Some properties of international maritime statistics

This paper tests two fundamental hypotheses concerning international maritime statistics. The first one deals with the question of stationary of the maritime market statistics. The second hypothesis tested is the assumption that the international maritime statistical time series are not distributed according to a normal of Guassian probability law, but rather belong to the same family of distributions with distinctly different critical parameters. Through well documented statistical methods, the paper concludes that the international freight rates observed on a day to day basis are generated by a random walk process. The paper finds that the shipping industry's conventioal wisdom is essentially correct. 'Last done' is as good a forecast of tomorrow's freight rate as any other generated by more sophisticated forecasting methods. Furthermore, freight rates and secondhand tonnage prices fluctuate closely together. The freight rates are generated by stochastic processes fully described by the Paretian family of distributions. The critical parameters of these distribution, the characteristic exponents, are such that the risk conscious ship operator can indeed reduce his exposure to risk by securing a correct set of freight rate contracts.