Temporal networks are a subject of intense study in the network analysis theory and applications ( Barabasi, 2005; Tang et al., 2009; Hill & Braha, 2010; Hanneke, Fu & Xing, 2010; Vespignani, 2011; Zhao et al., 2011; Pan & Saramäki, 2011; Perra et al., 2012; Starnini et al., 2012). Except for some preliminary attempts (Scherrer et al., 2008; Hempel et al., 2011; Sikdara, Ganguly & Mukherjee, 2016), temporal networks have not been studied in detail so far with time series econometric models. In this paper, first, we provide an overview of time series statistical estimators to study temporal networks (in particular as related to large networks) and study some of their statistical properties. Second, we apply this framework to the network data on artists and their exhibitions in Mainland China, collected from Artlinkart, a Chinese exhibit platform, covering the period of 1989-2015. A two-mode network analysis framework is developed over time. This allows us to outline the dynamic of the topography of contemporary art emerged after 1989. In this topography the period of 2006-2015 represents a relevant decade for shaping the contemporary art system as the main (private and public) museums of contemporary art have been established. Moreover, this analysis provides information on the position of the different museums as art gatekeepers in selecting and presenting young talented artists. In this regard, we are able to follow the impact of these museums choices over time. According to the signaling theory (Podolny, 2010), our assumption is that the museum choice of selecting an artist at time t, may be adopted by other museums in the following years. The purpose is to identify the key figures (»selectors«) in discovering new talented contemporary artists, whose choices have been confirmed and disseminated by other museums belonging to the same network (»disseminators«). In a young and immature market - as the Mainland China's one - characterised by high uncertainty (Burt, 1992; Podolny, 2001), when two actors initiated a relationship, the tie between them serves as a signal for other market members (Benjamin and Podolny, 1999; Podolny, 2001), which will revise their strategies. In this frame, the capacity to predict the link between actors and the properties of the future network instances using time series as a proxy, may have several market implications. Therefore, we propose a forecast model of time series to predict the properties of the temporal network in Mainland China contemporary art at a later time instance, extending the analysis in Sikdara, Ganguly & Mukherjee (2016) to the low frequency setting. This analysis represents one of the first attempts to map and study the Chinese contemporary art system using network analysis. It also provides important insights into the possibilities to use time series econometric tools to study the dynamics of temporal networks in future.