Empirical analyses of the effectiveness of interventions in the foreign exchange market in Brazil.
| Autor | Oliveira, Fernando Nascimento |
-
Introduction
Our objective, in this paper, is to analyze the effectiveness of the intruments that the Central Bank of Brazil (hereafter CBB) adopted to intervene in the foreign exchange market from January 2006 to April 2016. The instruments of intervention we consider are spot market operations and foreign exchange swap operations.
A first step to analyze the effectiveness of the central bank instruments is to measure the effect of different instruments in the dynamics of the foreign exchange rate. To do this, we specify a stochastic process of the foreign exchange rate, introduce the interventions of the central bank and estimate the process with the interventions.
We follow a literature of interventions of central banks in the foreign exchange market (Sarno and Taylor; 2001) and another literature that models the dynamics of foreign exchange rates using continuous time models (Erdemlioglu et al.; 2015). This latter literature describes these models as diffusion or jump-diffusion processes.
In this paper, we model the dynamics of the foreign exchange rate as jump-diffusion processes, considering spot and foreign exchange swap interventions of the CBB as jump processes. In this strategy, the instruments of intervention are effective if they affect the conditional mean of the process of the foreign exchange rate. In doing so, we follow Sarno and Taylor (2001) and Edison (1993). (1)
To estimate the parameters of the dynamics of the foreign exchange rate, we use the work of Hansen and Scheinkman (1995). The authors obtain moment conditions that allow the use of the Generalized Method of Moments (GMM). To test the robustness of the results we use different specifications, sample periods and a more common technique of estimation that transforms the continuous process into a discrete process (Chan et al.; 1992).
Our results show that the instruments used by the CBB were effective throughout our sample period. The results are robust to different specifications, sample periods and to the techniques of estimation mentioned above. The results are in line with those of other papers that analyze the effectiveness of foreign exchange interventions in Brazil on the level of the foreign exchange rate, such as Meurer et al. (2010), Chamon et al. (2017) and Kohlscheen and Andrade (2014).
The remainder of the paper is organized as follows. Section 2 briefly discusses the literature on interventions in foreign exchange markets by central banks. Section 3 describes the policy of interventions of the CBB in our sample period and the data that we use. Section 4 discusses the models that we estimate. Section 5 presents the results of the empirical analyses and Section 6 concludes.
-
Review of the literature on foreign exchange rate interventions
There are two ways by which central banks can intervene in the foreign exchange market: nonsterilized interventions, those that alter the domestic monetary base; and sterilized interventions, which do not alter the monetary base.
According to Sarno and Taylor (2001), foreign exchange intervention carried out by a monetary authority can be effective in influencing the exchange rate if it affects its conditional mean or the conditional volatility of the foreign exchange rate. Sarno and Taylor suggest that the effectiveness of intervention depends on the degree of industrialization of the economy, the dependence on the international capital market and the degree of substitution between financial assets with the appreciation or depreciation of the currency.
Central banks may announce transactions in the foreign exchange market, known as the coordination channel (Bryant; 1995). They can signal the goals they want to achieve, to influence the formation of expectations (signaling channel) (Mussa; 1981). Finally, they may also vary the amount of financial assets indexed to the existing dollar in the market and, thus, change the relative value of international assets vis-a-vis domestic assets (portfolio channel) (Dooley and Isard; 1983).
Dominguez (1998) finds that intervention policies always influence the conditional variance of the exchange rate; effects on volatility depend on the type of intervention. Estimated regressions showed that hidden interventions increase the volatility of the currency, while those made public during the analysis period lead to a reduction in volatility.
Domac and Mendoza (2004) study whether the measures carried out by the Central Bank of Mexico and the Central Bank of Turkey impact the conditional volatility of the exchange rate, since the adoption of floating exchange rate regimes in these countries. They used E-GARCH model and their empirical findings are that the volume and frequency of interventions decrease exchange rate volatility in these countries. These results corroborate the theory that currency intervention does not aim to defend a particular exchange rate, but it can be an important ally to contain adverse effects of temporary shocks in the exchange rate on inflation and financial stability.
Empirical studies on the effectiveness of interventions in emerging economies may differ from the results of studies on the most developed economies. This difference is due to the complexity of exchange rate regimes, the sophistication of the market and the regulatory systems of foreign transactions. In Brazil, for example, intervention tools are quite atypical, not just the traditional instruments of intervention such as the purchase and sale of foreign currency in the spot market or changes in interest rates. The CBB also influences the exchange rate through foreign exchange derivatives such as foreign exchange swaps.
The empirical literature on the effectiveness of interventions of the CBB on the foreign exchange market is still rather sparse. Papers that are more in line with ours include those of Meurer et al. (2010), Chamon et al. (2017) and Kohlscheen and Andrade (2014).
Meurer et al. (2010) show that the interventions affect foreign exchange dynamics, however they are more effective in periods of lower volatility of the foreign exchange rate.
Chamon et al. (2017) estimate the effects of announcements of swap programs by the CBB from August 22, 2013 to December 19, 2013. Their conclusion is that the initial announcement generated a persistent foreign exchange appreciation in the first weeks. However, daily interventions that occurred in the rest of the period did not create any additional impact on the foreign exchange rate.
Kohlscheen and Andrade (2014) use high-frequency data to study the effects of currency swap auctions by the CBB on the spot exchange rate. The authors find that official currency swap auctions affect the level of the exchange rate, even though they do not directly alter the supply of foreign currency in the market. The maximum impact occurs 60 to 70 minutes after the initial official announcement of an auction, and again typically shortly after the results of the auctions become public.
Other papers analyze different effects of intervention of the CBB on the foreign exchange market, such as those on the conditional volatility of the exchange rate and on the level and volatility of futures of foreign exchange rate. In the case of the former, Oliveira and Plaga (2011) use daily data of nominal exchange rate and the interventions for the period from January 1999 to September 2006. In all sample periods studied, including those with currency crises, some instrument of intervention affects the conditional volatility of the nominal exchange rate.
Regarding effects on future contracts of the foreign exchange rate, Janot and Macedo (2016) estimate the effects of CBB interventions on the return and volatility of the BRL/USD future exchange rate using intraday data from October 2011 to March 2015. Their results show that interventions not anticipated by the market affect the level of exchange rate persistently. They also show that the effects vary with the size of the intervention, and that higher volumes of intervention affect the exchange rate more strongly than minor volumes of interventions.
Nakashima (2012) shows the existence of abnormal returns in future contracts of foreign exchange rate near the foreign exchange auctions of the CBB. On the other hand, Nogueira (2014) finds that interventions of the CBB had an impact of the foreign exchange future only when they happened in an unexpected way.
-
Data
To investigate the effectiveness of the CBB's interventions in the foreign exchange market we use daily data. Our sample runs from January 02, 2006 to April 20, 2016. The whole sample period contains 2581 observations. Figure 1 shows how the foreign exchange rate evolved in this period.
The first type of intervention that we consider is spot interventions. We use the official daily spot interventions of the Central Bank of Brazil. We observe 1,209 interventions of this type. Of these, 1,160 are interventions in which the CBB supplied dollars to the market and 49 are interventions where the CBB bought dollars from the market. Figure 2 shows the date and intensity of the interventions in dollars. (2)
The second type of intervention we consider is foreign exchange swaps. In swap operations, the CBB exchanges the first difference of interest rate in a certain period with the depreciation (or appreciation) of the foreign exchange rate in the same period. In these operations, the CBB can be either short or long in the depreciation of foreign exchange rate.
We use the net daily supply of swap auctions as the intensity of the intervention. Our information comes from the Open Market Department of the CBB. In our sample period, there are 853 interventions with this instrument. Of these, 721 are ones in which the CBB is short on the depreciation of foreign exchange rate and 132 are those in which the CBB is long on the depreciation of foreign exchange rate. Figure 3 displays the interventions with this instrument.
-
The dynamics of the...
Para continuar a ler
PEÇA SUA AVALIAÇÃOCOPYRIGHT GALE, Cengage Learning. All rights reserved.
