Technical trading rule performance in the second-hand asset markets in bulk shipping

Type/no A04/00
Author Roar Os Ådland
This paper investigates the performance of technical trading rules applied to asset play in the international bulk shipping industry. The rules are tested on monthly returns in the Products Carrier segment for the period 1981 to 1998. The paper evaluates 1053 different parameterizations of three of the simplest and most popular trading rules in the financial markets: filter rules, moving averages, and support and resistance levels. Overall, the results provide strong support for the technical strategies. None of the trading rules generate negative cumulative wealth, and only one parameterization results in a mean return that is lower than the return from the benchmark buy-and-hold strategy. The results for the best-performing trading rule show that the mean return following buy signals is positive and the mean return following sell signals is negative, both significantly different from the buy-and-hold mean return according to standard statistical tests. Moreover, the returns following buy signals are less volatile than those following sell signals, as well as the returns of the buy-and-hold strategy. The best-performing trading rule obtains a mean return of 35.4% p.a. above the buy-and-hold annual return of 4.0%. Due to a low number of trades, the introduction of trading costs has little impact on the results. Adjusting for data-snooping biases according to White's Reality Check bootstrap methodology confirms the conclusion that the best-performing trading rule provides superior investing performance. However, the practical implementation in an illiquid market may reduce the theoretical excess return of the best-performing trading rule to a level where it is no longer significant. Moreover, the probability that an investor could have picked ex ante a trading rule with statistically significant excess return is small.
Language Written in norwegian