Print

Are we too optimistic about GDP growth in the second quarter of 2014?

Written by Armando Castelar.

This Friday, August 29, IBGE will publish its first estimates for Brazil’s GDP in the second quarter of 2014 (Q2 2014). Market analysts are waiting for the number with great expectation, as it may reveal whether Brazil has entered, or not, into technical recession (two quarters of negative growth). The results for GDP in Q2 2014 are important also because they will influence the overall result for the year. With consumer and business confidence at five-year lows, and credit and investment subdued while firms and workers wait for the October presidential elections, the economy is expected to expand only moderately in the second half of the year.

Most people believe that GDP contracted QoQ and YoY in Q2 2014, but there is substantial uncertainty regarding how bad this drop was. This is what show the results of weekly polls carried out by the Central Bank with market analysts to check on their forecasts for a number of economic indicators. Graph 1 reveals that market forecasts for YoY growth in Q2 2014 GDP have deteriorated continuously since mid-2013, with a more accelerated downturn starting in late May 2014. In the most recent poll conducted by the Central Bank among market analysts (August 22, 2014), forecasts for YoY change in Q2 2014 GDP ranged from -0.9% to 1.0%, with a median of -0.2% -- thus, quite a wide interval.

    Mkt froecasts GDP Q2 2014

A key reason why market analysts lowered their predictions for GDP growth over the last three months was the release of statistics showing a poor performance of manufacturing, retail and construction in Q2 2014. YoY, manufacturing output contracted 6.5% in Q2 2014, with potentially large impacts on wholesale trade, transportation and services provided to firms. Retail sales declined by 1.8%. The results were no better for construction: the output of manufacturing products used in construction dropped 9.1% YoY in Q2 2014, while retail sales of these products contracted 2.8%.

In light of these sharp contractions in manufacturing output, retail sales and construction, the median forecasts in the Central Bank’s poll seem somewhat optimistic. Indeed, the Central Bank itself recently published a much more bearish estimate of output growth in Q2 2014. This comes from an indicator of the level of activity calculated monthly by the Central Bank, known as IBC-BR, which has dropped 1.5% YoY in Q2 2014. As shown in Graph 2, the IBC-BR is a good coincident indicator of GDP: in the 10 years represented in the Graph (41 observations), the mean error was zero, with a standard deviation of 0.6 percentage point. That is, the IBC-BR suggests there is a two to one chance that the YoY variation in Q2 2014 GDP will be between -2.1% and -0.9%.

ibcbr vs PIB

It is important to take into account, however, that, as Graph 2 suggests, the IBC-BR tends to overestimate extreme GDP variations. In particular, in the four cases in which the IBC-BR signaled a contraction in GDP, it overestimated the fall in output by an average 1.0 percentage point. If this happens again, the YoY decline in Q2 2014 GDP would be closer to 0.5% than to 1.5%; thus, still below what market analysts anticipate.

A 1.5% YoY decline in GDP would translate into a 1.0% QoQ decline in Q2 2014 GDP, plus a downward revision in the QoQ change in Q1 2014 GDP, from 0.2% to -0.3%. This is a more bearish scenario than market analysts have in mind, but which is consistent with high frequency output indicators. Thus, QoQ, manufacturing output contracted 2.3% in Q2 2014, while retail sales fell by 3.1%. Moreover, retail sales of products used in construction dropped 5.2%.[i] In turn, a 0.5% YoY decline in GDP would translate into a 0.4% QoQ decline in Q2 2014 GDP, with a much less significant downward revision in QoQ growth in Q1 2014.

Taken together, these estimates should ring an alert that we may be in for an unpleasant surprise regarding GDP growth in Q2 2014. And for the year as a whole.



[i] All QoQ variations are seasonally adjusted.