The distribution of the benefits of waterfront reform

In this paper we introduce, for the first time, a methodology from the most recent literature of finance to reveal the duration of shipping cycles and then show the benefit of the use of this information to make more successful shipping loans. This is so as banks are willing to finance, during boom periods, shipping loans for new buildings but by this way ‘create’ oversupply and thus depress the freight market by their own actions. The information about cycles, especially their forecasting, is mostly important as shipping loans are based on project financing/cash flow financing, which means that ship revenue is of utmost importance. The Rescaled Range Analysis is applied here to 379 monthly freight trips—made stationary—between 1971 and 2002 (July), due to Hurst 1 Hurst, HE. 1950. Long-term storage capacity of reservoirs. April1950. pp.770–808. American Society of Civil Engineers. Paper No. 2447 [Google Scholar] and elaborated and popularized by Mandelbrot 2 Mandelbrot, BB. 1975. Stochastic models for the earth's relief, the shape and the fractal dimensions of the coastlines, and the number-area rule for islands. Proceedings of the National Academy of Sciences. 1975, USA 72. pp.3825–3828. [Crossref] , [Google Scholar]. The most important effect, however, is that shipping freight series exhibit non-normality and long-run dependence rendering the use of random walk models such as GARCH (Generalized Autoregressive Conditional Heteroscedasticity) problematic. Thus an adequate literature review is carried out with criticism against the models used. The cycles have been calculated as equal to 4.5 years and 2.25 years. This is almost compatible with the most recent paper of Stopford 3 Stopford, M. 21 September 2001. “Forecasting the dry bulk, tanker and container markets”. In Maritime Cyprus 21 September,  [Google Scholar]. The Hurst exponent was found equal to 0.93, alternating over the periods examined (0.65, 0.73, 0.62, 0.59 and 0.55) and indicating long-term persistence but seriously away from normal/random walk domain. Most studies have said the same using the Jarque--Bera test for normality but provided no alternative.