Research Publications

1. Macroeconomic Forecasting for Pakistan in a data-rich environment (with Dr. Kevin Lee, published in Applied Economics) - 2020

Abstract: This article forecasts the CPI inflation, GDP growth and the weighted average overnight repurchase rate in Pakistan using 161 predictors covering a period from July 2007 to July 2017. We use the naïve mean model and the autoregressive model as benchmark models and compare their forecasting performance against the dynamic factor model (DFM) and sophisticated machine learning methods such as the Ridge regression, the LASSO, the Elastic net and a few variants of Bagging. The main purpose of the article is to determine, how well the commonly used DFM which has been used for time series forecasting for a long time, performs against the recently developed penalized regression methods in forecasting key macro-economic variables in Pakistan. We forecast the variables of interest over 12 months forecast horizon. The forecast evaluation criteria used to compare the forecast performance of these models is the RMSE and MASE. For each variable of interest, we find that, for majority of the cases considered, one of the competing approaches outperform the benchmark models and other competing approaches at majority of forecast horizons. Our results show that, on the balance, the machine learning approaches perform better than the benchmark, the autoregressive and the DFM. 

2. The impact of COVID-19 on stock market and exchange rate uncertainty in Pakistan (with Dr. Kaneez Fatima and Mr. Mannan Zaheer, published in Business Review) 

Abstract: This paper examines the impact of coronavirus (COVID-19) on exchange rate (EXU) and stock market uncertainty (SMU) in Pakistan while controlling for the effects of interest rate and policy interventions by the Government and the Central bank to combat the pandemic. We employ the vector autoregressive (VAR) model over a sample period ranging from February 25, 2020, to May 6, 2021. We find that a shock to total daily coronavirus cases in Pakistan has a positive and significant impact on both the EXU and SMU. However, this impact is short-lived which may be attributed to a timely policy response and risk-averse nature of investors in Pakistan. This result is aligned with a vast literature on pandemics and investors uncertainty and remains robust to several robustness check exercises. 

3. The Impact of Economic Policy Uncertainty on Consumer Confidence in Pakistan (with Dr. Kaneez Fatima and Ms. Riffat Arshad, published in The Lahore Journal of Economics) 

Abstract: This study examines the impact of Economic Policy Uncertainty (EPU) on the consumer confidence index (CCI) in Pakistan. Using a sample from the start of 2012 up to February 2020, a vector errorcorrection model is used to gauge the impact of EPU on CCI. Our results show that a shock to EPU in Pakistan affects CCI negatively and significantly. The shock persists for a span of more than 20 forecast 1 Publishing research work focused on Pakistan is one of my job requirements; therefore, all my current publication is focused on Pakistan. However, with an international job, most likely research or teaching position at a U.S. university, my focus will be to conduct research on international issues, which is not the case now given the time constraint and current assignments. Finally, although, I am working mostly on Pakistan, my future work contains project on an important international dimension. 2 horizons, informing economic policy makers in Pakistan that sudden changes in the stance without proper communication can deteriorate consumer confidence. This is important as consumer confidence in Pakistan accounts for not only the current economic situation, but expected changes in key macroeconomic variables which is usually a key consideration when forward-looking policies are devised. Our results remain robust to alternate Choleski specifications and lag lengths in the model. 

4. The Macroeconomic Impacts of Entitlements (with Dr. Kaneez Fatima and Ms. Riffat Arshad, published in The Hilltop Review) 

Abstract: The worries expressed by Alan Greenspan that the long run economic growth of the United States will fade away due to increasing burden of entitlements motivated us to empirically investigate the impact of entitlements of key macroeconomic variables. A VECM is used to analyze the impact of entitlements on the price level, real output, and the long-term interest rate. The results show that a shock to entitlements leads to increase in output; however, this impact declines and lends support to the assertion made by Alan Greenspan. Several robustness checks are conducted and the results of the model qualitatively remains unchanged. 

5. The macroeconomic impacts of government debt in Pakistan (with Dr. Kaneez Fatima and Mr. Junaid Kamal, published in Business Review) 

Abstract: This paper examines the Ricardian equivalence hypothesis in Pakistan using a vector error correction model. The sample period extends from June 2002 to January 2020. The results are reported using variance decompositions and impulse response functions. The base model contains six variables and is estimated with 4 lags. We find support for the idea that wealth does not increase as government debt increases; Hence, it proves the fact that economic agents are rational actors and foresee current expansionary actions of the government that result in accumulation of debt as the present value of future taxation that they have to pay and not as an increase in wealth. The results remain robust to a change in sample, in Choleski ordering of the variables and in the lag length of the estimated models. 

6. Forecasting Group-Wise Imports and Exports of Pakistan (with Ms. Humaira Kiran and Ms. Sumbal Qureshi, published in The Pakistan of Applied Economics) 

Abstract: This study forecasts imports and exports of Pakistan at a disaggregated level. Both in and outof-sample forecasts are produced using conventional econometric time-series models and the Artificial Neural Network (ANN), a machine learning approach. The forecast performance is reported using the Root Mean Squared Error (RMSE). Given improved forecasts by the ‘iterative optimization’ nature of the long and short-term method, ANN outperforms other model in-sample. For the out-of-sample period, the autoregressive (AR) and ANN model outperforms other models for import groups, while all univariate approaches outperform each other for two out of six subgroups in out-of-sample forecasts. Hence, performing equivalently well for export groups. 

7. Forecasting the GDP Growth in Pakistan: The Role of Consumer Confidence (with Dr. Kaneez Fatima and Ms. Riffat Arshad, published in The Lahore Journal of Economics)

Abstract: This paper investigates whether consumer confidence improves the prediction of GDP growth over what are popularly construed as fundamental economic variables. We use monthly data concerning Consumer Confidence Index (CCI) and its sub-indices to forecast GDP growth for Pakistan. Employing a set of univariate and multivariate models and comparing their forecasting performance against the Naïve mean model, we find that adding the consumer sentiments with fundamental economic variables improves the forecast of GDP growth. Vector autoregressive model with current economic conditions index and economic fundamentals, we find, performs the best. The results have potential policy implications in terms of tackling unemployment and inflation, for economic growth stimulation.  

8. Inflation Forecasting for Pakistan in a Data-rich Environment (with Mr. Muhammad Ishtiaq, Ms. Sumbal Qureshi, and Dr. Kaneez Fatima, published in The Pakistan Development Review) Abstract: This paper uses machine learning methods to forecast the year-on-year CPI inflation of Pakistan and compare their forecasting performance against the comprehensive traditional forecasting suite contained in Hanif and Malik (2015). It also augments the comprehensive forecasting suite with the dynamic factor model which is able to handle a large amount of information and put all of these models in competition against the latest machine learning models. A set of 117 predictors covering a period of July 1995 to June 2020 is used for this purpose. We set the naïve mean model as the benchmark and compare its forecasting performance against 14 traditional and 5 sophisticated machine learning models. We forecast the year-on-year CPI inflation over a 24 months horizon. Forecasting performance is measured using the RMSE. Our results show that the machine learning approaches perform better than the traditional econometric models at 18 forecast horizons.