Bound to Rebound — what my econometric model says about Brazil’s GDP for the coming years

Adilson S. Proença
7 min readFeb 11, 2021
Brazil Long-Term Real GDP Forecast 2010–2025 (Adilson S. Proença)

Over the course of the last 12 months or so I have attempted to forecast Brazil’s Real Gross Domestic Product four times. In September 2019 my bet was anything between 1 and 1.5%; in February 2020 my model suggested a 1,2% uptick; in July that year it had plummeted to -7,05%; and now those very same variables suggest a milder fall at -5,50%. Why’s that?

In this article I aim at two goals. Firstly, I am going to talk you through the milder fall projected in comparison to July’s 2020 nosedive and, secondly, I am going to discuss the downside risks present. That is: what can go South and worsen the scenario forecast by the model.

Why a milder fall?

In one of his manifold books, the Nobel Prize winner Milton Friedman comments on the relevance of money in an economy. Asserting that seems to be as evident as the fact that we need oxygen to live. Yet, despite how evident that might be, one can only graps how indispensable a gasp of oxygen really is — and, by extension, how indispensable money can be for an economy — when one has no oxygen at all. This scarcity does not even need to mean complete scarcity; total inexistence (no money whatsover). All one need is fewer banknotes moving around to see just how deeply asphyxiated an economy can become.

That said, when I updated my econometric model in July 2020 to better reflect the downside effects that a then raging worldwide pandemic had cast over Brazil, the explanatory variables that made up my model (i.e. those which explain the past behaviour of the variable you’re trying to forecast) captured a gloomy horizon: an abrupt reduction in overall economic activity (all sectors); the ensuing rise of the unemployment rate; the consequent drop in consumption spending; the final contraction of GDP. That was the horizon towards which pointed that model if no counter measure in those early dark days were undertaken in order to lessen the certainly negative impacts of the pandemic. Those counter measures need to be highlighted.

They came about mostly in the form of a multi-billion injection of money into the Brazilian economy. From the emergency financial relief installments — which helped sustain a minimum level of consumption for a significant share of the population — to the relaxation of labour contracts — which contributed for fewer dismisalls and contained the unemployment rate. Those were the main measures responsable for cushioning the shrinkage of Brazil’s GDP in 2020. Study the numbers below.

GDP Components in % Participation on Total GDP, average 2010–2019 (Adilson S. Proença)

The pizza-shaped graph above showcases the percentage participation in GDP of each one of its components. It is frequently by summimg up these four components that one finds a country’s GDP, and I have used them in my econometric model. The components are: household consumption; private sector investiment; government spending; net exports (i.e. everything a country sells to foreign countries minues everything a country’s residents buy from other countries).

What stands out the most from the graph is just how understandably representative household consumption is over total GDP. In Brazil, for instance, in the 2010–2019 average, that share was as high as 63%. What that effectively means is this: over half of the income generated by the country stems the everyday expenses carried out by Brazilian households.

From that one can infer the relevance of that share of sepnding in the sense that events — such as the pandemic — can substancially affect the level of spending of regular families and that in turns spills over across other sectors. Another important share of GDP is level of spending of the private sector. When times are uncertain, companies tend fo roll back on their expenses. What that means is a highe level of people being made redundant, which only adds to the number of affected families that will consume fewer goods and services.

Government spending is no less representative and important and it’ll be fruther discussed in the upcoming paragraphs. Net exports, in absolute and aggregate terms, has put out a negative number, but that only means that under the period analysed (2010–2019), Brazil has spent more buying from other countries that it has made money by selling to other nations.

A further analysis of the components of GDP is looking at how much they individually vary from a given period to the another. In the bar and line graph below, I have showcased them on a anual basis starting as early back a 2010 up until 2019 (pre-pandemic levels) and then further into the future on a 6-year long-term forecast from 2020 to 2025. Taken together, these three main components — household spending (C) + private sector investment (I) + government spending (G) — represent nearly the totality of the Brazilian GDP and they’ve shown negative variation in the 2015–2016 economic downturn, when GDP plummeted by 6,8%.

For 2020 the econometric model has forecasted that household consumption, private sector investment and government spending contracted by -5,80, -9,10 and -2,70% respectively. The latter, in particular, would certainly have been lower (that is, less spending) if the pandemic hadn’t broken out and that’s because Brazil’s fiscal policy programme was supposed to roll back on its spending. Cirscunstances, however, have forced the government to do just the opposite. These components, the model suggests, tend to bounce back to pre-pandemic levels starting already in 2021, but, considered together (GDP), they shouldn’t recover fully before 2022–2023.

Downside risks — what can go South and negatively impact the forecast?

Fiscal Policy

Brazil’s fiscal policy (expenses vs income) has reached alarming levels. Unsurprisingly, this fact has worsened significantly throughout 2020 as the government needed to step up and increase its level of spending in the form of the billions of Reais handed out in financial relief packages. In practical terms, the deterioration of a country’s fiscal policy casts doubt on its solvency capacity (paying back its debt). This situation can trigger a series of macroeconomic events highly detrimental to a coutry’s long-term economic situation such as a spike in interest rates.

To illustrate that, the graph below depicts the ration between expenses and income (GDP). In other words, this means how much of its generated income in a given year a country spends. If, for instance, the ratio is 50%, it means that the country spent half of all the money it made. The graph brings in numbers of Brazil in comparison to a group of other South-American economies.

Fiscal Policy: Brazil, Argentina and other South-American economies (Adilson S. Proença)

What stands out the most is this: throughout the whole period under analysis, both Brazil and Argentina have shown expenditure levels well above that of all the other countries and even above the average (the thicker red line). the forecasts suggest that Brazil must have ended 2020 with a Debt-to-GDP ratio of 92% while Argentina must most likely have burst the 100% threshold! In additon, the coming years for both these economies suggest that a new age of worringly high Debt-to-GPD ratios has arrived!

The red dotted line across the graph is the threshold at 64%. A few years back the World Bank published a paper that sought to understand if there existed a a level of indebtment of a country in relation to its output (GDP) that could be deemed “critial”. In a nutshell, the study suggests that for emerging economies that level is somewhere at a rate of 64%. That is: when an emerging economy has spent 64% of all the money it can generate, this economy can be potentially affected in the long-temr.

With that in mind, observe just how Brazil bursts through this 64% threshold as early back as 2014, a moment during which the 2015–16 crisis was begining to unfold. Argentina bursts the threshold around 2017 — certainly Macri’s most challenging year. Uruguay nearmisses the threshold in 2014 and the data suggests that it has most likely burst through it in 2019.. As for the remaining countrys, they all present an ascending trajectory, but none seem to burst through the line, with the expection of Colombia and Bolivia which, suggests the model, should do that within the coming years.


There can be no going back to normality without a vaccine whose levels of effectiveness and supply are abundant on a global scale. Judging by what’s happend over the past weeks, I’d say that we’re in the beginning of the end, but that, unfortunately, does not mean that the final lap will be trouble-free. An obstacle (politics- or economics-related, or both) to the process of development and distribuition of the vaccice will have the practical effect of slowing down the pace of the recovery forecast by the model — which projects that Brazil should reach its pre-pandemic levels no earlier than 2022–23. If that unfortunate scenario comes about, then we witness more “U”-shaped recovery rather than the “V”-shaped one currently forecast by the model.

Presidential Elections in 2022

The upcoming preisdential elections will either be a push (as it usually is the case) or a drag to the rate of the recovery. In the months leading up to the election day there usually is a greater level of goverment-led investments in the economy. We can expect that for 2022, although I’d say it won’t be that big of a party given how badly affect Brazil’s fiscal policy already is. Additionally, one has to factor in the internal political strife that’s been taking shape for some time now. In the worst case scenario, that can bring about a troublesome political transition (something like we’ve seen in the USA recently). If that happens, economic recovery wil be significantly impacted given just how much it looks for a contry that apparently cannot ensure a peaceful political transition.



Adilson S. Proença

An International Relations degree holder; a language, history and economics aficionado; and a soon-to-be Economist who sees writing a thought-untangling act.