and recovery to come

There is a lot of talk about what the shape of the recession and recovery will be. For better or worse, analysts typically use alphabet-based terminology to communicate their opinion and hence one can see a lot of headlines including the letters U,V and W (as well as L). So which letter will it be?

Let’s first start with what the letters refer to. In most situations - but not all, and partly not even dedicated Wikipedia page - analysts talk about the level of GDP or the level of output gap (rather than the growth rate of GDP). The letter then is meant to approximate the shape of the curve capturing the level of GDP over the recession and recovery period. So, for example, V-shaped recovery is meant to suggest that after steep drop will be equally steep immediate jump in level of GDP that will take us back to original level. All three of these aspects – steep jump, its immediate nature and reaching original level – are important feature of V-shaped recoveries.

While V-shaped recessions/recoveries were the norm historically, the recessions and recoveries in last few decades typically follow different script. First deviation is that the onset of recovery is less rapid. This means that rather the sharp edge of letter V we have the soft edge of letter U. In U-shaped recession and recovery a period of rapid decline is followed by period of subpar growth before recovery gets under way. Second deviation is assuming that the level of GDP is permanently lower due to the recession, probably thanks to recession destroying some productive capacity (e.g. bankruptcy of otherwise profitable firms or permanent loss of some human capital). In such situation the shape does not have the second half and we talk about an (tilted) L-shaped recovery: after dropping there is no jump reversing the drop. Finally, in some cases the first wave of recession is followed by second recession immediately after, with the experience of southern European countries a recent example of such phenomenon. To such situations analysts refer to as a W-shaped recession.

Another modification of the shape of recovery – one especially relevant for current situation – focuses on symmetry of the recession and recovery. The U-shaped recovery implicitly implies a symmetry in the recession and recovery parts, which in other words means that the rise in GDP during recovery is as rapid as the decline was in recession. This will almost certainly not be the case in current situation with record-breaking speeds of decline observed during the lockdowns. So a different shape-metaphor has been proposed for our recovery: a Nike swoosh. In this shape the recession is rapid but recovery takes time as in U-shaped recovery, and hence the shape is not symmetric.

So which letter fits the current situation best? The answer is neither, since this recession and recovery will be highly unusual. First, the recession is more rapid than anything we have previously seen, among other things reflecting the fact that it includes a large supply-side component in form of government restrictions on economic activity. Relatedly, absent a second wave during summer, there is going to be jump in activity once the restrictions are lifted and/or people feel free to resume their life. This means that initially the recession and recovery will follow a clear V-shaped script, with drop in first and second quarters followed by jump in third. However, in all likelihood the jump in economic activity will be smaller than then the preceding drop, so that the level of GDP in third quarter will be below the level in last quarter of 2019, potentially substantially so. Afterwards, we will likely be in our normal recession, which can have both elements of U- and L-shaped recovery. Of course, W shape is a clear possibility in current situation given that we might experience a second wave later in the year.

How would one describe such complicated profile in a letter? One option is an incomplete V (or small v) recovery, with the dropping part of V longer than the rising part. Another option is referring to current situation as VU-shaped recovery, with the initial V-shaped profile followed by more gradual U-shaped recovery. Obviously, an incomplete-V and VU-recovery recovery just does not sound so good and in any case does not completely evoke the shape. To better describe the shape one might want to abandon the alphabet for world of mathematical symbols and refer to the recovery as (inverse) square-root sign: normal square-root sign has a smaller dip followed by larger jump; our current situation will have the opposite. (Depending on the author, this profile can also be meant by the reference to Nike swoosh.)

Now you might thing why is it important if we use the correct metaphor for the recovery in general, and why it matters that there is an element of V-shaped recovery in otherwise a U-,W-, or L-shaped recovery. The reason is simple: the effect on expectations and corresponding surprises. Many analysts nowadays talk back against expectations of V-shaped recovery, and partly rightly so, since complete V-shaped recovery is almost certainly not in the cards. However, this might lead to the opposite excess: by ignoring the fact that the recovery will have an element of V shape, readers might be surprised in summer to see positive economic numbers consistent with a V-shaped recovery and confuse them with signs that the recovery will indeed be a fully V-shaped. By building in expectations that there will be a large, but temporary jump in level of GDP followed by more gradual return to pre-pandemic levels, one will be pre-conditioned to read the summer and fall numbers with extreme attention to size of the jump. 

This is also partly related to the why growth rates will not be a useful representation of economic situation during the second half of the year: while in third quarter we are likely to see a very large positive growth rate corresponding to a large jump in level of GDP – growth which in some countries might match the decline in second quarter - the level will almost certainly remain well below the pre-pandemic value. The unsuitability of growth rates is further amplified by the changing base in the calculation of growth rate: a drop by 10% followed by increase by 10% leaves us 1% below the original value.

This discussion even has a political dimension: in absence of second wave the third quarter GDP growth will likely be record-breaking. Intriguingly, the release of the third-quarter GDP falls on October 30, 4 days ahead of the U.S. presidential election. Expect a lot of big tweets about the best economic performance in the history of the United States, or maybe even the universe.

P.S.: Below is chart illustrating V, U, and L recoveries (as well as combination of U and L). 

 

P.P.S.: Here is the newest OECS forecast for euro zone GDP - the blue line follows the incomplete-V (or as they say, half-way V) recovery.

 

How would the optimal/fair monetary policy response to pandemic look like?

My starting point is with what the pandemic is in terms of macroeconomy. It is a huge and unanticipated wealth shock. We do not have to discuss the huge part, I believe. The unanticipated part is more controversial: if you would ask epidemiologists, they would (and did) say that what we are experiencing is a clear possibility. However, the macroeconomic consequences were completely and utterly unappreciated. I view it from my perspective of somebody creating macroeconomic stress-testing scenarios for living: not only is the current outcome way out of the range of outcomes consider by regulators such as Fed, EBA or PRA – even though they were not trying, with latest CCAR being twice worse than the great recession - but I am fairly confident that if I would propose drop of GDP by 10-20% over 2 quarters I would be laughed at. 

Given that, I think that this outcome was clearly not in subjective distribution of those creating and pricing financial assets or just any contracts, on either side of the deals. This means that the shock creates ex-ante relative winners and losers: in financial markets bond holders are relative winners, for most part, with the residual-claim holders (=shareholders) suffering (almost) all the losses and bond holders not bearing (almost) any of the shock; in rent contracts, say between shop renter and shopping mall, the shop renter bears (almost) all the costs of not having any business, while shopping mall renting the space out is protected by the contract. Of course, with defaults on contracts the bond holders and those renting out will suffer as well, but this is only in the extreme outcomes and in any case their loss is much smaller compared to the stock holders and shop renters. Given the generally unanticipated nature of the shock, this distribution seems unfair, and indeed, governments around the world try to address this in some form. 

Ideally, there would be a renegotiation of the contracts, with bond holders and shopping malls taking a hit. Of course, we know from the contract literature that this is hard, so will not likely occur on massive scale. However, I realized that the same thing could be partly achieved by central banks, or at least in theory it could. Take the debt contract. Debt contracts are almost always written in nominal terms, so decreasing the real value of money via unexpected inflation shifts some of the real loss to bond holders, at least those who have long-term contracts. So the central bank could bypass the problem of contract-renegotiation by announcing it will increase the price level by say 5% or 10%. In practice this might be easier said than done, but there is always helicopter money, after all: by giving everybody 5% of new money supply now, we decrease the value of those holing the fixed-nominal-value side of the contracts by 5%, distributing the costs of the shocks more widely, evenly and I would say fairly. 

I think the strongest case for this is at levels of whole countries, say indebted country like Italy. In the old pre-euro days, in response to this shock Italy would experience a bout of inflation. This would re-distribute from the bond holders (and deposit holders) to the other people, alleviating some of the unfairness in the distribution of impact of this widely unanticipated shock by shifting it from tax-payers to creditors of the state. Of course, Italy does not have the option anymore, and doing this via default or deposit tax is not the same as it affects only one part of the link (e.g. hits bond holders, but not their creditors, making them insolvent). And of course, leaving the euro has such a large associated costs that it outweighs the associated benefits, so there is no such option for Italy. Before someone concludes that this shows that having common currency is bad, it is worth pointing out that this shock is extreme in that it was not part of our prior distribution, plus one could (or should?) do this at euro-zone, so this is not related to common currency per-se. 

P.S.: Of course this is all closely related to price level or NGDP targeting.

The Cyprus lesson on banking union

Lots of authors pointed out that the Cyprus deal highlighted the fact the EU does not have a banking union and that it probably will never have proper banking union one. While this observation is correct, there is something more worrying than that. Seeing, how complicated it was for Cyprus politicians to come up with a some sort of a deal, and that they did so only once they were days away from financial and economic meltdown of their country, makes me think how everything would look like if there indeed would be a banking union in Europe.

First note, that even if there would be full-fledged banking union, the authorities would be surely loath to simply bail out Cyprus banks without imposing losses on the creditors. Indeed the set of rules to be implemented in 2018 for dealing with failing banks provides for sharing the costs with the creditors of the problematic institutions. This would mean, that even in situation with banking union, uninsured depositors would be hit, albeit (probably) to a lower degree. The big question is whether the Cyprus political system and, even more, the Cyprus society would accept this? One does not have to be skeptic to be hard-pressed to imagine this. The truth is that as in the present situation, Cyprus would not have a choice but to budge. Nevertheless, the resentment towards EU would be even stronger than now, because the deal would be completely (as opposed to partially) made in Brussels, a favorite target of public rage.

However, what about the situation when the sovereign has to capacity to protect the creditors in banks? It is hard to imagine that Germany would ever allow EU institution to impose losses on the depositors - reaction similar to the reaction of Belgian politician to demands of Oli Rehn would for sure be forthcoming. But similar logic goes for the less controversial case of other creditors (such as bond holders) as long as they would be concentrated in Germany. The EU institution would then find it hard to prevent mutualization of losses on the national level since it would not really concern it (its money would not be used).

What this analysis is aiming to show is that even if EU has fully functioning institutions of banking union, the link between sovereign and banks is unlikely to be completely destroyed. This is because EU institutions simply lack the public legitimacy of domestic political institutions and hence they decisions are viewed as decision of outsiders. It is akin to girlfriend of your brother making decisions about the organization of your common home – while she is related to the family, when push comes to show, she is not part of the family.

The need for smarter CDS

The financial crisis was blamed on two new instruments – collateralized debt obligations (CDO) and credit default swaps (CDS). More recently, the naked CDS were banned in response to Greece woes. In the first instance, the blame was partly warranted, since CDS with no collateral postings and no central clearing place (CDS dealt over-the-counter), allowed some financial institutions to take larger risk on their balance sheets than would be otherwise possible. In the second case, it is now clear that banning of short selling was an empty gesture with no real (positive) effect. But in a way, CDS played their role in Greece crisis, even though exactly opposite than their critics claim – they are guilty of omission, not commission. It is now clear, that Greece crisis was created long before it became apparent, in the pre-2008 years, when the country was allowed to borrow far more than it can handle.

The point is that CDS and especially naked CDS are perfect instrument for speculation. This is because the fact that they often do not require collateral postings and allow speculation where it was not before possible, since there were no securities to sell short. It is more than abundantly clear from Michael Lewis great book “The Big Short”, that CDS at one hand allowed some institutions to take more exposure than they would be (and should had been) able to do otherwise, but on the other allowed the few who were brave and farsighted enough to speculate against the subprime madness.

Indeed what was the problem before the crisis was that there was not enough speculation against the bubble. The number and purchasing power of speculators was not large enough, even after enhancement provided by CDS. This applies to Greece case as well, because if there was enough speculators exposing the clear truth, that Greece is heading towards default, during the 2000s, Greece would not be able to indebt itself to its current degree.

But can anyone blame them? Look at the current situation, when authorities want to avoid credit situation (i.e. official default) by any means. One of the reasons of course is that it would trigger payments on CDS. While this behaviour is perfectly rational from the damage-containment perspective, it comes at huge price into the future. No speculator sane enough will bet that any country will go bankrupt, knowing that there will authorities doing everything possible to avoid paying him. In economic language, this clearly points towards something we know very well in respect to monetary policy – time inconsistency. No matter what they say before, government officials will bail out their peers, because at the given moment it’s the right thing to do. That is, it is the right thing if we take in respect only current period.

If we extend this analysis to past, speculators were probably right to lend to Greece whatever its situation was, believing that default will not happen. (Of course, those holding the bonds and selling them were hit, but so far no owner of CDS was paid). What is then to be done? I propose (without any knowledge of the peculiarities associated) new generation of smart CDS. CDS, which would take into account the possibility of government-led bailout and would be payable even in such case. This would improve markets in the Arrow-Debreu sense by allowing speculators to avoid the risk being denied profit due to government action. Furthermore, it would directly price the advantage accruing to too-big-to-fail institutions and countries, setting platform for aligning incentives.  

  

Where would 1998 Paul Krugman stand on current debate?

Just finished reading Krugman's take on Japan's slump. It is very interesting article in respect to our curent situation, but one thing striked me as really odd, given what Paul Krugman articulates now. When he discusses the possible fiscal answer to the problems of Japan, he stresses that "how much stimulus is needed, for how long - and whether the consequences of the stimulus for governemnt debt are acceptable" has to be considered. Now, this is perfectly acceptable position. Only that Krugman nowadays advocates continuous fiscal stimulus and ridicules those, who talk about the danger to US fiscal position and reaction of bond markets. The first attitude, after considering pros and cons, can still pass his original analysis. It is the second attitude, which is questionable. Not that USA would face imminent danger of debt markets pullout, Krugman is probably right that it does not. But he should know best of all, that the sudden reversions in market perceptions of sovereing debts can be, indeed, sudden. One could argue, that Japan shows, that the levels of governmen debt can go to much higher levels without any concerns (indeed, japanese government debt realative to GDP was roughly similar to current position of USA) . But Japan is definately in different position. Most important, it runs continuously current account surpluses, so the governemnt debt is financed by home savings. These savings are in a way trapped in the country and thus have no other choice than to be invested in japanese bonds. America is in exactly oposite position, since it depends on China's central bank (and others) for its financing. This means, that it is effectively vulnarable to changes of position of one big player. Arguing, that it is unlikely to happen, especially to America due to its central position in financial architecture, is variant of argument, that financial crises happen only in emerging countries. Guess what: they don't.