How the financial market is trying to predict the elections

Financial market economists and consultancies are investing in tools to try to anticipate this year’s electoral movements. The effort includes contracting electoral surveys, some of them exclusive, building projection models, algorithms and survey aggregators, in addition to the help of statisticians and political analysts.

The current assessment is that politics will not be as important to the economy as the economy will be to the election result this year — given that the two candidates leading the presidential race have already held office before.

Even so, there is an electoral premium embedded in Brazilian assets, such as foreign exchange and the stock market, which tends to influence the timing of investment decisions.

Clifford Young, President of Public Affairs at Ipsos for North America, has been following the Brazilian elections since 2002. He has participated in meetings with market representatives and company directors, here and abroad, to discuss the electoral scenario.

Ipsos’ three main models use data from election polls and election results in multiple countries. The first shows that presidents with less than 40% of positive evaluation have less than a 50% chance of reelection (with 25% in Datafolha, Jair Bolsonaro would have less than a 10% chance).

The second points out that, in 85% of the elections, the strongest candidate in the main subject of the campaign wins. In 2022, this theme is the economy (inflation and falling income), which disadvantages the current president. The third is an aggregator of polls, all showing Lula at the forefront today.

“We created a database on elections all over the world. Our experience shows that the best forecast is the combination of methodologies. We work like this in Brazil, in the USA, in other Latin American countries, etc.”, says Young.

The most widespread tool among economists is the so-called survey aggregator, an average of the results of surveys already released. Some analysts use microdata from their own polls and hypotheses of abstention and migration of votes to trace probabilities of evolution of these scenarios.

The chief economist at Ativa Investimentos, Étore Sanchez, for example, gives more weight to face-to-face questionnaires than those made by telephone and the internet. It also uses information on vote certainty and simulates the impact of changes in the economic scenario on the result.

According to him, Bolsonaro appears tied with former President Lula in the first round only in a scenario of deflation in the coming months and with the hypothesis of a 50% readjustment in Emergency Aid.

“If the elections took place this weekend, the models point to Lula winning in an eventual second round, with small chances of being elected in the first round”, he says.

Independent asset manager Kinitro Capital developed an algorithm that helped anticipate the migration of votes to Bolsonaro in 2018, used in conjunction with surveys and analyzes by a team of statisticians hired especially for the election period.

Carlos Carvalho, director of Kinitro, says that there was an anticipation of the electoral debate in the current election and that the hired team will deliver weekly voting intention surveys from July and daily from August.

For him, the impact of this election on the market will be smaller in relation to previous elections, but there is a component of volatility in asset prices that should be present until the day of the vote. “The market is more concerned with the candidates’ campaign speech than with who will win”, he says.

Alexandre Lohmann, economist at asset manager Constância Investimento, which uses a poll aggregator, says that a view of the electoral process is a demand from clients and that the dispute already has an impact. [e adia] investment decisions.

Today, Lohmann sees an electoral prize that prevents the dollar from migrating towards the R$4.00 level, a value that is more in line with economic fundamentals and with the assessment that, whatever government is elected, the margin for a totally anti-government policy market is limited.

For Lohmann, that uncertainty will disappear as the October election looms and a candidate takes the lead in the polls. “We are going to have a better definition of the electoral scenario that can lead to a drop in premiums and an increase in the markets”, he says.

Gustavo Cruz, strategist at RB Investimentos, believes that 2023 will be a positive year for the Brazilian economy, regardless of who wins the elections, but that the election result can benefit one or another sector on the Stock Exchange.

“We take these data, these probabilities, and strategize based on that. If candidate A wins, we understand that the recommended actions must be from that group. If candidate B wins, it’s another group of actions.”

The large number of projection models, surveys and analyses, however, is no guarantee that the electoral scenario will be easily anticipated by economists and political analysts.

Silvio Cascione, director of the consultancy Eurasia Group, says that there is an increasing number of institutes coming to the public with different sampling methods, which has generated a large dispersion of research results.

A report released by Eurasia in April, based on Ipsos models, pointed to Lula with a 70% chance of victory, Bolsonaro with a 25% chance and a 5% chance for a third way.

“If there is a more competitive election than today, we will start to see some polls with Lula in the lead and others with Bolsonaro. This tends to increase the noise about the result”, says Cascione.

“How will the market react if you start to see a tighter election? There may be a little more volatility on top of the uncertainty about the electoral scenario.”