Brazil: Making sense of macroprudential measures

 EmailPrint This Post Print This Post

This post is a guest contribution by Arthur Carvalho of Morgan Stanley.

The following is a discussion between our Latin American banks analyst and two members of the Latin American economics team.

Jorge Kuri: Like many market participants, I am somewhat confused about the implementation of monetary policy in Brazil. The central bank has elevated macroprudential measures as important contributors to bringing back inflation to target. Yet, most of the macroprudential measures implemented so far are aimed at slowing bank lending and, regardless if they’ll prove successful in doing so, I still don’t understand how this can make a meaningful difference in reducing inflation. I simply can’t see how car loans, for example, are the culprit here. Isn’t the inflation problem a combination of expansionary fiscal policy, high commodity prices and growing monetary base?

Gray Newman: Jorge, that’s a tall order. Let me take it in parts. First of all, you are right that the inflation Brazil is experiencing today isn’t simply about auto loans. Indeed, a strong currency is playing a big role. In real, that is to say, in inflation-adjusted, trade-weighted terms, the Brazilian currency is at its strongest level in over two decades. And this means that Brazilian consumers are enjoying a level of purchasing power that many haven’t seen in their lifetimes. I don’t have to tell you how many new iPad 2s I saw last month in the waiting lounge in New York before heading to Sao Paulo the day after the iPad’s launch to make my point.

And you’re also right, in our view, to argue that the fiscal and quasi-fiscal authorities are also co-responsible for the strong demand story in Brazil. In my view, faced with strong demand, the fiscal authorities should be withdrawing stimulus by running an overall budget surplus; instead, Brazil continues to run an overall budget deficit, meaning the authorities keep adding stimulus.

But I am not sure where that takes us when looking at the central bank. After all, the central bank can’t set fiscal policy or quasi-fiscal policy for that matter – both are starting points for the central bank. Look at Brazil’s situation: you can also argue that if emerging economies weren’t bidding up the price of commodities, then Brazil’s currency might not be at a multi-decade high, boosting consumer purchasing power. But there is not much Brazil’s central bank can do about commodity prices or for that matter the accommodative monetary policy from the US which is also playing a role in ‘exporting’ easy money and consumption to Brazil.

Jorge Kuri: Okay, Gray. But why go after car loans? What’s wrong with the monetary policy instrument, i.e., Selic rates, that Brazil has used for years?

Gray Newman: I hear from a lot of investors that yearn for the days when Brazil hiked interest rates aggressively. But look, the world is different. It’s not the same hiking interest rates in 2004 or 2005 when your currency is near multi-decade lows as when you are in 2011 and your currency is at a multi-decade high in large part due to global demand for your exports and an extremely accommodative monetary stance in most of the developed world. If you simply hike interest rates today, you run the risk of seeing your currency strengthen even further, and that can drive a further wedge in the economy by boosting consumer purchasing power while damaging your domestic productive sectors – what we call the Growth Mismatch (see “Brazil: Growth Mismatch Revisited”, This Week in Latin America, March 14, 2011).

And let’s be clear here: I don’t think that macroprudential measures should be seen primarily as an alternative to interest rates. I think the central bank is genuinely concerned about the rapid growth in consumer credit in Brazil and is acting prudentially to reduce the risk of credit bubbles forming before they take hold. But the authorities also recognize that macroprudential measures do have some impact on inflation and hence must be taken into account when using more traditional tools of monetary policy like interest rates. And it certainly doesn’t go unnoticed that macroprudential measures don’t tend to drive the currency stronger the way that interest rate hikes can.

On car loans and inflation, look at it this way: I don’t care where the credit is being given – if you boost credit you free up disposable income to consume more elsewhere. It should come as little surprise that the area of greatest pricing pressure is in services. If you boost demand, goods prices rise less because you can largely import if there is a local shortage. It is much harder to import a service: you may decide to come to New York for a show or go to a hotel in Miami, but for everyday services in Brazil – eating out in a restaurant, hiring a babysitter – strong demand is likely to push prices higher. Except for a beef and produce temporary uptick at the turn of the year, most of Brazil’s inflation problem has shown up in services.

Jorge Kuri: OK, so you’re arguing that we shouldn’t read the central bank’s focus on macroprudential measures as a sign that it is no longer concerned about inflation, but simply an additional tool that is useful because it has less impact on strengthening the currency?

Gray Newman: Jorge, I think we will make an economist out of you yet. Indeed, Arthur and I think that many Brazil watchers misread the latest quarterly inflation report. There is a legitimate debate about the costs and benefits associated with macroprudential measures, but the fact that the central bank is working with additional tools beyond interest rates does not mean that the central bank has abandoned its inflation target. And the fact that the central bank is pushing back the time before it believes inflation will return to the midpoint (4.5%) of its inflation target does not necessarily signal any lack of commitment to control inflation. Central banks, including Brazil’s, have long held that a reduction in inflation should come while minimizing GDP volatility. Given the magnitude of the price shock at the beginning of the year and given sticky service inflation, there is ample support for the central bank’s focus towards inflation in a horizon into 2012 – beyond the eight months remaining in 2011.

Jorge Kuri: Going back to the macroprudential measures, our reading from the February credit data is that the measures implemented so far have failed to curb growth. Total loans increased by 21.0%Y in February, up from 20.4% in January and 20.6% in December. Seasonally adjusted, total loans grew 1.8%M in February, faster than in January (1.3%) and December (1.4%). Looking only at consumer lending, the same holds true – consumer loans grew by 19.3%Y in February, up from 18.7% in January and 18.8% in December. And specifically, payroll and automobile loans continue growing at the same brisk pace as in December.

Gray Newman: Jorge, I am glad you want to go through the numbers, because we think they have been misread as well. And I am even happier to call in our Brazil economist, Arthur Carvalho, who monitors these series closely.

Arthur Carvalho: Thanks Gray. I am glad to contribute to this lively discussion. Jorge, the outstanding loan growth is influenced by many things other than originations, such as delinquency rates and payback on loans. From the policy-makers’ perspective in fighting inflation, what matters is the amount of new loans given on the margin, which are stimulating aggregate demand. For instance, the outstanding loans series you cite could indicate a lower growth, given that there were strong paybacks. Nevertheless, if originations are strong, the aggregate demand is likely to have expanded further, despite the lower outstanding loan growth.

I do acknowledge that our consumer origination series does not include all loans to consumers. But a big part of what we are leaving out, namely directed loans, are really insensitive to monetary policy. We tend to prefer to focus on originations and on the series provided by the central bank because we think it is the most appropriate series to measure whether monetary policy is beginning to work or not. Of course, you are right; if originations slow, at some point we are going to have to see this feed through to the total loan series you are tracking. Our only point is that when you are looking for the first signs in the first months since the December measures were announced, we think originations is likely to be a kind of early warning indicator.

Jorge Kuri: Thanks Arthur. We normally also analyze credit origination data; however, I would caution you from looking only at origination – looking at both originations and balances provides a better picture. First, origination data in Brazil have several drawbacks: earmark credit is excluded from the series; leasing is also excluded, and this is important for automobile lending as it represents close to a third of all automobile loans; more so, leasing does not carry IOF tax so it may gain share as the government is using IOF as a lever to increase the cost of credit; and origination data are normally inconsistent on credit renegotiations and credit line extensions, both a form of new credit. Second, while you are the expert on understanding the policy-makers’ perspective, we disagree that what matters most is originations. Balances provide a better perspective of the net amount of money that is affecting consumption. After all, repayments – which can only be analyzed through balances – reduce the leverage in the system, and hence consumption power. Also, the central bank and the finance minister always talk about targets based on loan balances, not originations.

But still, even when looking at origination data, it is hard to conclude that credit is slowing down. Seasonally adjusted, consumer origination was R$70-71 billion in the three months (Dec-10, Jan-11 and Feb-11) following the first round of macroprudential measures, basically unchanged from R$70-71 billion in the four months (July-10, Aug-10, Sep-10 and Oct-10) prior to the measures.

I also don’t have much confidence that the latest round of macroprudential measures will slow down consumer lending (i.e., higher IOF tax on consumer lending). Let’s look at the numbers: the increase in IOF tax from 1.5% to 3.0% means that the monthly installment on an average consumer loan will increase to R$428 from R$424 before. I don’t see how an additional R$4 per month will impact demand, especially when banks are showing strong willingness to lend and consumers seem less sensitive to price increases, given the tight labor markets and high confidence levels. You are closer to the central bank; do you think it agrees with this assessment? Do you think that perhaps the 1.5% increase was intended as a reminder or warning to the banks that the central bank has the tools to slow down credit? Certainly there is a disconnect between the banks’ credit growth guidance (15-20%) and the government’s target (10-15%).

Arthur Carvalho: First of all, you are indeed right that loan originations have not reverted dramatically. In part we suspect that this is due to very strong consumer confidence, which is keeping demand high. But after a long period of sequential increases in originations, we are seeing a flattening of the trend. I would look at the series over the past two or three years – after a steady climb, originations appeared to have peaked between October and December and are gradually slowing. But second and more importantly, the interest rates for final consumer have increased substantially and the maturity has shortened. We’ve seen rates on auto loans go up by 400bp and not return, while the average increase in personal loans has been 800bp. And perhaps more importantly, maturities – which play an important role in driving the monthly payments -are down by ten months. Higher monthly payments mean that in order to buy the same good, there will be less disposable income after debt service, hence less consumption of other goods and services.

On your point about what the origination series leaves out; yes leasing is an important part of the outstanding loan book to consumer, but since measures implemented in late 2008 (reserve requirements for leasing funding) have caused it to fall sharply. New leasing had been much less significant than auto loans. Of course, with the new restrictions, banks could go back into the leasing market, but that would most likely just lead the central bank to increase further the reserve requirements there.

Finally, measuring the effectiveness of these measures individually can be misleading – there are compound effects. Given that these have not been used before, there is a lot of uncertainty around how strong these measures are, even for policy-makers. So I do believe that the central bank is taking a gradual approach to putting new measures into place, in order to avoid any sudden disruption of the banking system. That said, what has been done up to now has not been enough to slow sufficiently the consumer demand for leverage. That sets us up, from our vantage point, for more measures.

Jorge Kuri: Thanks Arthur. I guess we’ll just have to wait and see. I am of the view that macroprudential measures work when credit slows down, not when prices and maturities adjust – especially at a time when many Brazilian consumers are enjoying all-time high purchasing power, as Gray mentioned. After all, loan originations remain at the same pre-December levels even though interest rates on personal loans have risen 800bp and maturities have fallen 10 months, as you pointed out. I would have expected a serious dent on origination from this. Separately, I appreciate your comment about not looking at the measures individually, but my sense is that banks look at them this way. In other words, they do blocking and tackling on a measure-by-measure basis. More than once I’ve been told not to worry about the 1.5% increase in IOF as the impact on the monthly installment – the ultimate measure of affordability in Brazil – will be fully offset by adding an extra 5-10 days to the average credit maturity, an imperceptible change for the borrower. Hence, I worry that tougher macroprudential measures may be around the corner.

Gray Newman: I can see from your vantage point how you could read the measures to date as having little or no impact. We think they have begun to work and we believe we are seeing the first signs. But I think where we all agree is that more is likely to come on the macroprudential front. How does that make you feel when looking at Brazil’s banking system?

Jorge Kuri: We remain cautious on banks until there is more visibility on the scope of policy actions or valuation multiples that compensate investors for earnings risk. We are worried not only about tougher macroprudential measures that may seriously constrain volume growth and margins, but also about asset quality deterioration. If macroprudential measures prove successful, consumer credit will likely not only be scarcer, but also more expensive and with shorter terms – hence negatively affecting affordability, which is already at risky levels. For some consumers this may limit their ability to roll over expensive debt, increasing the chances of default.

Gray Newman: If that happens, Arthur is going to have his hands full. My bias is that as long as the globe remains supportive of Brazil’s multi-decade high terms of trade and consequently multi-decade strong currency, the authorities are going to be both fighting the flows as well as the consequences of overly strong consumer demand. And although the response to date has been measured, I think the risk of overreach remains.

Arthur: I look forward to seeing you both in Brazil again. I expect we’ll have more measures and more data to discuss by then. Thanks for welcoming me into such a lively debate. Is this how you always treat economists or just the new kid on the block?

Source: Arthur Carvalho, Morgan Stanley, April 19, 2011.

Did you enjoy this post? If so, click here to subscribe to updates to Investment Postcards from Cape Town by e-mail.

OverSeas Radio Network

Leave a Reply

You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>




Top 100 Financial Blogs

Recent Posts

Charts & Indexes

Gold Price (US$)

Don Coxe’s Weekly Webcast

Podcast – Dow Jones

One minute - every hour - weekdays
(requires Windows Media Player)
newsflashr network
National Debt Clock

Calendar of Posts

April 2011
« Mar May »

Feed the Bull