No full-text available
To read the full-text of this research,
you can request a copy directly from the author.
Forecasting Australia's economic performance during the Asian crisis
Abstract
During the Asian economic crisis of 1997–98, published forecasts from a Bayesian vector autoregressive (BVAR) model consistently indicated that the crisis would have little or no effect on Australia’s economic performance, despite the deterioration in the trade balance. The worsening trade deficit led many other forecasters to predict a sharp fall in Australia’s GDP growth rate, as the countries most severely affected by the crisis represent over 60 percent of Australia’s export markets. This paper argues that the more pessimistic forecasts attached too much weight to the links between Australia’s external accounts and GDP growth. In particular, I show that forecasts for the period September 1997 to December 1998, conditional on the actual path of the merchandise trade balance, predict higher inflation and interest rates than unconditional forecasts from a model without the trade balance. There does, however, appear to be useful information in the individual components of the trade deficit. Conditioning on the actual paths of both exports and imports generally produces more accurate forecasts than conditioning on net exports. In particular, conditioning on the trade balance results in the least accurate forecasts for inflation and interest rates of any of the models considered here. On the other hand, conditioning on the individual trade flows produces the most accurate forecasts for inflation, and the second-most accurate for interest rates. Taken together, the results presented here lend support to the argument that Australia’s trade flows represent the outcomes of optimizing decisions, rather than defining constraints on economic growth.
... Summers (1999) which outlines the Bayesian Vector Autoregression (BVAR) model used at the Melbourne Institute. Subsequent work (Summers 2001) showed the effectiveness of forecasts generated using the BVAR specification during the Asian financial crisis . A relatively recent study is Tsiaplias & Chua (2010) The structure of this paper is as follows, section 2 outlines the innovations state space model including the new multivariate specification. ...
... Summers (1999) which outlines the Bayesian Vector Autoregression (BVAR) model used at the Melbourne Institute. Subsequent work (Summers 2001) showed the effectiveness of forecasts generated using the BVAR specification during the Asian financial crisis. A relatively recent study is Tsiaplias & Chua (2010) which examines the use of large data sets using the same approach as Stock & Watson (2002a,b). ...
Innovations state space time series models that encapsulate the exponential smoothing methodology have been shown to be an accurate forecasting tool. These models for the first time are applied to Australian macroeconomic data. In addition new multivariate specifications are outlined and demonstrated to be accurate.
... Not until the mid-1990s were there enough data points to make this method more useful for economy-wide modelling in Australia. Since then a number of studies have been undertaken, including Moreno (1992), Weber (1994), Fisher (1996), Huh (1999, Brischetto and Voss (1999), Summers (1999Summers ( , 2001, Pagan (2000), Suzuki (2001), Willard, Voss and Zha (2001). 2 In this paper we examine the impact of international economic conditions on the Australian economy using results derived from the VAR literature. In particular we take an updated version of Dungey and Pagan (2000) to decompose the influences on the Australian economic cycle over the 1990s and examine the role of domestic monetary policy. ...
... The literature is somewhat divided here; earlier VAR models of Australia such as Weber (1994), Smith and Murphy (1994) and Moreno (1992) had full recursivity. More recently, Suzuki (2001) finds that the empirical evidence does not support the small open economy restriction and Summers (1999Summers ( , 2001 reports feedbacks. Brischetto and Voss (1999) and Willard, Voss and Zha (2001) both ...
In the presence of large international disturbances small open economies are faced with difficult policy choices. International conditions impact on domestic outcomes. Using a structural VAR model of the Australian economy I explore the ways in which domestic monetary policy contributes to the output outcomes experienced in the economy. The focus is on the impact of international shocks. Monetary policy is modelled using a cash rate response to GNE, inflation and real exchange rate shocks. The results show that removing the focus on either GNE or inflation leads to lower GDP outcomes for the economy. The challenge for domestic policy is to recognise and respond to international and domestic shocks to the maximum benefit of the domestic economy.
... The sum of coefficients prior is controlled by using the hyperparameter, µ h 5 ≥ 0. Sims and Zha (1998) assume that a good forecast of θ i at some point t = τ is the average of the lagged values of θ i , denoted as θ i , but θ i does not help to predict the values of θ j for j ̸ = i. They introduced n dummy variables, θ (Doan et al., 1984;Sims and Zha, 1998;Robertson and Tallman, 1999;Summers, 2001). ...
Parametric mortality models capture the cross section of mortality rates. These models fit the older ages better, because of the more complex cross section of mortality at younger and middle ages. Dynamic parametric mortality models fit a time series to the parameters, such as a Vector-auto-regression (VAR), in order to capture trends and uncertainty in mortality improvements. We consider the full age range using the Heligman and Pollard (1980) model, a cross-sectional mortality model with parameters that capture specific features of different age ranges. We make the Heligman-Pollard model dynamic using a Bayesian Vector Autoregressive (BVAR) model for the parameters and compare with more commonly used VAR models. We fit the models using Australian data, a country with similar mortality experience to many developed countries. We show how the Bayesian Vector Autoregressive (BVAR) models improve forecast accuracy compared to VAR models and quantify parameter risk which is shown to be significant.
... The sum of coefficients prior is controlled by using the hyperparameter, µ h 5 ≥ 0. Sims and Zha (1998) assume that a good forecast of θ i at some point t = τ is the average of the lagged values of θ i , denoted as θ i , but θ i does not help to predict the values of θ j for j = i. They introduced n dummy variables, θ Sims and Zha, 1998;Robertson and Tallman, 1999b;Summers, 2001). ...
http://handle.unsw.edu.au/1959.4/51582
Future mortality rates are uncertain and the risk that estimated mortality rates will be higher than observed rates has negative financial implications for providers of living benefits including life annuities and pensions. This thesis studies time trends and cohort trends in mortality rates to determine the number of factors that drive mortality changes. An econometric analysis of mortality improvements is used to give a clearer picture of the stochastic nature of mortality rates in a lower dimensional data space as this thesis uses cointegration analysis for dimension reduction. A multi-country analysis of standardized mortality rates finds evidence of stochastic trends and a significant number of common factors. However, no evidence of common stochastic trends is found. An analysis of Australian mortality rates establishes there are non-stationary and stationary mortality rates by age. The common stochastic trends across age-groups which are exhibited within the Australian data lead to the characterization of mortality rates using a stochastic trend model. Dimension reduction is performed using the Heligman and Pollard (1980) parametric mortality model. The trends in the data are reflected using flexible Vector Autoregressive (VAR) models allowing for correlation between the estimated Heligman and Pollard model parameters. Bayesian Vector Autoregressive (BVAR) models which additionally quantify parameter risk are shown to significantly improve the forecast accuracy when fitting the developed HP-BVAR model to data from 1946-1995 and then comparing its out-of-sample forecasts to observed data from 1996-2007 for Australian mortality rates. Allowing for parameter uncertainty shows it to be a significant component of total risk since the results are realistic probabilistic forecasts. The HP-BVAR model is applied to the calibration of the longevity stress margin of the life insurance capital charge. The structure and magnitude of the current simplification by APRA result in a longevity stress margin that is found to be too prudent and too generalised. An alternative age-dependent simplification is proposed.
... The model incorporates both foreign (U.S.) and domestic sectors. In Summers (2001) there is no block exogeneity assumption, although more recent versions of the model have adopted this. Willard, Voss and Zha (2001) have developed a BVAR model where the exogenous U.S. sector is based on the Atlanta Federal Reserve Bank model of Leeper and Zha (1999). ...
The objective of the structural VAR model developed in this paper is to identify the effects of Australian monetary policy on the economy, within a small open economy framework. The methodology applied is the Bayesian techniques of Sims and Zha (1998) and Waggoner and Zha (2000) to incorporate prior information into the VAR model. Results consistent with economic theory are achieved by employing a block exogeneity assumption combined with additional contemporaneous and lagged linear restrictions. The paper examines the effects of introducing a mark-up pricing model restriction in to the monetary policy environment. In summary, the effects of restricting the model with this assumption, in comparison to the base model, is to lengthen the negative output gap produced by a wage inflation shock. Conversely, a price inflation shock induces a relatively larger fall in real wages that subsequently leads to a more pronounced above trend phase in output. The effect of a cash rate increase in the mark-up restricted model is to have a very similar effect on reducing the level of GDP to the unrestricted case.
... The sum of coefficients prior is controlled by using the hyperparameter, µ h 5 ≥ 0. Sims and Zha (1998) assume that a good forecast of θ i at some point t = τ is the average of the lagged values of θ i , denoted as θ i , but θ i does not help to predict the values of θ j for j = i. They introduced n dummy variables, θDoan et al., 1984; Sims & Zha, 1998; Robertson & Tallman, 1999; Summers, 2001). The weighted dummy observations for the case n=3 and p = 2 is: ...
Mortality risk models have been developed to capture trends and common factors driving mortality improvement. Multiple factor models take many forms and are often developed and fitted to older ages. In order to capture trends from young ages it is necessary to take into account the richer age structure of mortality improvement from young ages to middle and then into older ages. The Heligman and Pollard (1980) model is a parametric model which captures the main features of period mortality tables and has parameters that are interpreted according to age range and effect on rates. Although time series techniques have been applied to model parameters in various parametric mortality models, there has been limited analysis of parameter risk using Bayesian techniques. This paper uses a Bayesian Vector Autoregressive (BVAR) model for the parameters of the Heligman-Pollard model and fits the model to Australian data. As VARmodels allow for dependence between the parameters of the Heligman-Pollard model they are flexible and better reflect trends in the data, giving better forecasts of the parameters. Forecasts can readily incorporate parameter uncertainty using the models. Bayesian Vector Autoregressive (BVAR) models are shown to significantly improve the forecast accuracy of VAR models for mortality rates based on Australian data. The Bayesian model allows for parameter uncertainty, shown to be a significant component of total risk.
... Both Tavlas (2000) and Grenville (1999) support the contention that Australia was not seriously effected by the Asian crisis attributing this to Australia's relatively more mature financial markets. McKibbin (1998McKibbin ( , 1999 comes to similar conclusions, while Summers (2001) presented predictions that the effects of the Asian crisis on the Australian economy would not be as large as some feared. Duncan and Yang (2000) attributes Australia's resilience to the Asian crisis to the low dependence on manufacturing exports in Australia, as there is much pressure on this sector from those economies involved in the Asian crisis. ...
During the East Asian currency crisis of 1997-98 the potential transmission of the crisis to developed markets such as Japan, Australia and New Zealand, was of considerable policy concern. Potential channels consist of anticipated movements stemming from common factors, spillovers and contagion. The empirical results show that the transmission of volatility in the East-Asian currency markets to the developed markets in the region is not due to contagion, but rather attributed to common world factors. Spillovers have a minor role in the case of Japan and to a lesser degree, Australia. Copyright Blackwell Publishing Ltd/University of Adelaide and Flinders University of South Australia 2004.
Vector autoregressions (VARs) are richly parameterized time series models that can capture complex dynamic interrelationships among macroeconomic variables. However, in small samples the rich parametrization of VAR models may come at the cost of overfitting the data, possibly leading to imprecise inference for key quantities of interest such as impulse response functions (IRFs). Bayesian VARs (BVARs) can use prior information to shrink the model parameters, potentially avoiding such overfitting. We provide a simulation study to compare, in terms of the frequentist properties of the estimates of the IRFs, useful strategies to select the informativeness of the prior. The study reveals that prior information may help to obtain more precise estimates of impulse response functions than classical OLS-estimated VARs and more accurate coverage rates of error bands in small samples. Strategies based on selecting the prior hyperparameters of the BVAR building on empirical or hierarchical modeling perform particularly well.
Reflecting the importance of commodities for the Australian economy, we construct a dynamic stochastic general equilibrium (DSGE) model of the Australian economy with a commodity sector. We assess whether its forecasts can be improved by using it as a prior for an empirical Bayesian vector autoregression (BVAR). We find that the forecasts from the BVAR tend to be more accurate than those from the DSGE model. Nevertheless, for output growth these forecasts do not outperform benchmark models, such as a small open economy BVAR estimated using the standard priors for forecasting. A Bayesian factor augmented vector autoregression produces the most accurate near-term inflation forecasts.
Most companies find that obtaining information about the future is particular to ensuring future organisational success. The analysis of macro and micro environmental factors and conditions is one of the main procedures in strategy formulation for a business company. Forecasting is a decision tool that has a variety of managerial application areas, one of which is the strategic decision-making. The study mainly discusses the potential uses of forecasting within the strategic, particularly marketing, management, and strategic decision-making. It points out both the internal and external strategic decision areas in companies and highlights the potential application areas of forecasting within these areas. The paper also highlights how functional forecasting is as a managerial tool that provides the information needed in critical strategic decision situations.
This paper estimates a small open economy New-Keynesian model using data from Australia and the US economy with Full Information Maximum Likelihood method. Our estimated US structural parameters are in line with those of Giordani (2004) on Canadian data. For Australian parameters we find that the real exchange rate has little effect on output gap and inflation, whereas the foreign output gap has a significant effect on both of them. Moreover, our results show a prominent role for backwardlooking behaviour for Australian inflation.
It is no longer necessary to argue for the importance of white-collar crime. Its devastating financial and physical effects
are obvious. The task now is to develop better ways to control and prevent white-collar crimes. In this paper, we argue that
in order to reduce white-collar crime we must first identify the specific opportunity structures associated with the offenses
we wish to prevent. That is, we must identify the features of the settings that allow the crime to occur. Three major theories
are reviewed here that can help in this task: routine activity theory, crime pattern theory, and situational crime prevention
theory. These theories address how crime opportunities are formed by immediate environments and then discovered and evaluated
by potential offenders. We demonstrate that they can also be used to uncover how specific forms of white-collar crimes are
committed and help structure analyses of the underlying opportunity structures associated with these offenses.
In this chapter, we draw direct comparisons between corporate crime and terrorism in order to improve our understanding on
how to better control each. We acknowledge obvious differences between the corporate criminal and the terrorist organization,
but also raise important similarities between them. Namely, corporations and terrorist organizations both strive to survive
in highly competitive environments and have adopted more complex organizational structures over time to achieve this goal.
However, it is this organizational complexity thatmakes it difficult for criminal justice officials to detect and prosecute
illicit behavior. After drawing upon the literature in both areas to describe organizational complexity, we offer several
(surprisingly similar) recommendations to better control the corporate criminal and the terrorist organization.
Persistent heterogeneity and state dependence explain the persistence of criminal offending over time in two distinct ways.
The first asserts that offending is persistent over time due to a stable individual trait while the latter suggests that offending
has a causal relationship with future crime because it increases the risk of negative consequences. In this chapter, we examine
Occupational Safety and Health Act (OSHA) violations to assess which of these explanations is more consistent with corporate
crime over time. Using fixed-effects time series analysis (which controls for unobserved heterogeneity), we find an inhibitory
effect associated with OSHA inspections, thus challenging the persistent heterogeneity explanation.
This paper applies general strain theory (GST) to the explanation of white-collar crime, including (a) occupational crimes
committed by higher class individuals, (b) economic offenses such as fraud and embezzlement, which are committed by lower
as well as higher class individuals, and (c) corporate crimes. Several strains or stressors are said to be especially relevant
to the explanation of such crimes, including the blockage of economic goals, the experience of a range of other economic problems,
the inability of achieve status goals, and a variety of work-related stressors. Whether these strains result in white-collar
crime, however, is said to be influenced by such things as coping skills and resources, social support, opportunities for
white-collar crime, social control, the perceived costs and benefits of white-collar crime, and association with criminal
others.
The criminal career paradigm represented a successful shift in criminological thinking, and ensuing research has generated
important descriptive information about the key dimensions of active criminals; spurning both theoretical (developmental/life-course
criminology) and methodological/statistical advances. Yet, the paradigm has failed to take into account acts of criminality
that do not fit into the stereotypical image of street offending, in particular white-collar crime. The current study utilizes
group-based trajectory modeling to examine trends of criminal behavior in a sample of convicted white-collar criminals over
a more than 10-year follow-up period. Three offender trajectories are identified (low rate, intermittent, and persistent offenders)
and suggest the importance of recognizing the variability of offending in a white-collar crime sample, and the overlap between
white-collar and common crime criminal careers. This study also suggests the importance of recognizing both static and dynamic
factors in the understanding of criminal careers. This research confirms a heterogeneous view of white-collar crime which
recognizes that the white-collar crime category includes within it a broad diversity of offenders, and suggests that it is
important to recognize that different models of explanation may be needed to provide explanations for different types of offenders.
The focus of this chapter is ‘fear of crime’ and ‐ within severe data limitations and conceptual controversies ‐ it interrogates the variations in fear of different forms of crime, white-collar and other. It begins by examining the state of ‘fear of crime’ policy and what we might mean by ‘fear’ in this sort of arena; goes on to review fears of and concerns about those white-collar crimes that have been researched; and ends with a discussion of their implications for thinking and policy about fear of crime generally and about what we can learn about it from studying fears about white-collar crimes.
This chapter discusses how globalization facilitates white-collar crime, and how it can foster more effective mechanisms for
the prevention and control of white-collar crime. It uses the framework of routine activity theory to explain transnational
white-collar crime, then presents a range of offence types with examples of cross-border offending. The essay then turns to
a discussion of regulatory institutions that can comprise a system of transnational white-collar crime control. It illustrates
the evolution of global regulatory systems with the examples of transnational corruption control and the international cooperation
to combat money laundering. The chapter concludes with the observation that, as is the case of conventional domestic “street
crime,” the effective prevention and control of whitecollar crime in an era of globalization requires the involvement of public,
private, and nonprofit institutions.
Securing employee adherence to workplace rules and company policies is a key antecedent of successful functioning within organizations.
It is important for companies to be able to effectively motivate rule following behavior among employees, i.e. to be able
to limit white-collar crime. This chapter compares the utility of two approaches to securing such behavior. Those strategies
are as follows: (1) the sanction-based command-and-control model and (2) self-regulation approaches that are linked to activating
employee’s ethical judgments. Research findings suggest that, while command-and-control strategies influence employee behavior,
self-regulatory strategies have a stronger influence. Studies also explore the basis of these ethical judgments, and find
that the primary factor shaping them is the procedural justice that employees experience in their workplace. These results
suggest that the roots of employee policy adherence and rule following behavior lie in the procedural justice of the organization.
Overall, this analysis highlights the important role ethical judgments play in motivating both rule following and policy adherence
among employees in work settings and provides practical suggestions for shaping those judgments.
In 2003, China's energy consumption amounted to 1678 million tonnes coal equivalent (MtCE), making China the world's second largest consumer behind only the United States. China is now also one of the largest oil importers in the world. With an economy that is expected to maintain a rate of growth of 7–8% for decades, China's role in the world energy market becomes increasingly influential. This makes it important to predict China's future demand for energy. The objective of this paper is to apply the Bayesian vector autoregressive methodology to forecast China's energy consumption and to discuss potential implications. The results of this paper suggest that total energy consumption should increase to 2173 MtCE in 2010, an annual growth rate of 3.8%, which is slightly slower than the average rate in the past decade. The slower growth reflects expected slower economic growth and a decline in energy consumption due to structural changes in the Chinese economy.
This paper examines whether the disaggregation of consumer sentiment data into its sub-components improves the real-time capacity to forecast GDP and consumption. A Bayesian error correction approach augmented with the consumer sentiment index and permutations of the consumer sentiment sub-indices is used to evaluate forecasting power. The forecasts are benchmarked against both composite forecasts and forecasts from standard error correction models. Using Australian data, we find that consumer sentiment data increase the accuracy of GDP and consumption forecasts, with certain components of consumer sentiment consistently providing better forecasts than aggregate consumer sentiment data. Copyright © 2009 John Wiley & Sons, Ltd.
This paper examines the empirical literature on financial market contagion in Asia during the 1997-98 financial crises with respect to existing tests of contagion. Empirical evidence shows that contagion affects both developed and emerging markets and does not seem to vary with the relative fundamental economic health or trade and financial linkages of the Asian economies. Contagion occurs across both asset types and geographical borders and tends to have larger effects in equity markets than in currency and bond markets. There is evidence to support the hypothesis that contagion is regional and transmitted through developed markets. A discussion of the behavior of correlation coefficients in the presence of contagion and financial crises suggests that they are not a reliable metric for detecting contagion. (c) 2006 The Earth Institute at Columbia University and the Massachusetts Institute of Technology.
A structural vector autoregressive model of the Australian economy that allows for international shocks from the USA, Japan as well as world commodity prices is specified and estimated for the period 1979-1999. A block exogenous structure linking the three countries is imposed. The international linkages are modelled using a factor structure to circumvent problems from estimating large scale dynamic models. The factors are estimated recursively using a Kalman filter and are found to represent aggregate demand and liquidity shocks for the USA and Japan respectively. The key empirical result is the USA shocks are the dominant source of international shocks on the Australian economy with the Japanese economy having a dampening effect on the USA shocks. The empirical results also show that Australian monetary policy responds to domestic conditions rather than international monetary policy.
This paper evaluates different policy options to reduce unemployment by using a version of the TRYM model. For the purpose of this paper, the TRYM model has been modified in several respects, particularly by combining the private business and government trading enterprise sectors. For the long run, the neoclassical model closure means that the unemployment rate converges to an exogenously set NAIRU rate. For the short and medium run, periods well in excess of ten years, policy simulations show that macroeconomic policy changes, wage changes, labour productivity changes, and NAIRU changes affect employment and unemployment. Further, these policy effects are produced whether the model begins in a disequilibrium situation of unemployment above the NAIRU or at the long run equilibrium growth path with unemployment equal to the NAIRU.
A modified Treasury Macroeconomic model is used to assess the relative effects of policy options to reduce unemployment. Simulation results are reported for the Australian economy starting at either a high or a low rate of unemployment. Over the next 10 years or so, all policies are projected to generate cyclical gains and losses in macroeconomic outcomes, including unemployment, relative to the base-case scenario, before converging to an exogenous long-run equilibrium growth path. The structure, amplitudes and other details of the cyclical responses vary both with the policy option and with whether the starting economy unemployment rate is close to the non-accelerating-inflation rate of unemployment or at a much higher rate. Copyright 2005 The Economic Society Of Australia.
Although Australia has an equivalently large trading relationship with Japan and the US, current macro models often incorporate only US variables in the external sector of Australia. This paper explores the consequences of including both US and Japanese effects in the international sector of a SVAR model of Australia. The results indicate the significance of the Japanese effects. Excluding Japan results in an overstatement of the impact of US based shocks on the Australian economy. When Japan is included, US based shocks remain dominant in explaining Australian outcomes, but the responses are moderated compared with a model incorporating only a US based external sector. This has important implications for domestic policy responses to international shocks. Without the influence of Japan, domestic monetary policy will over-react to a US based shock.
This paper examines the effects of monetary policy in Australia using a small structural vector autoregression model. The model we use is a modification of the small open economy model developed for the G6 economies (the G7 less the United States) by Kim and Roubini (1999). The success of the Kim and Roubini model across these economies makes it a natural starting point to analyse monetary policy in Australia. Consistent with Kim and Roubini, we find no evidence of the price or exchange rate puzzles identified in the literature. We demonstrate that, in the Australian context, modelling the interdependence between the domestic interest rate, the foreign interest rate and the nominal exchange rate is critical to resolving these puzzles. Further, we demonstrate that the Kim and Roubini model can be further usefully refined when applied to the Australian data. We use the model to perform various monetary policy experiments and the results of these experiments accord well with both the existing SVAR literature and previous empirical work for Australia. We find that monetary policy shocks have a delayed and gradual effect on the price level and a small temporary effect on output. We also use the model to examine the effects of shocks to the Australian economy and the role of monetary policy in response to these shocks. Generally, we find that monetary policy has served to dampen both output and price fluctuations. A qualifying aspect of our results is the model’s instability, for example, around the most recent change in the implementation of monetary policy in Australia.
This paper develops a forecasting procedure based on a Bayesian method for estimating vector autoregressions. The procedure is applied t o 10 macroeconomic variables and is shown to improve out-of-sample forecasts relative to univariate equations. Although cross-variable responses are damped by the prior, considerable interaction among the variables is shown to be captured by the estimates We provide unconditional forecasts as of 1982:12 and 1983:3. We also describe how a model such as this can be used to make conditional projections and to analyze policy alternatives. As an example, we analyze a Congressional Budget Office forecast made in 1982: 12 Although no automatic causal interpretations arise from models like ours, they provide a detailed characterization of the dynamic statistical interdependence of a set of economic variables, information that may help in evaluating causal hypotheses without containing any such hypotheses.
If dynamic multivariate models are to be used to guide decisionmaking, it is important that probability assessments of forecasts or policy projections be provided. When identified Bayesian vector autoregression (VAR) models are presented with error bands in the existing literature, both conceptual and numerical problems have not been dealt with in an internally consistent way. In this paper, the authors develop methods to introduce prior information in both reduced-form and structural VAR models without introducing substantial new computational burdens. Their approach makes it feasible to use a single, large dynamic framework (for example, twenty-variable models) for tasks of policy projections. Copyright 1998 by Economics Department of the University of Pennsylvania and the Osaka University Institute of Social and Economic Research Association.
Vector autoregression (VAR) models are widely used for policy analysis. Some authors caution, however, that the forecast errors of the federal funds rate from such a VAR are large compared to those from the federal funds futures market. From these findings, it is argued that the inaccurate federal funds rate forecasts from VARs limit their usefulness as a tool for guiding policy decisions. In this paper, we demonstrate that the poor forecast performance is largely eliminated if a Bayesian estimation technique is used instead of OLS. In particular, using two different data sets we show that the forecasts from the Bayesian VAR dominate the forecasts from OLS VAR models—even after imposing various exact exclusion restrictions on lags and levels of the data.
Constructing forecasts of the future path for economic series such as real gross domestic product growth, inflation, and unemployment forms a large part of applied economic analysis for business and government. Model-based forecasts are easier to replicate and validate by independent researchers than forecasts based on expert opinion alone. In addition, the forecaster can formally investigate the source of systematic errors in model forecasts, and a forecast model s performance can be established before it is used by a decision maker. ; The authors of this article describe a particular model-based forecasting approach, a vector autoregression comprising six U.S. macroeconomic variables. They focus attention on the technical hurdles that must be addressed in a real-time application and methods for overcoming those hurdles, such as conditional forecasting to handle the staggered release of data and matching quarterly with monthly data. ; By emphasizing the practical problems of forecasting economic data using a statistical model, the authors draw on experience in using such a model at the Federal Reserve Bank of Atlanta. Although the model studied is small and highly aggregated, it provides a convenient framework for illustrating several practical forecasting issues. The focus on a simple model provides potential users with a road map of how one might implement such a forecasting model in specific applications.
For the first-order univariate autoregression without constant term, the joint density (corresponding to a flat prior) for the true coefficient and its least squares estimate is estimated by Monte Carlo and graphically displayed. The graphs show how a symmetric distribution for the coefficient conditional on the estimate coexists with an asymmetric distribution for the estimate conditional on the coefficient. Prior densities implicit in treating classical significance levels as if they were Bayesian posterior probabilities are calculated. They are shown to depend sensitively on the estimated coefficient and to put substantial probability on values of the coefficient above one. Copyright 1991 by The Econometric Society.
draft of this paper is available by ftp from
: Models used for policy analysis should generate reliable unconditional forecasts as well as policy simulations (conditional forecasts) that are based on a structural model of the economy. Vector autoregression (VAR) models have been criticized for having inaccurate forecasts as well as being difficult to interpret in the context of an underlying economic model. In this paper, we examine how the treatment of prior uncertainty about parameter values can affect forecasting accuracy and the interpretation of identified structural VAR models. Typically, VAR models are specified with long lag orders and a diffuse prior about the unrestricted coefficients. We find evidence that alternatives that emphasize nonstationary aspects of the data as well as parsimony in parameterization have better out-of-sample forecast performance and smoother and more persistent responses to a given exogenous monetary policy change than do unrestricted VARs. JEL classification: E44, C53 Key words: Bayesian infer...
Nowadays the Central Bank of a country is the monopoly supplier of legal tender currency. The commercial banks are committed to making their deposits convertible at par into such currency. So the banks need to keep reserves in the form of currency and deposits at the Central Bank. The Central Bank primarily conducts its policy by buying and selling financial securities, e.g. Treasury bills or foreign exchange, in exchange for its own liabilities, i.e. open market operations. Academic economists generally regard such operations as adjusting the quantitative volume of the banks’ reserve base, and hence of the money stock, with rates (prices) in such markets simultaneously determined by the interplay of demand and supply. Central Bank practitioners, almost always, view themselves as unable to deny the banks the reserve base that the banking system requires, and see themselves as setting the level of interest rates, at which such reserve requirements are met, with the quantity of money then simultaneously determined by the portfolio preferences of private sector banks and non-banks. This difference in perceptions is discussed again in section 6.4.
Macroeconomic shocks such as wil price increases induce a systematic (endogenous) response of monetary policy. We develop a VAR-based technique for decomposing the total economic effects of a given exogenous shock into the portion attributable directly to the shock and the part arising from the policy response to the shock.
In applications of structural VAR modeling, finite-sample properties may be difficult to obtain when certain identifying restrictions are imposed on lagged relationships. As a result, even though imposing some lagged restrictions makes economic sense, lagged relationships are often left unrestricted to make statistical inference more convenient. This paper develops block Monte Carlo methods to obtain both maximum likelihood estimates and exact Bayesian inference when certain types of restrictions are imposed on the lag structure. These methods are applied to two examples to illustrate the importance of imposing restrictions on lagged relationships.
Previous empirical study on the effects of monetary policy shocks in small open economies has generated puzzling dynamic responses in various macroeconomic variables. This paper argues that these puzzles derive from an identification of monetary policy that is inappropriate for such economies. To remedy this, it is proposed that a structural model be estimated to explicitly account for the features of the small open economy. Such a model is applied to Canada with tightly estimated results overall. The dynamic responses to the identified monetary policy shock are consistent with standard theory and highlight the exchange rate as a transmission mechanism.
In the existing literature, conditional forecasts in the vector autoregressive (VAR) framework have not been commonly presented with probability distributions. This paper develops Bayesian methods for computing the exact finite-sample distribution of conditional forecasts. It broadens the class of conditional forecasts to which the methods can be applied. The methods work for both structural and reduced-form VAR models and, in contrast to common practices, account for parameter uncertainty in finite samples. Empirical examples under both a flat prior and a reference prior are provided to show the use of these methods. © 2000 by the President and Fellows of Harvard College and the Massachusetts Institute of Technolog
The issue of uncovering the effects of monetary policy is far short of resolution. In the identified VAR literature, restrictions have been imposed to identify the effects of unpredictable monetary policy disturbances. We offer critical views on the unreasonable assumptions in the existing work and argue for careful economic argument about identifying assumptions. We display a structural stochastic equilibrium model in which our VAR identification would produce correct results while drawing attention to the serious lack of time series fit in most of the DSGE literature.
This article puts forward the view that there are considerable benefits to be gained from foreign borrowing and that policy actions of a macro nature should be diverted towards internal balance.
In Section 2 a brief history of the evolution of approaches to the current account is set out. Recent theories imply that the current account indicates the optimal amount of foreign borrowing both by firms and households, at least with zero fiscal deficits. The present conventional Australian approach to current account issues (Section 3) would seem to be based on theories which were relevant in an era of pegged exchange rates.
In Section 4 it is suggested not only that the current account should not be a target of macroeconomic policy, but that any microeconomic policy directed at reducing it should be justified by the existence of externalities in the borrowing/debt process.
We show how to correctly extend known methods for generating error bands in reduced form VAR's to overidentified models. We argue that the conventional pointwise bands common in the literature should be supplemented with measures of shape uncertainty, and we show how to generate such measures. We focus on bands that characterize the shape of the likelihood. Such bands are not classical confidence regions. We explain that classical confidence regions mix information about parameter location with information about model fit, and hence can be misleading as summaries of the implications of the data for the location of parameters.
This paper compares the evolution of the Australian current account balance over the period 1954-94 against an optimal current account derived from a consumption-smoothing model. The findings indicate that the Australian current account was not used to smooth consumption optimally in the period prior to the relaxation of capital controls in the early 1980s. The results also suggest that, in the period since the mid-1980s, Australia's current account deficits have become excessive and that the increase in net national saving required to satisfy its external borrowing constraint is about 2 to 4 percent of GDP. Copyright 1998 by The Economic Society of Australia.
We present a data oriented analysis of the effect of different kind of economic shocks on Chilean output growth and inflation over the last 40 years. Two important results are: (1) foreign shocks only explain 17% of the variability of the output growth in the period 1984-2006 whereas it used to account for the 47,2% of output variability in 1966-1983; (2) The participation of foreign shocks to explain the Chilean inflation reaction becomes more importan in the last twienty years because of the price liberalization and Chile's openness to international trade. Results highlight specific features of the Chilean economy not present in other countries.
Does monetary policy A structural vector generate recessions? Working Paper 98-12, Federal autoregression model of monetary policy in Australia
- C A Sims
- T Zha
- A Brischetto
- G Voss
Sims, C.A., & Zha, T. (1998). Does monetary policy Brischetto, A., & Voss, G. (1999). A structural vector generate recessions? Working Paper 98-12, Federal autoregression model of monetary policy in Australia. Reserve Bank of Atlanta.
Improving Business Cycle Analysis: Frontier Techniques'. forecasts of the Federal Funds rate in a policy model
- Melbourne
- Forecasting
- J C Robertson
- E W Tallman
Melbourne conference titled 'Forecasting and
Robertson, J.C., & Tallman, E.W. (1999). Improving
Business Cycle Analysis: Frontier Techniques'.
forecasts of the Federal Funds rate in a policy model.
Dummy observation priors for VARs. computer code. I retain sole responsibility for Unpublished lecture notes, Princeton University, http: / any remaining errors. This work was sponsored / eco-072399b.princeton
- Sharing Robertson For
- Both
- Expertise
- C A His Sims
Robertson for sharing both his expertise and his
Sims, C.A. (1999). Dummy observation priors for VARs.
computer code. I retain sole responsibility for
Unpublished lecture notes, Princeton University, http: /
any remaining errors. This work was sponsored
/ eco-072399b.princeton.edu / yftp / TimesF99 /
in part by Australian Research Council grant
DummyObs2.pdf
C79930704.
Prior parameter Lahiri and two anonymous referees for helpful uncertainty: some implications for forecasting and policy analysis with VAR models. Mimeo, Federal Reserve comments. I am especially grateful to
- Ron Bewley
- Don Harding
- Olan Henry
- J C Robertson
- E W Tallman
Ron Bewley, Don Harding, Olan Henry, Kajal
Robertson, J.C., & Tallman, E.W. (1999). Prior parameter
Lahiri and two anonymous referees for helpful
uncertainty: some implications for forecasting and policy analysis with VAR models. Mimeo, Federal Reserve
comments. I am especially grateful to John
Bank of Atlanta.
Conditional forecasts ics, Empirical Economics, the Economic Record, and the in dynamic multivariate models. Review of Economics Australian Economic Review, among others
- D F Waggoner
- T Zha
Waggoner, D. F., & Zha, T. (1999). Conditional forecasts
ics, Empirical Economics, the Economic Record, and the
in dynamic multivariate models. Review of Economics
Australian Economic Review, among others. Peter reand Statistics 81, 639–651.
ceived his Ph.D. from the University of Iowa.
Vector autoreg-Conference of Economists, and a University of ressions: forecasting and reality. Economic Review. Federal Reserve Bank of Atlanta
- J C Robertson
- E W Tallman
Robertson, J.C., & Tallman, E.W. (1999). Vector autoreg-Conference of Economists, and a University of
ressions: forecasting and reality. Economic Review.
Federal Reserve Bank of Atlanta, 4-18.
Dummy observation priors for VARs
- C A Sims
Sims, C.A. (1999). Dummy observation priors for VARs.
computer code. I retain sole responsibility for
Unpublished lecture notes, Princeton University, http: /
any remaining errors. This work was sponsored
/ eco-072399b.princeton.edu / yftp / TimesF99 /
Vector autoregressions: forecasting and reality. Economic Review. Federal Reserve Bank of Atlanta
- J C Robertson
- E W Tallman