After writing a post on the topic in the early days of the coronavirus outbreak, I have decided to rdo a part two about the consequences of the uber strict lockdown and future consequences.
So Argentina has, by many indications, one of the world’s strictest social distancing schemes: people have been under stay at home orders for over 100 days, all forms of travel are heavily restricted, if not banned outright, and wearing face masks outside the house is mandatory. Anecdotally, I have only left the house seven or eight times in over three months, longer than the time I had a wound in my foot and couldn’t leave.
Have these measures been successful? Maybe. The country has had about 70 thousand cases (daily graph here) of coronavirus and about 2000 deaths. The vast majority of cases are concentrated in just two provinces: the City and Province of Buenos Aires, which make up about 45% of the population. So far, the city had partially relaxed some of its measures in May, including allowing for daily walks and letting certain shops and restaurants to open under strict conditions; this coincided with a big jump in cases (mostly due to the authorities ignoring their own forecasts about the peak of infections happening by the end of May and early June), so the measures were rolled back. The lockdown has been extended and toughened, for this reason, until the second week of July.
Somewhat worryingly, 45% of tests come back positive, according to Our World in Data (aka neoliberalism dot com), which is probably indicative of severe undertesting, since the country’s tracking and tracing system is very recent and not as efficient as would be desirable.
The lockdown, and the government, have so far enjoyed widespread support, and social distancing measures have generally been sustained across party lines – most divisions appear regarding how to respond to the economic consequences. Despite support for the President having reached nearly 95% in the early days of the quarantine, there are signs that public opinion is slowly turning – extending a stringent lockdown only gets a small majority according to some polls, rather than near-unanimous consensus as it used to. This has mostly been due to exhaustion and to the economic fallout, which will be explained shortly.
This exhaustion also has an unseemly underbelly: the economy is devastated, and the poorest people in the country can no longer afford to stay at home and not work, under insufficient government aid and mounting debts.
Income and labor
For the last decade, both economic growth and the labor market have remained stagnant, while real wages remained basically constant, all three of which stedily and slowly deteriorated since the recession started in 2018 . This dynamic obviously worsened this year, although the data is still out; wages are growing below inflation (after a small recovery), and both activity and employment are still obviously going to fall, as even the most optimistic bounce predictions for May and June would still be a solid 20 points below 2019 and 15 below February.
So far, the coronavirus pandemic has not been good for personal finances, as might be expected. During the first quarter, the unemployment rate was already 10.4%; employment fell 0.5% every month since March (-3% throughout the year and -1.5% during the lockdown), and registered private sector payrolls fell behind 6 million for the first time in over a decade: 230 thousand people lost their jobs in April alone, and another 90 did in March. The bulk of these job losses seem to have come from the private sector (69%), and 26% from self-employed workers (public sector employment increased 0.4%, meanwhile). Some losses are due to seasonal differences in agriculture, fishing, and other resource-focused activities; nonetheless, almost all sectors of the economy have lost jobs.
A recent UCA report indicated that up to 860 thousand people may have lost their jobs during the lockdown (up to May, at least), and that 70% of those (602 thousand) were unregistered workers (who only make up 35% of the total workforce). That nearly 900k figure points to, if accurate, at least a 4% hike in unemployment during the second quarter alone. The government tried to prevent job losses by banning firing people and by providing assistance for wages; this has the rather unfortunate effect of only helping registered workers (about 50% of the workforce) while doing barely anything for self-standing workers (15%) or informal workers- who also have much lower incomes.
UCA also has put out one of their hard to parse poverty estimates (extremely petty critique here), at 45% (or 10 points higher than currently), and some of the country’s most respected experts claim it’s already gone over 40% for sure – for example, during May the poverty rate could be estimated at about 39.8%, using income distribution data and the official poverty line. This would be the highest level since 2007 (or 2003,poverty data is squirrely), and the biggest jump since both the 1989 hyperinflation and the 2001 collapse (more information here).
As for wages, 19.8% of workers in the Buenos Aires metro area had no income to speak of, and 44.2% had lower wages; 39.3% of workers aren’t working or were suspended, and 8.2% lost their job altogether. A recent survey points to 80% of the population having lost income during the pandemic; a later study points to 45% having done so afterward, and even more recent studies point to 53% of the population having endured “significant” income losses. Following official data, real wages have decreased for three consecutive months in May: they stagnated last year,and were growing above inflation until March ; then, they fell well behind (0.1% in wages vs 1.5% inflation) in May. Part of this was intentional: to prevent job losses, the largest unions and the biggest corporations agreed to a 25% pay cut for workers who weren’t going to work; still, this doesn’t bode well for the labor market as a whole. Nevertheless, this won’t nearly be enough even before 2020, income was already 20% lower than in 2018, and a similar 18% drop is estimated for GDP per capita after 2020. During the lockdown, 70% of the population made less than 20 thousand pesos (about 200 dollars) a month while income inequality skyrocketed.
Looking at the job market as a whole, it doesn’t get much better: a full third of companies are planning on laying off some employees after it is allowed, 70% are unsure about whether or not they will be able to afford any pay raises during the second semester, and 64% of workers are afraid of losing their job in 2020 – with 51% saying that they owuld be laid off if their employer was seriously affected by the pandemic.
Even if the labor market fully recovers, which it is probably not going to, there’s still the issue of debt: almost 90% of people fell behind on payments in May, and 86% in June especially utilities and rent (since evictions were also banned). This means that even if the job market and the economy recover, people are bound to not spend their money on consumption to boost the economy anyways – and a good signal to see how expensive stimulus might be is that two thirds of adults are already receiving money from the government.
Economic activity and real output
Economic activity will definitely take a beating in 2020. GDP had already gone down 5.4% in the first quarter of 2020 (that is, before any type of lockdown) and looking forward it doesn’t get pretty: there was a 10% annual drop in March and a 25% drop in April, leading to the same level of output as in 2004:
Economic activity, 1993-2020
The chart actually points to something interesting – since 2011, the country has basically remained stagnant, growing about 1% until 2017 and then de-growing 0.2% throughout the decade. Stagflation, the combination of no growth and high inflation, has obviously weakened the country’s ability to recover – and some even venture to say GDP might not exhibit any genuine growth until 2023, since April saw the largest drop in activity recorded anywhere in the last 120 years.
Only 28% of companies have been fully active during the lockdown: a quarter has had trouble paying salaries, 40% couldn’t pay suppliers or utilities, and two thirds have fallen behind on their taxes; 43% expect to lay off workers, and 88% expect to see profits fall. This points into a larger trends: businesses don’t expect to do well at all in this environment. Just going back to the February status quo is expected to take between half the year and 18 months, with a full quarter of companies believing it will take at least two years to recover. 65% of companies expect to be doing worse off in the third quarter than they were, with significant concerns over actually being able to pay off salaries. And some reports estimate that 30 thousand companies will go out of business in 2020, with chances of a recovery (per the latest Leading Index report) put at a meager 4%. Since March, in fact, 19 thousand companies have gone out of business, with about 290 thousand workers being affected in some shape.
The largest sectors in the economy
The activities that have seen the largest losses account for a large share of employment, and over 900 thousand have already been laid off. On a sector-by-sector basis, activities responsible for 70% of employment collapsed by almost 40% in March alone. Industry and commerce in particular have been hit particularly hard, with revenues (according to said report) plummeting by 50%, in sectors employing over a million people each (or, put together, about 35% of the entire workforce).
In fact, the most heavily affected sectors were construction and hotels and restaurants (the latter, for obvious reasons); the two largest sectors, retail and industry, have also taken a beating in 2020. So I’ll focus on those sectors (retail, industry, and construction) one by one, since after all, they account for about 40% of both GDP and employment; another 10% goes to farming, which will be examined more closely as it relates to food prices (obviously relevant to inflation), and exports.
Let’s start with industry. According to government data, industrial production plummeted by -27% in May and -33% in April, with the biggest drops in cars (-100 in April, since none were produced and -85% in May), steel (-72% and -56%), metals (-65% and -44%), textiles (-60% and -34%), and, with a huge chasm, chemicals (-11% in both months ) and food (-1% and -5%). Since these industries account for a majority of output in the sector, it helps paint the dire portrait it is going through right now: two thirds of factories could not operate normally in April and half in May, with a breakdown by sector that points to the trends previously exhibited. A recent report by the Industrial Union shows that production has decreased by 30% in April (and 17% compared to March), exports have plummeted by 58%, and employment has decreased by 2.4% (or about 40 thousand jobs lost in the past 12 months, with a tenth of that in April alone). Other estimates for industrial output, by FIEL (a well regarded think tank) has industrial output continuing to decrease in May, with an 18% drop (compared to -13% in April and -8% in March) – and with consumption goods staying strong (5% growth in durable goods and 10% in non durables) compared to cratering industrial input production (-13% in parts and -57% in capital goods); as a result, consumption goods have fared relatively well, recovering quickly and even thriving, but industrial inputs like parts and capital goods have plummeted by at least 20% in all three months. Nearly 60% of industrial capacity was idle in April, and that number soars to over 80% for steel and metals, and over 90% for textiles and cars – but food and chemicals still held strong. 80% of industrial companies expect layoffs, although they also expect a small bounce in activity moving forward.
Looking at the big sectors, it doesn’t bode too well either for the biggest parts of industry either. Car production was literally 0 in April, and fell by 47% in June, 84% in May, and 20% in March. Car sales also plummeted, falling by 45%, 74%, and 19% respectively (sales may have bounced 2% back in June, but they are still far behind 2018 or 2017 levels). Steel production fell by 30% in June, 50%the previous two months in a row, and 32% in March; and metal production plummeted by 23% in March, 50% in April, and 37% in May.
Industrial production, base 2016, and idle capacity (right axis)
Regarding construction, it’s in an even worse position. Construction activity plunged by 45% in May, 75% in April, and 45% in March. Home sales have plummeted, from over a thousand in March to just 7 in April and 700 in May; the supply of homes to rent is going down, while demand is going up, and a new rent control bill could, paradoxically, raise rents by up to 20% and already has increased them by 4% (double the monthly inflation rate) in June in anticipation; rent control measures will simultaneously reduce demand because of rent increases (“real” rent went down in 2019, as housing costs increase by only 33% compared to 55% annual inflation), and supply by creating enough restrictions to contracts. Employment in March has gone down a third in April, between 2019 and 2020, which means 65 thousand construction workers lost their job in 2020 so far (and that’s only counting registered workers) – and 57 thousand of them did so in March and April alone. For an indication of how things might move in May, both sales of construction materials and cement shipments fell 33%, although they both exhibited a significant recovery regarding April.
90% of construction companies have had their revenue reduced by at least 20%, and almost a third did not have any revenue, 86% rate the lockdown as (very) negative. It also has negatively impacted 94% of realtors, 90% of builders, 84% of architects, and 89% of developers.
Construction activity indicators, base 2012
Both construction companies and industries expect the third quarter to be a slight improvement, but 2020 would still be the worst year on record for both, with widespread closures and massive employment losses.
Retail sales have also been affected in extremely negative ways: as previously noted, up to 100 thousand stores may close this year, and 17 thousand workers have already been laid off. According to sectoral data, retail sales were 35% lower than in 2019 in May, and 50% lower in both March, April, and May – resulting in their lowest levels in 11 years. Supermarket sales stayed above water in March and April, mostly boosted by a big increase in February and March – but sales still crashed 14% in April. Shopping centers suffered greatly, with sales dropping 50% in March and 97% in April – and the sector estimates that 15% of all stores in shopping centers have already closed, and that 41% of stores in smaller malls may close this year.And speaking of stores, anywhere up to a 100 thousand of them may close during 2020 – so far, 24 thousand have. VAT revenue, considered a general indicator of how consumption is doing, has stayed below inflation for five months in a row, having accumulated an 11% reduction since February and dropping nearly 40% in real terms (adjusted for inflation, which is in the 45% range) for four consecutive months. Consumer confidence, in general, has deteriorated significantly since the pandemic outbreak began (with a small bounce in June), although trust in the government has exhibited the opposite behavior.
Argentina has had, for the longest time, an inflation problem. I don’t want to get too tangled up in it, but it probably has three main causes: the government, for a variety of reasons, has consistently spent more than it can possibly raise, therefore resorting to seigniorage systematically (i.e. printing money, MMT style); secondly, a cluster of institutional and historical factors make saving and holding pesos risky and unreliable; thirdly, and as a consequence, the country has an incredibly high demand of US dollars. This results in a perverse dynamic where monetary expansions that aren’t coupled with high interest rates (or just positive; the country has also had a tendency of having a negative real interest rate) end up draining the Central Bank’s reserves (under a fixed peg) or devaluing the currency (under extremely rare floating regimes); as a result, the exchange rate has an incredible amount of power in controlling the inflation rate.
After big devaluations in 2018 and 2019, the US dollar has been kept under control by increasingly stricter currency controls; at the same time, the Central Bank’s reserves have dwindled, from their highest levels in decades in 2018 to about a quarter of that now. The other side of this policy is that the real exchange rate, especially compared to leading trading partners like Brazil and Chile, has suffered and recent boosts to competitivity that improved the trade balance are slowly withering away (more on that later).
The other factor that has contained inflation is the quarantine: the demand of local currency is much higher because people are holding more pesos to remain liquid in times of great uncertainty, while the velocity of spending has dwindled due to, you know, nobody being able to leave the house. This, added to widespread price controls and utility freezes (to protect consumers), have kept price increases at a breezy 1-2% a month until restrictions are lifted.
The real question is: is this actual disinflation, or has inflation been repressed? Dynamics haven’t actually changed, and as soon as the lockdown is over, inflation will “reheat” as every single one of the measures that contained it is lifted or proven unsustainable.
The price controls, especially on food, have started taking their toll on producers, as costs have increased and many hadn’t increased their prices since January to boost sales in the summer ;though the government has recently softened the price caps, a quarantine that is sustained for much longer could risk food shortages. Utility freezes are a particularly thorny issue, since hikes are massively unpopular but the costs start piling up: in 2015 they accounted for three quarters of a 6.6% deficit, and since the exchange rate is lagging, soon the oil sector will lose profitability and imports will soar, worsening the trade balance significantly.
The second big issue is the currency market: exchange controls have made the US dollar “cheaper”, and the proliferation of parallel and black market rates has exploded the gap between the two (the “FX gap”) to the point where it was over 80%. This created incentives to buy US dollars for purely speculative reasons, even in the official market: the number of currency buyers doubled in May (from 1.2 to 2 and a half million), and in March and April it was still 50% higher than in the summer – the average purchase was USD 190, ten bucks below the maximum allowed amount (to a total of 500 million dollars form Central Bank reserves). A consquence is that a low real exchange rate both increases imports and reduces exports; this means that a strategy to protect the local currency will ultimately weaken it by draining Central Bank reserves (even with no domestic purchases) and hastening a devaluation.
Seignorage as a percentage of GDP (I didn’t make this one)
The final factor to keep in mind is monetary emission. The government has, in 2020, nearly doubled the monetary base, from 1.3 to 2 trillion pesos; most of this has occurred during the lockdown, resulting in an annual growth of 77%. So far the money didn’t cause any trouble; as soon as restructuins end, and considering the real interest rate is in the double digit negatives, the excess cash could simply go into the currency market until the Central Bank has no other resort than to devalue the peso. This has resulted in both grave concerns from economists and the government already promising it will “vacuum” these pesos and consider raising interest rates and reducing credit, which would have a negative impact on economic activity.
Repressed inflation is already estimated at 20%, and as soon as the quarantine ends the government will have to handle some incredibly tricky tradeoffs between inflation, output, and competitivity. A likely case is a big shock to inflation because the government would not simply allow a prolonged recession (which is already on track to be the worst in the country’s history). Currently, 2020 could reach a level well below 40%, with the nasty side effects being punted off into 2021 if the quarantine drags on long enough. So far wages have behaved properly (as mentioned, real salaries have decreased in line with wage-cutting agreements by the unions) but a high inflation, deep recession scenario would also restart struggles between unions and business leaders regarding price and wage levels – potentially, added to a big devaluation, spiraling the country’s price level out of control, but not into hyperinflationary territory.
During the quarantine, spending has increased and revenue has plummeted as a result of lower economic activity, so 2020 is on track tohave the largest primary deficit since the 14% recorded in 1975 – anywhere from 5.5% to 8% of GDP, defined mostly by how long the quarantine (and the corresponding stimulus) drags on.
Compared to last year, the deficit is ten times higher, having reached 260 billion, and overall spending has doubled to 597 billion. This has also meant that, as previously mentioned, the Central Bank has printed a trillion pesos so far this year to finance government sending, and the debt market has dried up. This has resulted in the Central Bank handing out 8% of GDP in monetary emission to the Treasury – which has been described as a ticking time bomb for inflation by most serious economists. Currently half of all spending is financed by printing money, which means the government will have a horrific choice once the quarantine ends: either cut back on spending to reduce its reliance on the money machine, or basically allow inflation to accelerate – and both options could nip any recovery in the bud.
Compared to most other countries, the fiscal stimulus package has been small: only 5% of GDP has been dedicated to counteracting the effects of the pandemic, consisting on 2.7% of fiscal aid and 2.3% of low interest credits to allow businesses and self-employed workers to pay off salaries and other expenses. This stimulus package has a vital flaw at its core: it doesn’t reach the right people, regardless of size. Currently, two main facilities exist: a very means-tested program for the government to cover up to half of an employees’ salary and a 10 thousand lump sum payment per household in which no adult is a registered worker. The obvious flaw is that this only covers the absolute poorest people and registered private sector employees, but nobody else – with a big gap that masks a variety of precarious economic situations that are not being addressed. Credit to firms, on the other hand, has been criticised by “friendly” voices such as ECLAC for being too small.
Why is the government only funding itself through the money printer, when most of its neighbours have taken advantage of low international rates to fund more generous packages through debt? Due to a high debt burden, the country has decided to restructure its sovereign debt (largely in US dollars) with creditors. The government’s plans pointed to wrapping up the issue by April; instead, after a disastrous first offer to creditors, negotiations dragged out for three months as the government unveiled bad faith proposals that under 40% of bondholders accepted, since they implied cuts of over 50% to interest payments. Finally, the country caved and unveiled a final (this time apparently for real) offer, which only reduces payments by 47%, saving the country thirty billion dollars (or around two billion per year for 15 years) and, contrary to the government’s own intent, payments start in February of next year rather than in 2023.
A worrying aspect is that and that the government played hardball for two months over an amount under 6 billion dollars – spread out over a decade and a half. The biggest concern right now isn’t that the country will default (it already technically has, since the April deadline coincided with the maturity of a bond that had to be paid) but rather how credible the restructuring offer is, especially on terms of future fiscal responsibility: creditors are wary that, as has happened three times already, if the current President doesn’t start making part of the effort to pay , then nobody ever will – leading to a fourth debt restructuring in 20 years.
2020 will have the largest trade surplus in 110 years, surpassing 20 billion dollars. This means that the Central Bank will increase its reserves, as exporters are legally bound to sell their hard currency (and they actually pay hefty export taxes, so the Bank makes a profit buying the dollars).
The problem with this is where this surplus comes from: for the last 2 years, imports have fallen at a breakneck pace, while exports basically stayed at the same level. This follows a longstanding pattern: when the country grows, imports (for capital goods and especially parts and inputs) grow too, and faster; if you add in a lagging exchange rate and fuel subsidies (which result in higher fuel consumption, lower local production, and higher fuel imports all at once), you can very quickly see a trade deficit coming. A second aspect is clearly that a lower real exchange rate reduces incentives to export, so the trade balance also worsens that way. If you break up exports and imports on quantity and price, you can see a worrying trend: the amount of imported goods decreased due to the recession, but the quantity of exports basically plateaued and then dropped because of lower output in the past three months.
The big risk moving forward is that, once the country starts recovering, a non-real trade surplus will vanish as imported parts pick up; there’s also the fact that two thirds of exports are industrial goods (about half of that agricultural products, mostly soy and corn derivatives, and half manufactured goods, led by chemicals and metals) and that commodity prices are, at best, not increasing, and at worst, collapsing.
Exports and imports (left axis, in millions of USD) and real exchange rate (base 2001, right axis)
The goose with the golden eggs is the agricultural sector, considering 55% of exports are either commodities or industrialized products; the problem is that high export taxes and an uncompetitive exchange rate (soy exporters got paid about 45 pesos per dollar, compared to an official rate of 70 and a black market rate of 120) have led everyone to put off sales, which have stagnated or even decreased in a similar, or even larger, harvest, to wait out the government until lack of reserves forces a devaluation.
The fact that many of the country’s trading partners in the region, especially Brazil, have sharply devalued their currencies and we have not is also cause for concern: as the real exchange rate worsens from domestic appreciation, it also loses its power if our trading partners are also more competitive.
Conclusion: an L shaped recovery
So far, the picture I’ve painted seems bleak. A stagnant economy who’s had the same GDP for an entire decade, while inflation slowly and steadily picks up, has to confront a massive economic shock. The 2018-2020 (at least) recession seems to be the longest one on record in the country’s history, going on for 10 quarters strong, and of an unprecedented intensity: the country seems to be on track for its largest ever drop in GDP, since the IMF expects a 9.9% contraction, the World Bank and the OECD about 8% each, and local economists predict a recession of 11.6%.
GDP, in billions of pesos, and annual economic growth
The problem is not so much the recession, which is kind of priced in at this point, but the recovery: the government is currently modelling its efforts on the 2003 program, where a vibrant external sector, high spending, and a primary surplus created by a mix of export taxes and frozen pension spending would result in a boom and “Chinese rates” of growth peaking at the double digits. This vision has a number of problems that make it unrealistic: for once, inflation wasn’t a problem back then, since it had been around 1% with the occasional deflation for a decade. The public sector also had half the size it currently does (23% of GDP vs 42%) and a balanced budget rather than the largest deficit in 40 years. The international panorama was also much better for export-led growth.
As previously stated, the fundamentals are also stacked any recovery achieved through consumption and government spending: the Treasury is all but bankrupt, and it has to actually renegotiate its sizeable debts with the IMF and the Paris Club as well; high debt means stimulus checks are going straight to landlords and utility companies for a while. The painful dynamics of inflation mean that seigniorage to fund spending is a non starter, and the Central Bank doesn’t have the firepower to sit on the exchange rate for another full year.
The likeliest scenario is more of the same: tepid growth only made possible with a frozen dollar and an overstretched state, while no real growth occurs and there is only a series of equally sized contractions and recoveries that are perfectly timed so the good parts are on election years.
If the country is to succeed, it will actually be forced to be sensible for once: sustainable fiscal policy, focused on small deficits or balanced budgets, and achieved through changes in the tax code and another reform to the massive pension system. The next logical step would labor reform, by allowing for simplified processes of hiring and firing and by creating incentives for employers to actually register their “illegal” workers. Reforming the external sector by keeping it competitive, and without preventing local companies from importing basic inputs they need is also a must to preserve Central Bank reserves and allow the government to pay off its debts.
Reasonable fiscal policy isn’t so much a matter of “running the government like a company” or “tightening the belt”: long term, only weaning the Treasury off the money printers and creating an actual local market for investment, debt, and bonds will stop prices from getting out of control.