Clase de Mantenimientos por el tipo de acción

There are several applied studies that examine whether changes in interest rate significantly influences economic activity. For example, Bernanke and Gertler (1995) and Mojo and Peersman (2003) produce empirical works with evidences, which support that monetary policy operates through the interest rate channel. Bernanke & Gertler (1995) used the SVAR approach to analyse the effects of monetary policy shocks in the US economy. These authors applied a semi-structural VAR and identified the innovation in federal funds rate as the exogenous shock (i.e. a monetary policy instrument), and their system includes real GDP and GDP deflator as measures of economic activity. Bernanke and Gretler’s results show that output declines in response to a positive monetary policy

64 _{This view is supported by Post Keynesians such as (Arestis & Sawyer, 2006; Hansegenn, 2006; Palley,}

2003), as well as by mainstream economists such as (Bain & Howells, 2003; Goodhart, 2007; McCallum, 2001).

65 _{Money supply channels have been argued for by many authors; however, here we present (McCallum,}

2001, p. 157) who purported that the “model without monetary aggregates does not imply that inflation is a non monetary phenomenon.” Although the theoretical model seems to suggest misspecification, McCallum (2001) asserts that this effect is very minimal in quantitative terms. He therefore concluded his observation by claiming that ‘‘policy analysis in a model without money and based on interest rate rule is not fundamentally misguided.’’

shock. They indicate that the general price index responds after the fourth quarter; this seems to show that it lags behind the response in output. These findings confirm the operation of the interest rate channel through which monetary policy impacts the real economy in the US. On the same topic, Mojon and Peersman (2003) examine the monetary transmission process in 10 countries in the Euro area. Using the method of structural VAR, they evaluate cross-country differences in the transmission mechanism. Mojon and Peersman included variables such as world commodity price index, US GDP and short term interest rates.66 The world commodity index and the US GDP group are assumed exogenous and they are used to represent world inflation and capture the so-called ‘price puzzle i.e after a increase in monetary policy shock inflation goes up rather than going down’ associated with the VAR studies (Favero, 2001). Evidence from Mojon and Peersman’s study indicates that output and price react to random shocks from the interest rate. Firstly, output temporarily falls and reaches a trough around the fourth quarter, and thereafter, price levels decline in response to a positive shock in the interest rate. These results are compatible with the general outcomes of VAR studies. The stylized facts of the effects of contraction of monetary policy shock on output, prices and interest rate are aggregate output initial falls, which is represented by a j-shaped response; aggregate prices initially rose steadily and starts to decline after the 3-quarter; and interest rates initially rose in response to a monetary policy shock (Favero, 2001; Christiano, Eichbaum, and Evans 1996).

Regarding the research on developing countries, we selectively surveyed the empirical studies that examine the interest rate channel in economies with a fixed peg (e.g. Oman) and flexible exchange rates (India, Uganda, and South Africa).67 In this view, we present the work of (Al-Raisi, Pattanaik, & Al Raisi, 2007) in Oman, (Kapur & Patra, 2010) in India, (Antigi-Ego, 2000) in Uganda and (Smal & de Jager, 2001) in South Africa.

Al-Raisi et al. (2007) investigate the transmission mechanism in Oman, using two econometric methods which include the structural New Keynesian model with three equations (i.e. Output gap, New Keynesian Phillips curve and monetary policy reaction function), and the SVAR approach. The structural equation model and SVAR both produce evidence that suggests that changes in interest rates do not influence aggregate demand and

66 _{SVAR and semi-SVAR represent the structure of the economy. These can be recursive or non-}

recursive depending on the hypothesis being tested. Most SVARs are non-recursive because this represents a system of equation with each equation describing a particular structure of the economy.

67 _{Oman has a similar exchange rate policy setting to Namibia. India and South Africa both have monetary}

aggregate supply in Oman. They noted that these results are ascribed to the lack of responses by market-determined interest rate to interest rate policy in Oman. Al-Raisi et al.(2007) argue that the weak transmission mechanism of monetary policy is due to the fact that Oman does not have an independent monetary policy. This might be explained by the fact that households in Oman respond to monetary policy in the anchor country. In addition, Al-Raisi et al. (2007) also discover the evidences of the interest rate puzzle (or IS puzzle) and the ‘Phillips curve puzzle’ which are common occurrences in the analysis of transmission mechanism using the SVAR method. Interest rate puzzles turn up in other empirical studies about transmission mechanism such as Mojo & Peersman (2003) and Westerway (2002). The IS puzzle implies that an increase in real interest rate leads to an increase in aggregate demand instead of a decline; while the Phillips curve puzzle denotes the empirical finding of a negative relationship between output and inflation (i.e. prices increase when monetary policy is tightened). To solve the puzzles, Kim & Roubini (2000) and Mojo & Peersman (2003) included the current world oil price index in US dollars in the VAR model; while Favero (2001) included the world commodity price index to counter the problem of the price puzzle as exhibited by the Phillips curve relation.

Kapur and Patra (2010) applied the Generalized Method of Moments (GMM) to estimate the structural New Keynesian model in order to examine monetary policy effects without any reference to money supply in India.68 They modelled monetary policy within the so-called ‘live policy-making environment’ as referred to by (Westerway, 2002). The sample period is from 1997:2 to 2009:3 and the variables in the model are: GDP, GDP deflator, repo rate by the Reserve Bank of India, US Federal Fund rate and the World index on non-fuel commodity prices. Evidence from their model suggests that aggregate demand as measured by output gap reacts to monetary policy through the interest rate channel. Kapur and Patra (2010) also found that aggregate demand reacts with at least three quarters delay; while inflation takes seven quarters to react to a change in the interest rate in India. They conclude that monetary policy has an impact on real activity and inflation with waning effects in the long run. Kapur and Patra’s study resonates well with our thesis that embraces the consensus view which does not emphasize the role of money.

Antigi-Ego (2000) examines how interest rate compares with monetary base targeting as a monetary policy instrument in the Ugandan economy. He constructed a small structural VAR model that captures the structural dynamic features representing Uganda’s

68 _{GMM is favoured by economists such as Biha, Galles and Jondeau (2004) because it captures the forward}

looking component of monetary policy better than the OLS and VAR methods. The incorporation of forward lags makes the estimation to include beliefs about the future conditions of the economy.

economy. Antigi-Ego used the model to compare the monetary base and interest rate operating procedures for monetary policy with a sample from 1981:1 to 1997:4. The SVAR results indicate that the transmission effects from interest rate is rapid compared to the effects from base money. He claimed that it takes less than six months for a 1% rise in the interest rate to cause an approximately equal fall in inflation. Antigi-Ego reveals that base money is slower in Uganda and that transmission effects take a year for a change in base money to impact on the interest rate through the money market. Therefore, he argues that there is favourable evidence to support a move to an interest rate setting strategy in Uganda.

Smal and de Jager (2001) investigate the monetary transmission mechanism in South Africa with the aim of giving a description of how monetary policy has evolved in the past two decades. In year 2000, South Africa adopted the Inflation-Targeting in their monetary policy framework with the inflation target set in a range of 3-6 percent. Smal and de Jager’s macro-econometric model comprises three equations that define aggregate demand, aggregate supply and monetary policy rule to represent the reaction function of the South Africa Reserve Bank. In their model the repo rate is the monetary policy instrument by which the reserve bank influences variables such as money, credit and other asset prices. Smal and de Jager’s results indicate that the repo rate has a significant impacts on real output and inflation in South Africa. The study further shows that monetary policy effects are felt after four to six quarters which thus confirms the existence of the interest rate channel in South Africa.

Brischetto and Voss (1999) examined monetary policy effects in Australia using the structural VAR model similar to Kim and Roubini (2000). Their model includes variables such as World Oil price index in US dollars, Federal fund rate, domestic output, domestic price index, monetary aggregates, domestic policy rate, and exchange rate.69 The oil price index is included to capture anticipated inflation, while the Federal fund rate is included to control the response of domestic monetary policy to US financial variables. In this study Brischetto and Voss used the official cash rate as policy instrument which has been an official instrument over the sample period in Australia. Brischetto and Voss (1999, p. 1) described the results of monetary policy shocks thus: “it has delayed and gradual effects on the price level and small temporary effects on output”. In addition, the results are consistent with other empirical works in Australia and other similar economies.

69 _{The model by Kim and Roubini (1999) has the following variables: World oil price index in current US}