چرخه های کسب و کار در بازار نوظهور

Emerging market business cycles feature a higher variability of consumption relative to output and a strongly countercyclical trade balance. An equilibrium business cycle model in which agents learn to distinguish between the permanent and transitory components of total factor productivity shocks using the Kalman filter accounts for these features. Calibrated to Mexico, the model accounts for the behavior of consumption and the trade balance for a wide range of variability and persistence of permanent shocks relative to transitory shocks. Estimation for Mexico and Canada suggests more severe informational frictions in emerging markets than in developed economies.

Learning about the “nature” of shocks plays an important role in explaining salient features of emerging market economy (EME) business cycles, namely, the higher variability of consumption relative to output and the strong countercyclicality of the trade balance. We establish this result in the context of a small open economy model in which the representative agent observes all the past and current total factor productivity (TFP) shocks and knows the stochastic properties of the distributions of trend growth and transitory components, but does not observe the realizations of the individual components. Using the available information, she forms expectations about trend growth (or permanent) and transitory (or cycle) components of TFP shocks using the Kalman filter.1 To reconcile the key differences between emerging and developed economy business cycles, the model features two different signals that reveal information about the permanent and transitory components of TFP. The first signal is total TFP growth and, therefore, in addition to revealing information, it determines the productivity of the economy. The second one is an additional noisy signal that reveals information about the permanent component of TFP, which is modeled as the trend growth shock plus independent and identically distributed (i.i.d.) noise. This trend growth signal allows us to vary the degree of information imperfection while keeping all other structural parameters unchanged including the parameters of the TFP process and the size of permanent and transitory shocks. The structural estimation suggests that the accuracy of the trend growth signals for Mexico is significantly lower than that for Canada. Starting from a baseline imperfect information model for Mexico and reducing the noisiness (variance) of the trend growth signal, the model moments move closer to those of developed economies regarding variability of consumption and cyclical behavior of the trade balance. This structural analysis shows that the degree of uncertainty that agents face while formulating expectations can help explain key differences of EME business cycles compared with developed countries. Earlier research found that the predominance of trend growth shocks is crucial to explain salient features of emerging market business cycles. In our setup, however, two key mechanisms are sufficient for the model to generate “permanent-like” responses even when trend growth shocks are not predominant. First, under perfect information, in response to a positive and persistent trend growth shock, the agent reduces her labor supply due to the wealth effect while increasing her investment. When the persistence of the trend growth shock is higher than a threshold (around 0.2 in our calibration), the decline in labor supply leads to a fall in output even after capital starts to accumulate. This leads the model to generate low correlations of output with consumption and investment. Under imperfect information, when a positive, persistent trend growth shock hits, the agent only gradually realizes that the economy was hit by such a shock. This, in turn, contains the fall in hours worked, preventing a decline in output. The second key mechanism that helps explain EME regularities is related to the TFP being modeled as trend plus cycle. In this case, the beliefs about the contemporaneous trend growth shock relative to the cycle shock can be higher even when the variability of the trend growth shock is lower than that of the cycle shock. Because when the agent observes a high TFP growth today, she assigns some positive probability to a negative cyclical shock that hit in the previous period and was not fully incorporated to the beliefs. A negative cyclical shock leads to a below-trend TFP growth on impact but in order for TFP to revert back to trend, it leads to above-trend TFP growth in the subsequent periods until it dies off. Since the agent takes into account such a possibility, it is optimal for her to revise her beliefs about the previous period's cyclical shock backwards in the opposite direction to the TFP growth observed in the current period. To further elaborate the second key mechanism, let's define total TFP as View the MathML sourceAt≡eztΓtα.2ΓtΓt represents the cumulative product of growth shocks defined by View the MathML sourceΓt=egtΓt−1=∏s=0tegs. z and g are Normal AR(1) processes. The growth rate of A can be written as View the MathML sourceln(gtA)≡ln(At/At−1)=αgt+zt−zt−1. Under the imperfect information assumption, the agent optimally decomposes the signal, View the MathML sourceln(gtA), into trend growth, gt, and change in the cycle, zt−zt−1zt−zt−1. 3 When updating the beliefs about the changes in the cycle,the agent finds it optimal to update her beliefs not only about the contemporaneous cycle shock but also its first lag. This backward revision has no implications for the already executed decisions in the previous period. However,it implies,for example,that in response to a positive signal,the agent may improve her beliefs about the change in cycle (that is, zt−zt−1zt−zt−1) by not only improving her beliefs about the contemporaneous cycle shock (zt) but also lowering her beliefs for its first lag (zt−1)zt−1). Therefore,a given upward updating of zt−zt−1zt−zt−1 can be attained by improving the beliefs about contemporaneous cycle shock, zt, by less than she would in a setting without the backward revision of zt−1zt−1. 4 The permanent component of TFP shocks captures major structural changes in the economy driven by policy regime switches such as trade or financial reforms (as in Aguiar and Gopinath, 2007). These changes are likely to have permanent effects on TFP, as opposed to business cycles that do not alter trend growth but simply are mean reverting fluctuations around a stable trend. Capturing these policy regime switches adequately requires an explicit modeling of these two components separately. In addition, Baxter and Crucini (1995) find that, in an incomplete markets environment, the effects of an international business cycle shock vary greatly depending on whether shocks are permanent or transitory. Their findings provide an additional rationale for the explicit modeling of trend-cycle decomposition as well as a reason for why the agents would want to know about this decomposition for both their economy and for foreign economies that they have financial linkages with. The motivation for introducing a learning problem to decompose total TFP into trend and cycle relies on the uncertainty surrounding the duration of structural changes in EMEs. Once a reform takes place in an EME, agents face a high degree of uncertainty as to when and if the next government will undo the reform.5 This view is also supported by the earlier literature on emerging market business cycles that hinged on uncertain duration of reforms, particularly in the context of exchange-rate-based stabilization programs (see, for instance, Calvo and Drazen, 1998 and Mendoza and Uribe, 2000, among others). In this context, this paper underscores that the uncertainty regarding the duration of these structural breaks contributes significantly to the salient differences between emerging and developed economy business cycles. Most time series data (particularly at high frequency) in EMEs are shorter than in developed economies making the informational frictions more acute. For example, the median length of quarterly gross domestic product (GDP) series available for EMEs is 96 while that in developed economies is 164 quarters (see International Financial Statistics of the International Monetary Fund (IMF)). Looking at employment, the median length for EMEs is about half of that for developed economies (44 vs 80 quarters).6 For EMEs, series such as Emerging Markets Bond Index (EMBI) spreads start as late as the mid-1990s or early 2000s. Moreover, yield curves are also short since most EME government bonds have at most a 10-year maturity. Some series, such as hours worked, are missing altogether for many EMEs. This paper primarily contributes to the emerging market business cycles literature including Aguiar and Gopinath (2007), Garcia-Cicco et al. (2010), Mendoza (1995), Mendoza (2010), Neumeyer and Perri (2005), and Uribe and Yue (2006), among others.7Aguiar and Gopinath (2007) show that introducing trend shocks to an otherwise standard small open economy real business cycle model can account for the salient features of economic fluctuations in EMEs.8 In order for the perfect information model to account for the two key features of EME business cycles, a high variability of innovation to trend shocks and a low autocorrelation of the trend growth shocks are necessary. The imperfect information model relaxes these assumptions considerably. In a related paper, Garcia-Cicco et al. (2010) argue that RBC model could imply spurious dynamics such as a need for highly dominant trend shocks as well as a near-unit root behavior of the trade balance—GDP ratio.9 The imperfect information model performs well on these two dimensions since it does not need to resort to highly dominant trend growth shocks, and it implies a stationary, downward sloping behavior for the trade balance—GDP ratio. This paper makes an important methodological contribution to a vast literature on macroeconomic models with learning. To our knowledge, it is the first paper to incorporate a learning problem with permanent shocks and persistent AR(1) transitory shocks into a dynamic stochastic general equilibrium model. In this literature, Boz (2009) investigates the business cycle implications of learning about persistent productivity shocks in the context of emerging market business cycles. Her model does not allow for permanent shocks. Van Nieuwerburgh and Veldkamp (2006) study U.S. business cycle asymmetries in an RBC framework with asymmetric learning regarding transitory TFP shocks. Their analysis focuses on whether learning regarding transitory TFP shocks can induce asymmetries in output growth over the business cycle. Another study on U.S. business cycles is Edge et al. (2007) who show that uncertainty regarding the nature of productivity shocks helps explain some of the U.S. business cycle characteristics. They model signals as trend plus i.i.d. shocks, whereas we model signals as trend plus AR(1) cycle shocks. Similarly, Guvenen (2007) studies learning about earnings utilizing a signal extraction problem with AR(1) plus noise shocks. In a parallel work to ours, Blanchard et al. (2008) follow a similar modeling strategy with trend growth and transitory shocks to explore the contribution of news and noise shocks to macroeconomic volatility. The literature on “news shocks” (for example, Cochrane, 1994, Jaimovich and Rebelo, 2009 and Schmitt-Grohé and Uribe, 2008, among others) emphasizes the role of expectations. As shown by these studies, the standard RBC model with Cobb–Douglas preferences implies counterfactual dynamics on labor supply in response to positive news shocks; labor supply drops on impact due to positive wealth effect—similar to the dynamics of labor supply in response to highly persistent trend growth shocks. Recent studies have focused on building frameworks that deliver empirically plausible dynamics of labor, for example, Jaimovich and Rebelo (2009) “quasi-GHH preferences.” Our analysis shows that an alternative modeling approach could be the introduction of learning in an environment with trend growth shocks. Highly persistent trend growth shocks have similar economic interpretation as news shocks, and the gradual learning in our framework leads to realistic dynamics of labor supply. Finally, this paper relates to Quah (1990) who aims to resolve the dispute about whether consumption is excessively smooth in the U.S. data. Quah (1990) argues that one way to resolve this excessive smoothness observed in the U.S. data is to assume that the labor income is an integrated process plus a trend-stationary one. Unlike us, he assumes that the econometrician is imperfectly informed while the agents are fully informed. He finds that consumption will appear excessively smooth to the imperfectly informed econometrician since the consumption decisions were made by the fully informed economic agents. The rest of the paper is structured as follows. The next section introduces the model, the information structure and the consequent learning process. Section 3 presents the quantitative results. Section 4 concludes and discusses extensions for further research.

When the agents are imperfectly informed about the trend-cycle decomposition of productivity shocks, and they solve a learning problem using the Kalman filter to estimate the components of the TFP, an RBC model of a SOE's performance in matching the emerging market business cycles improves greatly. The key ingredients for these results are the existence of trend shocks, the existence of transitory but persistent cycle shocks, and uncertainty regarding the decomposition of TFP into its components. Our analysis contributes to the emerging market business cycles literature, which has largely emphasized the role of financial frictions, terms of trade shocks, and trend shocks but overlooked the role of uncertainty and informational frictions. The role of uncertainty that the agents face while formulating their expectations about the persistence of changes in economy-wide TFP appears to contribute significantly to the emerging market business cycles. By introducing imperfect information and learning about the underlying fundamentals of the economy in a tractable manner, we open up a new line of research. For example, studying optimal policy (fiscal or monetary) in an environment where the agents learn about permanent and transitory shocks to economic fundamentals can deliver interesting insights. An application of the framework could be on uncertainty about the shocks to the commodity prices. Another useful application could be to build the signal extraction problem developed in this paper into a two-country environment allowing different levels of informational frictions across the two economies to explore cross-country portfolio allocations, consumption correlations, and other relevant issues.