Black-Necked Swans and Active Risk Management

Introduction

Taleb (2007) proposed the Black Swan Theory to explain high-impact, hard-to-predict and rare events, which are far beyond the realm of normal expectations, i.e., black swans. He did not consider, however, what to do with the black swan’s cousin, the black-necked swan. The black-necked swan, unlike the Australia-bound black swan, is born covered in white, but its neck gradually darkens from 3 months onward. Hence there is the “black-necked swan problem” concerning whether a given swan born white will stay white. And yet, an even harder problem is when and where it might possibly turn black (See Figure 1).

Figure 1.

Black-necked swan

Risk management should take into consideration the possible appearances of these “black swans” and “black-necked swans”. Once in a while a driver gets a traffic ticket; he/she starts to watch out for a few weeks and then forgets about everything until another traffic violation. Good risk management practice entails active vigilance and continuous monitoring of the financial situation and relevant macroeconomic variables. Early warning systems, surveillance technologies, and accurate, up-to-date and easily accessible databases are important tools to support this kind of active risk management.

This chapter describes statistical models and methods for on-line surveillance and early warning/quick detection of risk associated with “black-necked swans” and even “black-swans”. Traditional risk management considers risks associated with adverse events that occur with substantially higher, albeit small, probability. We first give a brief review of a number of important ideas in traditional risk management that is mostly passive and precautionary in nature. We then refine some of these ideas for active risk management, which also involves sequential detection of structural changes and surveillance of financial risks. In this connection, recent advances in on-line surveillance and sequential detection are also reviewed. We conclude with some remarks and further discussion of the advantages of active over passive risk management in today’s complex financial markets.

Measures Of Financial Risks And The Basel Accord

In this section, we first summarize different categories of financial risks and commonly used risk measures; a classic reference on these topics is Jorion (2004). Then we review the regulatory background underlying statistical methods for risk management, focusing primarily on the Basel Accord. The Basel Accord was developed in 1988 by the Basel Committee on Banking and Supervision (2005, 2006) and was later widely endorsed by banks in G10 countries. It provides a unified statistical approach to measuring the risks in banks and specifying the amount of capital needed to hedge against these risks. The first Basel Accord introduced 8% as the minimum solvency ratio, defined to be the ratio of capital to risk-weighted assets. We focus primarily on the internal ratings models, introduced by the second Basel Accord (Basel II), which use rigorous statistical models to calibrate the solvency ratio for each bank.