J. Shanghai Jiaotong Univ. (Sci.), 2016, 21(6): 655-661 DOI: 10.1007/s12204-016-1778-0 Scaling Behavior of Bulk Freight Rate Volatility Before and After Noise Reduction DAI Siyu 1 (), CHEN Feier 1,2∗ (), ZENG Yadong 1 (), ZENG Xin 3 ( ) (1. State Key Laboratory of Ocean Engineering, School of Naval Architecture, Ocean and Engineering, Shanghai Jiaotong University, Shanghai 200240, China; 2. Collaborative Innovation Center for Advanced Ship and Deep-Sea Exploration, Shanghai 200240, China; 3. College of Transport and Communications, Shanghai Maritime University, Shanghai 201306, China) © Shanghai Jiaotong University and Springer-Verlag Berlin Heidelberg 2016 Abstract: Analysis of freight rate volatility characteristics is a hot topic after year 2008 due to the effect of financial crisis in marine transportation. In this paper, we analyze the bulk freight rate index by detrended fluctuation analysis (DFA) method and discover the scaling behavior. Switching points (SPs), as the indicators of scaling behavior, can be eliminated after Vondr´ak noise reduction technique. Therefore, we conclude that high- frequency noise is the cause of SP. Key words: bulk freight rate, noise reduction, detrended fluctuation analysis (DFA), Vondr´ak filtering, scaling behavior CLC number: F 551 Document code: A 0 Introduction As the freight rate market has gradually transformed into a market where freight rate can be bought and sold for investment purposes like any other financial asset or commodity, more attention has been paid to the dy- namics and volatility of freight rate market [1] . Freight rate volatility represents the fluctuation or dispersion of the freight rate in shipping market. In finance, the anal- ysis methodology for forecasting the direction of prices by means of the study of past market data is called technical analysis. The unique properties of freight rate market, such as the lag when responding to information change, the specialty of non-storable commodity and the involvement of non-shipping market participants, result in the insufficiency of fundamental analysis and the indispensability of technical analysis [2] . In this pa- per, the Baltic dry index (BDI), which is the under- lying asset of forward freight agreement (FFA) and is Received date: 2015-07-23 Foundation item: the MOE (Ministry of Education in China) Project of Humanities and Social Sciences (Nos. 12YJCGJW001 and 14YJC630008), the Inter- discipline Foundation of Social Science and Engineering of Shanghai Jiao Tong University (No. 15JCMY11), the Fund of Center for Teaching and Learning Development of Shanghai Jiao Tong University (No. CTLD16B3002), the National Students Innovation Program of China (Nos. 201610248001 and IPP12002), and the National Natural Science Foundation of China (No. 51409157) ∗E-mail: chenfeier@sjtu.edu.cn considered by the investment community as a leading indicator of economic activity in shipping market [2] , is investigated and analyzed. A phenomenon in freight rate volatility that fas- cinates us is the scaling behavior, which means the volatility of freight rate index will show different char- acteristics when divided by different time steps. The scaling behavior of complex systems with a great va- riety of agents of different perspectives and conflicting interests has long been the focus of discussion [3-6] . As is widely known, different types of investors with differ- ent trading horizons exist in freight rate market. Short term traders such as market makers, voyage charterers and speculators mainly pay attention to the dynam- ics of freight rate in shorter time scale. They reevalu- ate the market situation and execute transactions at a relatively high frequency. Trades are often with large volume for a minor deviation of prices. Nevertheless, long-term participants, such as shipowners, banks and the government, concentrate primarily on the long-term market trends and execute the transactions when large price movements indeed happen over a long trading horizon. Existence of diverse trading horizons in freight rate market attaches great significance to the analysis of scale properties. Switching points (SPs), as the in- dicators of scaling behavior, are also worth analyzing. Therefore, we are left with the problem of developing a method which can provide instructions for market par- ticipants with different trading horizons.