The occurrence of periods of stress in international financial markets has been a challenge to economists and i nancial researchers for long. Academic interest in exploring underlying economic forces goes back, at least, to the great depression of the 1920s.
More recent prominent examples of periods of stressed markets include the crash of 1987, the so-called 1989 mini-crash, the 1997 Asian currency crisis, and the 1998 Russian debt crisis, which caused the collapse or near collapse of several financial intermediaries worldwide. Also, a purely exogenous event such as September 11, 2001 caused substantial market stress.
Furthermore, various other recent cases of market stress and related cases of individual financial distress add to a growing interest in understanding periods for stressed markets. Sometimes, although less frequently, stressed markets may also relate to run-ups in prices; a large stock market run-up with the start of the Gulf War in January 1991 may serve as an example.
Also, market stress may sometimes not at all be obvious from overall period price changes; a typical example would be frequently observed trading patterns such as so-called "one-day reversals". These tend to occur under high intraday volatility as well as hectic trading but typically do not end with large overall price movements on a market close-to-close basis.
Th e above observations led to the concept of "stressed markets", which is assigned to situations during which unusual economic circumstances prevail. One then distinguishes related market behavior from what is otherwise assumed to be "normal". This concept makes the analysis of markets under stress a separate and relatively recent research topic.
A central characteristic of stressed markets is heavy intraday trading activity that goes along with high intraday price volatility. Such periods are obviously of particular relevance for risk management as well as financial engineering applications.
During recent decades, important advances have been made in the area of economic models, which aim at an explanation of market stress and the occurrence of crashes. Stress dynamics critically depend on the nature and diversity of market participants, their motives for entering the market, and the extent of consistency in their response to worsening conditions.
While economic models make assumptions about the underlying market structure, the potential diversity in the economic background also calls for empirical methods in the study of market stress. Such methods include quantitative approaches in finance, which take care of the special stress circumstances as well as frequent approaches based on extreme value theory.
How do securities behave in situations of market stress? Important empirical observations of market behavior under stress, which is dif erent from normal behavior, include two main areas: nonlinear crosssectional dependence between and liquidity and nonlinear dependence between asset returns.
Consider the following first point: nonlinear cross-sectional dependence between assets returns relates to a typical observation under periods of market stress in that asset return correlations seem to be different than under normal market conditions.
In particular, correlations during strong market downturns seem to be higher than otherwise. This "diversification meltdown" may partly explain increased risk and the sharp movements in overall market indices. However, there is evidence that behavior is stable in a statistical sense.
In other words, it is a standard feature of a complex asset return dependence structure, which is nonlinear and not fully described by a linear dependence concept such as correlation. The feature obviously affects risk-management decisions and also indicates that standard methodologies would underestimate risk during market stress.
The second point is even more involved since we commonly assume that market volatility under normal market conditions is driven by information arrival and trading activity. However, under market stress, the patterns may not hold. Liquidity, which is a constant side variable under normal market conditions, starts to play a dominant role under market stress.
As such, nonlinear asymmetric relationships appear. The Bank for International Settlement’s Quarterly Report for the year 2000 commented that: "The illusion of permanent market liquidity is probably the most insidious threat to liquidity itself".
While markets, typically, become more liquid as prices rise and more participants enter, they become sticky when many participants want to exit at the same time. Risk management is af ected by potential market stress, which implies that common assumptions on market mechanics are violated. Such assumptions include that the liquidation of a position would have no ef ect on the market, that positions can be liquidated in a relatively short time period, and that the bid-offer spread remains stable.
As such, empirical evidence indicates that during times of stress, bid-offer spreads widen and market depth may become asymmetric between the buy and the sell side. Also, the ef ect of order flows on price movements becomes stronger. At the same time, no single measure so far seems fully appropriate to capture market liquidity or liquidity risk.
More recent prominent examples of periods of stressed markets include the crash of 1987, the so-called 1989 mini-crash, the 1997 Asian currency crisis, and the 1998 Russian debt crisis, which caused the collapse or near collapse of several financial intermediaries worldwide. Also, a purely exogenous event such as September 11, 2001 caused substantial market stress.
Furthermore, various other recent cases of market stress and related cases of individual financial distress add to a growing interest in understanding periods for stressed markets. Sometimes, although less frequently, stressed markets may also relate to run-ups in prices; a large stock market run-up with the start of the Gulf War in January 1991 may serve as an example.
Also, market stress may sometimes not at all be obvious from overall period price changes; a typical example would be frequently observed trading patterns such as so-called "one-day reversals". These tend to occur under high intraday volatility as well as hectic trading but typically do not end with large overall price movements on a market close-to-close basis.
Th e above observations led to the concept of "stressed markets", which is assigned to situations during which unusual economic circumstances prevail. One then distinguishes related market behavior from what is otherwise assumed to be "normal". This concept makes the analysis of markets under stress a separate and relatively recent research topic.
A central characteristic of stressed markets is heavy intraday trading activity that goes along with high intraday price volatility. Such periods are obviously of particular relevance for risk management as well as financial engineering applications.
During recent decades, important advances have been made in the area of economic models, which aim at an explanation of market stress and the occurrence of crashes. Stress dynamics critically depend on the nature and diversity of market participants, their motives for entering the market, and the extent of consistency in their response to worsening conditions.
While economic models make assumptions about the underlying market structure, the potential diversity in the economic background also calls for empirical methods in the study of market stress. Such methods include quantitative approaches in finance, which take care of the special stress circumstances as well as frequent approaches based on extreme value theory.
How do securities behave in situations of market stress? Important empirical observations of market behavior under stress, which is dif erent from normal behavior, include two main areas: nonlinear crosssectional dependence between and liquidity and nonlinear dependence between asset returns.
Consider the following first point: nonlinear cross-sectional dependence between assets returns relates to a typical observation under periods of market stress in that asset return correlations seem to be different than under normal market conditions.
In particular, correlations during strong market downturns seem to be higher than otherwise. This "diversification meltdown" may partly explain increased risk and the sharp movements in overall market indices. However, there is evidence that behavior is stable in a statistical sense.
In other words, it is a standard feature of a complex asset return dependence structure, which is nonlinear and not fully described by a linear dependence concept such as correlation. The feature obviously affects risk-management decisions and also indicates that standard methodologies would underestimate risk during market stress.
The second point is even more involved since we commonly assume that market volatility under normal market conditions is driven by information arrival and trading activity. However, under market stress, the patterns may not hold. Liquidity, which is a constant side variable under normal market conditions, starts to play a dominant role under market stress.
As such, nonlinear asymmetric relationships appear. The Bank for International Settlement’s Quarterly Report for the year 2000 commented that: "The illusion of permanent market liquidity is probably the most insidious threat to liquidity itself".
While markets, typically, become more liquid as prices rise and more participants enter, they become sticky when many participants want to exit at the same time. Risk management is af ected by potential market stress, which implies that common assumptions on market mechanics are violated. Such assumptions include that the liquidation of a position would have no ef ect on the market, that positions can be liquidated in a relatively short time period, and that the bid-offer spread remains stable.
As such, empirical evidence indicates that during times of stress, bid-offer spreads widen and market depth may become asymmetric between the buy and the sell side. Also, the ef ect of order flows on price movements becomes stronger. At the same time, no single measure so far seems fully appropriate to capture market liquidity or liquidity risk.
 |
| Stressed Markets |
