Bank competition, financial reform, and institutions: The importance of being developed

In this paper, I estimate the degree of market power at the bank-level for 84 banking systems worldwide. Subsequently, I analyze the sources of bank competition, placing emphasis on the impact of financial reform and the quality of institutions. I find that financial liberalization policies reduce the market power of banks in developed countries with advanced institutions. In contrast, banking competition does not improve at the same pace in countries with weaker institutions and a lower level of institutional development. The results hold across a wide array of identification tests and estimation methods. The main policy implication to be drawn is that a certain level of institutional development is a precondition for the success of reforms aimed at enhancing the competition and efficiency of banking markets.     

Overnight borrowing, interest rates and extreme value theory

We examine the dynamics of extreme values of overnight borrowing rates in an inter-bank money market before a financial crisis during which overnight borrowing rates rocketed up to (simple annual) 4000 percent. It is shown that the generalized Pareto distribution fits well to the extreme values of the interest rate distribution. We also provide predictions of extreme overnight borrowing rates using pre-crisis data. The examination of tails (extreme values) provides answers to such issues as to what are the extreme movements to be expected in financial markets; is there a possibility for even larger movements and, are there theoretical processes that can model the type of fat-tails in the observed data? The answers to such questions are essential for proper management of financial exposures and laying ground for regulations.     

Time varying CAPM betas and banking sector risk

This paper employs the [Bai and Perron, 1998] and [Bai and Perron, 2003] structural break methodology to investigate whether the CAPM betas for banking sector stocks are time invariant. I find evidence for three large structural shifts in my monthly (1941.02-2008.01) sample. The third break corresponds with a decline in the perceived riskiness of banking stocks in the period starting in 2000.04. The banking sector was thus priced to be less risky during the period associated with rising leverage and financial sector risk.     

Periodically collapsing Evans bubbles and stock-price volatility

This paper analyzes stock-price volatility in the presence of periodically collapsing Evans bubbles. We derive a volatility formula that establishes a link between the bubble component and stock-price volatility. We demonstrate how to fit the volatility equation to stock-market data.     

Log-concave probability and its applications

Summary. In many applications, assumptions about the log-concavity of a probability distribution allow just enough special structure to yield a workable theory. This paper catalogs a series of theorems relating log-concavity and/or log-convexity of probability density functions, distribution functions, reliability functions, and their integrals. We list a large number of commonly-used probability distributions and report the log-concavity or log-convexity of their density functions and their integrals. We also discuss a variety of applications of log-concavity that have appeared in the literature.Received: 31 December 2003, Revised: 22 March 2004, JEL Classification Numbers: C40, D40, D80. Correspondence to: Ted BergstromWe thank Ken Binmore and Larry Samuelson for encouragement and suggestments.     

VaR-implied tail-correlation matrices

Empirical evidence suggests that asset returns correlate more strongly in bear markets than conventional correlation estimates imply. We propose a method for determining complete tail-correlation matrices based on Value-at-Risk (VaR) estimates. We demonstrate how to obtain more efficient tail-correlation estimates by use of overidentification strategies and how to guarantee positive semidefiniteness, a property required for valid risk aggregation and Markowitz-type portfolio optimization. An empirical application to a 30-asset universe illustrates the practical applicability and relevance of the approach in portfolio management.     

Realized volatility transmission: The role of jumps and leverage effects

This paper is the first to employ a multivariate extension of the LHAR–CJ model for realized volatility of Corsi and Renó (2012) considering continuous and jump volatility components and leverage effects. The model is applied to financial (S&P 500), commodity (WTI crude oil) and forex (US$/EUR) intraday futures data and allows new insights in the transmission mechanisms among these markets. Besides significant leverage effects, we find that the jump components of all considered assets do not contain incremental information for the one-step ahead realized volatility. The volatility of S&P 500 and US$/EUR exchange rate futures exhibits significant spillovers to the realized volatility of WTI. Moreover, decreasing equity prices appear to increase volatility in other markets, while strengthening of the US$ seems to calm down the crude oil market.     

A new measurement method of investor overconfidence

We present an alternative measurement method of investor overconfidence, using unique survey data on stock market predictions of investors. We apply the Parkinson estimate based on extreme bounds around the stock forecast to deduce investor confidence. The results support overconfidence.     

ARDL bounds tests and robust inference for the long run relationship between real stock returns and inflation in Australia

This paper conducts an empirical investigation into the long run relationship between real stock returns and inflation in Australia by employing the ARDL bounds tests. There exists a stock return–inflation long run relationship, and the long run parameters are non-linear functions of those of the conditional error correction model. The OLS estimates of the latter model constitute the long run parameter estimates and their standard errors are estimated by delta methods. The long run model estimates so constructed can be biassed and inconsistent, and the delta method is derived assuming asymptotic normality, which does not hold in this investigation. In this paper, to overcome these limitations of the traditional methods, we employ the bias-corrected bootstrap method. As a consequence, the robust and reliable statistical inference can be made on the long run return–inflation relationship. The empirical results show that the expected inflation had no significant effect on real stock returns, while the observed inflation had a significant and negative effect. Furthermore, the data generating process of the returns–inflation relationship was not affected by the change in monetary policy regime in the early 1990s. These findings imply that Australian stocks have been very effective instruments for hedging against expected inflation. Because of the resilience of Australian economy to the current global financial and economic crisis, this finding has implications for long term domestic and foreign investors in Australia.     

RECURRENT BUBBLES*

We study rational bubbles in a standard linear asset price model. We first consider a class of bubble processes driven by multiplicative i.i.d. shocks. We show that a bubble process in this class either diverges to infinity with probability one, converges to zero with probability one, or keeps fluctuating forever with probability one, depending on investors' “confidence” in expected bubble growth. We call a bubble process having the last property “recurrent.” We develop sufficient conditions for a bubble process to be recurrent when it is driven by non‐i.i.d. shocks, when the risk‐free interest rate is not constant, and when the process is driven by non‐i.i.d. shocks and the risk‐free interest rate is not constant. In the last case we demonstrate via simulation that there can be a prolonged period in which both the bubble and the interest rate stay close to zero.     

An alternative identification of nonlinear dynamic panel data models with unobserved covariates

I provide the nonparametric identification of nonlinear dynamic panel data models. I relax the assumption of covariate evolution in Shiu and Hu (2013) by the results of Hu and Shum (2012). The assumptions include first-order Markov assumptions and a restriction on the evolution of the covariate.     

A ‘technically level playing-field’ profit efficiency analysis of enforced competition between publicly funded institutions

The study demonstrates how recent advances in the data envelopment analysis (DEA) of profit efficiency, and in combining DEA and stochastic frontier analysis (SFA) to measure producer performance, can be used to obtain a ‘technically level playing-field’ profit efficiency performance measure. A three-stage analysis is used to provide an empirical investigation of public-sector universities facing regulatory pressures to attain both profit efficiency and output quality. Stage one employs directional distance function analysis to decompose DEA profit efficiency measures into technical and allocative components. As these performance measures incorporate the impacts of managerial inefficiency, the operating environment and luck (statistical noise), stage two applies SFA to total (radial and non-radial) slacks to obtain a degree of disentanglement from the latter two impacts. The stage-two results are used to define a common environment and common luck scenario, thus permitting university producers’ inputs to be adjusted so as to reflect operation within this more level playing-field scenario. Stage three recomputes the first-stage DEA for this scenario, to yield an (input-focused) ‘technically level playing-field’ measure of profit efficiency that unifies the regulatory financial audit and output-quality indicators in a single, best-practice-benchmarked, performance measure.