By Rafael García-Tinajero, MPP’20, former Senior Financial Markets Analyst at the Central Bank of Mexico
During the summer of 2019, I had the opportunity to work with the Secretariat of Economic Development of the State Government of São Paulo. It was my first time working in Brazil, and it was my first time working on an economic development project, so this was an intimidating experience for me at the beginning. I decided to write this to help a future student in similar situations during her internship or maybe in her new job, so follow me through this journey.
The problem and objective were clear; access to credit in Brazil is expensive. We needed to find a way to make it less costly. Without the proper training, someone may think: “ok, let’s just set a maximum interest rate banks can charge for credit to borrowers”. However, interest rates are prices, and after micro I and II, we know that ceiling and floor prices affect the quantity supplied and create a deadweight loss. Setting a maximum interest rate can also incentivize banks to find other revenue streams like fees. For these reasons, I discarded this idea.
From my previous job at the Central Bank of Mexico, I knew that the State Government is limited in this matter as the Central Bank is in charge of setting the short-term interest rate, and usually, this short-term rate acts as a floor for longer-term rates. Banks charge a premium over the short-term rate based on the maturity of the loan, and risk associated with the borrowers, to name a few factors. But I also knew from micro classes that interest rates as prices are subject to information asymmetry. During this process, Konstantin Sonin’s course was beneficial, but I am not going to spoil all the fun from micro II, so I am going to skip it. LOL.
At this point, I already had a conceptual framework, but I didn’t have the data to see what was going on with the credit market in Brazil. When I started looking for the data sources, I realized it would be an endless task without making use of one of the cool things I learned in the intro to programming course. It took me one day to write a code to do the web scrapping needed for the data collection and a few hours to get all the data gathered by the algorithm.
I had everything that I needed, so what was the data telling me?
At the beginning of 2015, the annual growth of Brazil-based bank’s loan portfolio decelerated, and from the second half of 2016 till the beginning of 2018, it contracted. Over the last year, the loan portfolio expansion has started to accelerate, although it has been growing at a slower rate compared to the pre contraction period. Moreover, domestic credit to the private sector as a share of GDP went from 66.8% in 2015 to 61.8% in 2018, placing Brazil well below the median (88%), among the G20 economies.
Companies and individuals from the Southeast region have historically concentrated almost half of the loans local banks lend. This concentration is explained mostly by the state of São Paulo, as their companies borrow 0.82 cents of reais out of each real lend to Southeast region businesses. It is not that surprising, taking into account that São Paulo represents 32% of Brazil’s GDP.
The 2015-2016 recession and the increase in the cost of loans can partially explain the contraction of credit. Since the economy is not in a recession anymore, the critical factor in analyzing is the current high cost of credit.
Central Bank’s tight monetary policy helps explain the increase in the cost of credit. From 2014-2016 the rise in interest rates to the public was driven by a tighter monetary policy. However, since September 2017, when the easing cycle started, the cost of credit has been explained primarily by a risk premium. This risk premium is even higher than in pre-recession levels. Low market competition is not the main reason behind this rise in risk premium. The market concentration of banks’ participation to total credit, measured by Herfindahl-Hirschman Index, has decreased in recent years, signaling higher competition among banks.
I am going to stop here and continue in another edition of “The one when I used the Harris’ Toolkit®”. Hopefully, by this point, you already realized what you are learning is useful, and I haven’t even mentioned the caipirinhas the toolkit gave me access to.