Chapter 1: Introduction & Significance of the Thesis

1.1Significance of Research

Accurate estimation of national output and particularly that of potential output ofan economy serves pivotal importance in business cycle theory. Actual and potentialnational output also serves as important components of Monetary and Fiscal policyanalysis, implementation and assessment. Central banks around the world attributegreat importance to national income accounts (NIA) as they serve as the core inputdata for policy making. Estimating the Output Gap in the short run provides insightsin the aggregate supply and aggregate demand in the economy which directly impactactual and expected inflation. While in the medium to long run information in thetrend and magnitude of the Output Gap provides valuable policy input on the state offactor utilization (labor, capital) and impact of technological advancement on output;hence the growth path of the economy. Considering the importance of this topic,various elaborate and intricately-sophisticated economic and econometric methodsare employed in order to draw statistically relevant estimates of Output Gap.Estimation of Output Gap is of great importance because; a positive or negativevalue of the Output Gap signals demand pressure and excess factor capacity,respectively. A positive Output Gap may be a sign of demand pressures on inflation(over heating) in the economy while a negative Output Gap may show that excessfactor capacity remains unutilized or underutilized in the economy. An output levelwhere Non Accelerating Inflation Rate of Unemployment (NAIRU) is achieved ishighly desirable in an economy and this is important policy input for Monetary andFiscal policies.

…………

1.2 Motivation of Research

Managing inflation, stabilizing the exchange rate and fine-tuning interest ratesare core responsibilities of a Central Bank in any country. Central banks employMonetary and Exchange rate policies to tackle variations in inflationary tendencies,so that a steady and continuous economic growth path can be achieved over anextended period of time. Monetary policy therefore is a set of discretionary and non-discretionary tools used by the Central bank to eventually ensure high economicgrowth. Business cycle fluctuations, their ‘causes’ and ‘remedial measures’ are a sourceof extensive debate in economics theory over much of the last century. The role ofprudent and proactive Monetary and Fiscal policies has been much stressed ineconomic thought (especially in Keynesian economics). The role of reliable andaccurate macroeconomic variables is paramount in formulation and assessment ofafore mentioned policies. In light of the above, estimate of the potential output of aneconomy serves as yardstick for evaluating the relative standing within the businesscycle.

………..

Chapter 2: Analysis and Estimation of Output Gap usingUni-Variate Filter Methods

2.1. Non-Structural Methods to Estimate the Output Gap

By Non-structural methods we intend to state those methods which moresoundly based in statistical methods than in economic theory, these methods canarguably be termed as ‘model free’ methods of estimation. This is not to say that themethods described as Non-Structural (or ‘model free’) have no economic sense,rather these tools have been assessed over more than a decade of economic researchand have been reliably tested and tried. This definition holds within its self the one ofthe foremost criticisms of economic theorists. The reason why these methods arepopular among practicing economists and econometricians is the ease of use of thesemethods. As compared to other structured methods these methods require fairlysimple data input and results are well rounded in by statistical standards. : is the prescribed penalty on the volatility of trend component, as the aboveproblem is to be minimized therefore, a higher value of this penalty would tend tosmooth the time series in a higher proportion than that of a lower value. Hence a timeseries with higher frequency would naturally justify a higher value of this penalty.The vast literature on HP filter prescribes a value of 100 for annual frequency, 1600for quarterly frequency and 14400 for monthly frequency time series data. Theavailable literature on HP-filter also points to other criterion to arrive at differentvalue of this penalty.12The first term in the HP filter algorithm intends to minimize the variation in thecyclical component of the time series (i.e. it is the deviation of the time series from itstrend) while the following term acts to reduce the variability in the trend componentof the time series.

…………

2.2. Band Pass filter (Baxter-King Filter Method)

Within the literature used to estimate Output Gap, the band-pass filterproposed by Baxter and King is widely used. A band pass filter is a spectral filterwhich penalizes variations in the time series by imposing an arbitrarily imposed upperand lower band, this way those observations which are above or below the limits arefiltered out resulting in a smoothed time series over time. Baxter and King’s (1995)band-pass filter is constructed as the difference between two optimal approximatedlow-pass filters with the constraint that the filter weights sum to zero. Baxter andKing's band-pass filter is designed to pass through components of time series withcycles between six and 32 quarters. Baxter and King recommend to use the valueK=3 for annual data and K =12 for quarterly data. A higher value of K, should give abetter approximation to an ideal filter (smaller leakage and compression) but limitsthe number of observations available to the researcher. Figure 3 shows Output Gapcalculated using Baxter-King filter method.

…………

CHAPTER 3: ESTIMATING THE OUTPUT GAP OF PAKISTAN ONANNUAL......... 20

3.1. Structural Methods for Estimation of Output Gap .........20

3.1.1. State-Space Model .........21

3.1.1.2. Discussion on Output Gap Estimates for Pakistan Economy .........23

3.2. Annual Assessment of Output Gap .........24

3.3. Quarterly Output Gap Assessment......... 27

3.4. Conclusion & Discussion on Results .........30

CHAPTER 4: ESTIMATING THE OUTPUT GAP OF PAKISTAN ONQUARTERLY.........39

4.1. Introduction......... 39

4.2. Literature Review......... 40

4.3. Methodology .........41

4.3.1. State Space Method .........42

4.3.2 Structural Vector Auto Regression (SVAR) Method .........44

4.3.3 Wavelet Filter Method .........49

4.4. Conclusion and Policy Implications......... 50

CHAPTER 5: ANALYZING AND FORECASTING OUTPUT GAP ANDINFLATION......... 54

5.1. Introduction .........54

5.2. Literature Review .........56

5.3. Methodology, Data and Estimation......... 58

5.4. Results and Discussion .........66

Chapter 7: Bayesian and Multivariate Leser Filter Methodfor Estimating the Output Gap of Pakistan

7.1 Introduction

Potential output is an estimate of maximum output attainable by the economyutilizing the existing stock of capital and labor without adding pressures oninflationary tendencies (Conway and Hunt, 1997). Estimates of output gap28canfunction as useful indicator for macroeconomic policy assessment andimplementation. Such estimates are also useful as fiscal and monetary policies needto be aligned with the information whether the economy is above or below itssustainable long run capacity. In case when the output gap is positive i.e. ActualOutput > Potential Output, implies that there is excess demand in the economy, and incase where output gap exhibits a negative value (i.e. Actual Output < Potential Output)it implies the existence of excess supply. In essence of Blanchard and Quah (1989)shocks to total output can be categorized as either (A): Short run fluctuations(transitory) or (B): Permanent (persistent) shocks, to the production capacity of theeconomy. In light of the above, it is imperative to know whether the change in outputis caused by demand shocks or supply shocks or whether the economy is capable ofmore supply.The recent decade has been unrelentingly dismal for the economy of Pakistan.High unemployment, persistent double digit inflation since 2004, high fiscaldependence on monetization of deficit, rising unemployment and slowdown in realGDP, along with political instability and adverse law and order conditions havecompounded the impact of economic slowdown in recent years. Only recently sinceQ1 - FY2014Q1 economy of Pakistan has depicted signs of recovery. In this scenarioit is important to estimate and analyze the short run trend of total output of Pakistan.

………….

Conclusion

In order to consolidate the output gap estimates from the 5 methods we posittwo simple methods; an average of estimates at each point in time (see Figure 1 and2), in the second method we state a weighted average of the estimates at each point intime. The weighted average series is distributed on the intuition that; typicallyquarterly time series data included a greater irregular time series component (i.e.noise) and ideally output gap is assumed to be the cyclical component of the timeseries, extending this intuition weights in the aggregated output gap series should bedistributed in the 5 methods as inversely proportional to the variance of the series, i.e. for output gap estimates from the 5 afore mentioned methods; the higher the variancethe lower the weight in the aggregated output gap series.

…………

Reference (omitted)