beyondbrics

By Tony Volpon of Nomura Securities

Is Brazil facing a credit bubble? With credit to GDP rapidly rising (from around 25 per cent in 2005 to 47 per cent now) in an environment of rapid demand growth and massive capital inflows, investors are right to worry. But looking both at the data and the unique characteristics of Brazil’s credit market, we see no credit bubble – though we do see an overstretched consumer and under-appreciated downside risks to future economic growth.

The first thing to understand in the case of Brazil is that though all forms of credit are growing strongly, the largest growth is in so-called “earmarked” or government subsidized credit, mostly corporate lending by the giant state-owned BNDES development bank. While in April of this year consumer loans stood at R$582bn ($371bn), or 32.8 per cent of overall lending, earmarked loans stood at R$614.7bn, 34.6 per cent of the total (Figure 1). While rapid BNDES loan growth raises its own risks in the form of contingent government liabilities, these loans are far safer and cheaper than consumer loans.

The most unique and important characteristic of consumer loans in Brazil is their very high price. Average consumer loan rates as of April were 46.0 per cent a year (against, for example a 12.25 per cent policy rate and a 6.55 per cent inflation rate). These extremely high loan rates are a major impediment for any credit bubble because outstanding debt quickly exhausts disposable income. A credit bubble needs low – in real terms often negative – interest rates so that consumers can expand debt levels with little to no impact on income, especially in permissive lending environments where it’s easy to roll over debt. The most recent, relevant example was the low, often zero, interest rate “teaser” loans made during the subprime real estate bubble in the U.S.

Another usual characteristic found in any credit bubble is the link to asset markets. Cheap credit against liquid, long lived assets, such as real estate or equity, generates speculative, positive feedback between asset prices and the value of loan collateral, raising both over time. Rising asset values allow investors to increase indebtedness and buy more of the asset, further boosting prices. Again, in the U.S. real estate bubble this was key, further compounded by home equity loans that also boosted consumption, growth and income. In the case of Brazil most loans are very short (average loan terms are 567 days, or about 1.5 years) and taken out to purchase durable goods (such as cars). Real estate lending, for example, represent only 8.3 per cent of total loans.

Thus facing expensive, short term loans, it is difficult to get the bubble to float. Nonetheless exactly because loans are so expensive and short tenor, the average Brazilian consumer already looks overstretched. While it is difficult from the data to say how the loan burden is distributed along different income groups, looking at the aggregate data already shows the average consumer handing a little over 18 per cent of disposable income to pay interest (Figure 2). But when one takes into account the need to amortize these relatively short tenor loans, consumers may be already be handing over close to 30 per cent of their disposable income to service their debts.

While we see little risk of a credit bubble, the expensive, short term nature of consumer lending in Brazil means that the risks to growth are already high, and generally under-appreciated by the market and the Central Bank of Brazil (BCB), who seem exclusively focused on current tight labor markets as the major upside risk to inflation. While the current loan loss rates of 3.3 per cent is quite low, the current high debt burden means that debt levels should, from now on, at best grow in line with nominal GDP.

Faster growth will quickly translate into higher loan losses. As credit markets have been a big part of the Brazil’s economic boom, lower loan growth will contribute to moderate aggregate demand. But besides this “best case” scenario, the current high debt burden means that any negative unexpected shock to credit availability or income growth could, as consumers raise savings to pay back debt, add an important downward driver to growth. Brazil may not be seeing a classic credit bubble, but the consequences of its rising consumer debt burden could still be severely negative for future growth.

Related reading:
Brazil risks tumbling from boom to bust, FT
Brazil credit bubble fear as defaults rise, FT
Guido Mantega: What, me worry? beyondbrics
Keeping an eye on the Brazilian bubble, beyondbrics