Search results

Probability of Default (PD) is a crucial credit risk parameter. International accords have motivated banks and credit institutions to adopt objective systems of evaluating and monitoring the PD. This study examines retail unsecured loans of a major Greek bank during the period of the financial crisis. It focusses on the stochastic behaviour of the financial states of the loans. It is tested whether a first-order Markov chain (MC) model describes sufficiently the transitions from one state to another. Moreover, Poisson regression models are estimated in order to calculate the limiting transition matrix, the limiting state probabilities and the PD. It is proved that the MC of the financial states of loans is non-homogeneous suggesting that the transition probabilities from one financial state to another are not constant across time. From the Poisson regression models, the transition probability matrix is estimated from one state to another in alternative time periods. From the limiting transition matrix, it is shown that if a loan is delayed then it is very likely to move towards the next worst case. The findings of this research could be useful for bank management.

As a field, we should put more emphasis on interpreting the magnitude of coefficient estimates rather than only assessing statistical significance. To support this claim, I demonstrate how focusing only on statistical significance can lead to incorrect and incomplete conclusions in many common applications of the linear regression model. Moreover, I demonstrate why interpreting coefficient estimates in common non-linear estimators (e.g., probit, logit, Poisson, and negative binomial estimators) requires additional care compared to the linear regression model.

In this chapter, we explore the relationship between organizational complexity and firm-level innovation. We define and operationalize a new construct, experienced complexity, which is the extent to which the organizational environment makes it challenging for decision makers to do their jobs effectively. We distinguish experienced complexity from structural complexity, which is the elements of the organization, such as the number of reporting lines or integrating mechanisms, that are deliberately put in place to help the organization deliver on its objectives, and we argue that structural complexity correlates positively with firm-level innovation, while experienced complexity correlates negatively with innovation. Using a novel dataset combining survey and objective data on 209 large firms, we find support for our arguments.

It is widely recognised that universities and other research institutions (hereinafter, PRIs) are sources of knowledge in their regional and national economies. They have a broad impact on economic growth through several activities, including educational partnerships, industry-sponsored research, job placement, technical assistance to industry and the creation of start-up firms. This is essentially an issue of knowledge transfer from PRIs to industry, which may take several forms and be either direct or indirect in nature.¹

Purpose – Despite the commonly held stereotype that immigration and crime go hand in hand, there are but a few studies that examine the relationship between immigration and crime across macro-social units, including neighborhoods, cities, and metropolitan statistical areas (MSAs). Even fewer focus on homicide, particularly homicide disaggregated by motive and circumstance. The current study addresses this shortcoming by examining the relationship between immigration and homicide across large cities in the United States.

Methodology – We extend prior work by disaggregating homicide into different “types” based upon motive and circumstance to determine whether immigration is linked not only to overall homicide rates but also to specific types of lethal violence that some suggest may be higher in places where immigrants are more prevalent.

Findings – Cities with greater immigrant concentration have lower homicide rates. There is a significant and fairly strong positive relationship between immigration and gang-related homicides.

Value – This analysis with disaggregated homicide adds to the findings that immigration is not associated with increased crime. Its finding of a correlation between immigration and gang-related homicides points to the next question that needs to be addressed with appropriate data.

People centrally located within networks enjoy a variety of benefits. Why some people achieve such advantage while others do not is not well understood. Using a novel dataset that maps the workflow network among software engineers at a Fortune 500 technology company, I trace the evolution of the position of 804 new entrants to the firm over a period of three years. Findings paint a consistent picture as to the determinants of position in this network: one’s team and one’s job at time of entry, rather than intellectual or social endowment, play the strongest roles in determining one’s subsequent centrality and autonomy.

We know a great deal about the ways in which routines of news coverage may bias newspaper content, but little about how different article retrieval practices influence newspaper data assembled by scholars. Using the New York Times as a source of data on social movement activity, we compare depictions of protest by the African-American Civil Rights movement over time produced using the two most common article retrieval methods: index versus full-story coding. Full-story coding clearly offers more depth and greater breadth in terms of the events identified. Moreover, many of the same event characteristics associated with selection bias in newspaper reporting (e.g., size and confrontational nature of a protest event, presence of counter-demonstrators or police, and event sponsorship by a recognized social movement organization) are selected upon again when stories are indexed by New York Times staff.

An important precondition for resource redeployment is that firms are aware of the commercial applications for which their resources can be used. We take an inventing-firm perspective and ask: how many new commercial applications will a firm associate with an existing technological invention? We note that both technological and organizational characteristics determine the number of distinct applications firms consider feasible for a given technological invention. In particular, we suggest that inherently fungible technologies, that is, technologies that have a broad impact on other technological fields (highly general technologies), will be associated with a larger set of commercial applications. We also suggest that linking applications to an inherently general technology can be challenging when the technology is already embedded in organizational (commercial) routines. Proprietary data from an online marketplace allow us to investigate the applications firms consider feasible for their technological inventions. In line with extant work, a firm assigns a greater number of applications to more general technologies. As expected, however, this relationship is shaped by how the technology is embedded within the organization. Our results have implications for redeployment as firms may face challenges in the initial step of redeployment when fungible resources need to be linked to emerging market opportunities.

The causal relationship between the size of welfare benefits and the birth decisions of women on welfare has been explored in a number of studies using a variety of analytical approaches applied to vital statistics data, data from the Current Population Survey, the Panel Study of Income Dynamics, or similar survey data. These studies typically use non-experimental methods to relate differences in birth rates or birth decisions across states to differences in welfare benefits levels. Analyses of this type have been criticized on several grounds. Benefits across states may be correlated with unobserved interstate differences that may also be related to birth decisions. Very often, these studies measure the key independent variable, welfare benefits level, as the cash benefit guarantee under the Aid to Families with Dependent Children (AFDC) program for a household of fixed size, varying this amount by state of residence. Actual benefits paid will vary with household size, number of AFDC-eligible household members, other sources of income, and other factors.

We propose a generic algorithm for numerically accurate likelihood evaluation of a broad class of spatial models characterized by a high-dimensional latent Gaussian process and non-Gaussian response variables. The class of models under consideration includes specifications for discrete choices, event counts and limited-dependent variables (truncation, censoring, and sample selection) among others. Our algorithm relies upon a novel implementation of efficient importance sampling (EIS) specifically designed to exploit typical sparsity of high-dimensional spatial precision (or covariance) matrices. It is numerically very accurate and computationally feasible even for very high-dimensional latent processes. Thus, maximum likelihood (ML) estimation of high-dimensional non-Gaussian spatial models, hitherto considered to be computationally prohibitive, becomes feasible. We illustrate our approach with ML estimation of a spatial probit for US presidential voting decisions and spatial count data models (Poisson and Negbin) for firm location choices.