Bitte verwenden Sie diesen Link, um diese Publikation zu zitieren, oder auf sie als Internetquelle zu verweisen: https://hdl.handle.net/10419/169168 
Erscheinungsjahr: 
2015
Quellenangabe: 
[Journal:] SPOUDAI - Journal of Economics and Business [ISSN:] 2241-424X [Volume:] 65 [Issue:] 1/2 [Publisher:] University of Piraeus [Place:] Piraeus [Year:] 2015 [Pages:] 67-86
Verlag: 
University of Piraeus, Piraeus
Zusammenfassung: 
The paper deals with maritime risk, which we consider important, no doubt, for ship-owners acting in volatile markets. Traditionally, risk is measured by "standard deviation". Other risk measures like "excess kurtosis", "excess skewness", "long-term dependence" and the "catastrophe propensity" were ignored. Risk in 1900 was based on the mathematical laws of Chance and influenced greatly by Probability theory due to Pascal and Fermat (1654). Economists, but maritime ones, have understood, however, that the "random walk" model, and the "efficient market hypothesis", failed to interpret reality since Black Monday (1987) at least. The traditional treatment of risk assumes that 95% of the observations fall within 2s from their mean. However, the daily data of 4 time-charter routes ("Baltic Panamax Index", May 1996-February 2012) showed otherwise. Moreover, variance varies from one decade to next, even under stable mean. Risk is related to dispersion, which is defined the same in "normal" and "chaotic dynamic systems". All maritime studies (1997 - 2013), however, reported excess skewness, excess kurtosis, absence of normality and serial correlation...but no remedy provided. As far as the reference to the assumption that observations are "independent and identically distributed" is concerned, maritime time series analysis shows "long term dependence" indicated by a high "Hurst exponent"~1. The paper uses "Rescaled Range Analysis"-a nonparametric method, to identify the "Noah effect" (i.e. the propensity of time series towards catastrophe; measured by alpha exponent). Combined with nonlinear forecasting methods, short and long term risk is thus in this paper forecast. Finally, it shows using daily data, that "risk and dependence" vary on data's calendar time used.
Schlagwörter: 
Risk
excess kurtosis-skewness risk
non-normality risk
long term dependence
"Noah" effect
forecasting catastrophe's risk
BPI 1996 - 2012
Hurst exponent
JEL: 
C65
C53
E44
G17
Dokumentart: 
Article

Datei(en):
Datei
Größe
1.37 MB





Publikationen in EconStor sind urheberrechtlich geschützt.