Welcome to another issue of The Journal of Operational Risk. In this issue we would like to welcome a new editorial board member, Dr Anna Chernobai from the University of Syracuse. Dr Chernobai has been helping us tremendously for quite a while and has also been working towards the development of operational risk modeling with a number of high-quality articles and books. It is an honor for us to have her on our editorial board. When the Journal started four years ago Dr Chernobai was a fresh PhD graduate and since then has made an amazing contribution to the area.
In March I had the opportunity to speak and attend the OpRisk & Compliance 2009 (USA and Europe) conferences organized by our sister title. It was inspiring to realize that the number of participants has kept stable compared with previous years despite the crisis. It is a good sign that banks and corporations are still keen to invest in the education of their risk managers. Obviously the current crisis was at the forefront of most discussions. I feel energized by the mindset of the participants who have kept their minds open to the necessary changes in risk management that must follow. I agreed with the panelists and most of the participants who recognized the failed risk governance as one of the key points that needs to be fixed. I am sure that these discussions will permeate risk management in the foreseeable future.
Regarding the state of operational risk research, I would like to ask potential authors not to feel discouraged by the crisis to submit to the Journal. We have seen a reduced flow of papers recently that we are blaming on the current economic environment. I would like also to re-emphasize that the journal is not just for academics to publish in. We at The Journal of Operational Risk incentivize readers to submit papers to the "Forum" section. This section is aimed at discussion of current events without too much concern with the technical aspect, formulas and mathematics. We at the journal will be extremely happy to see more submissions with more practical, current views of relevant matters that affect your day-to-day work.
In this issue we bring you four research papers. This is a particularly exciting issue in which we bring a paper that deals with a new frontier for operational risk that is tackling business control from an engineering standpoint as well as costs and productivity. I highlight the role that sophisticated Bayesian techniques like Markov chain Monte Carlo are playing in the state-of-the-art modeling of operational risk. Two articles deal with the subject. There is also an article on modeling tails of popular operational risk severity distributions. A very interesting issue indeed. In the first paper, "Operational risk management with process control and business process modeling", Cernauskas and Tarantino propose an approach for modeling, monitoring and controlling operational risks in financial institutions based on a methodology that integrates business process modeling with statistical and engineering process controls. The authors address an interesting combination of efficiency improvement and operational risk, which in my view is the next frontier for operational risk.
In the second paper, "Modeling operational risk data reported above a time varying threshold", Shevchenko and Temnov, both regular contributors to the journal, tackle the issue where operational risk losses are usually reported above a certain threshold. They present maximum likelihood estimation and Bayesian Markov chain Monte Carlo approaches that provide a better fit for a time varying threshold.
In the third paper, "A comparison of tail performance of the Champernowne transformed kernel density estimator, the generalized Pareto distribution and the g-and-h distribution", Buch-Kromann introduces a new tail-dependent parameter estimation method for the Champernowne distribution, computed by conditional maximum likelihood, and show that, by using this new method, an estimator is obtained that in general outperforms the benchmark estimators with respect to tail performance. At the same time the new estimator provides a density estimate on the entire axis superior to the g-and-h distribution, and unlike the generalized Pareto distribution estimator, which provides a density estimate only above the threshold. The estimator's performance is investigated in a Monte Carlo simulation study and their application to operational risk is illustrated.
In the fourth paper, "Dynamic operational risk: modeling dependence and combining different sources of information", Peters, Shevchenko and Wüthrich model dependence between operational risks by allowing risk profiles to evolve stochastically in time and to be dependent allowing for a flexible correlation structure. Therefore, the dependence between frequencies of different risk categories and between severities of different risk categories, as well as within risk categories, can be modeled. The model is estimated using the Bayesian inference methodology, allowing for a combination of internal data, external data and expert opinion in the estimation procedure.
Comparison of tail performance of the Champernowne transformed kernel density estimator, the generalized Pareto distribution and the g-and-h distribution