Benefits of multi-operator spectrum sharing in wireless networks heavily depend on the traffic misbalance in the networks belonging to different operators. In this paper, we study the likelihood that such misbalance occurs in networks with high traffic dynamics. An extensive business portfolio for heterogeneous networks is presented to analyse the benefits due to multi-operator cooperation for spectrum sharing. High resolution pricing models are developed to dynamically facilitate price adaptation to the system state. By using queuing theory, we quantify the operators' gains in cooperative arrangements as opposed to non-cooperative independent operation. In addition, Markov model is used that can handle wider range of different distributions of traffic arrivals and service rates. A tractable analysis and quantitative results are provided for those gains as a function of the number of cooperating operators. Under the condition that there is a traffic underflow in one band, it has been shown that with capacity aggregation model, the operator operating in other band can take advantage of additional channels with probability close to 1. In capacity borrowing/leasing model, this advantage is not unconditional, and there is a risk that the operator leasing the spectra will suffer temporary packet losses. When cognitive models are used in a network with high traffic dynamics, 50–70% of the spectra may be lost due to channel corruptions caused by the return of primary users. The gains of traffic offloading from a cellular network to a WLAN are quantified by an equivalent increase in opportunistic capacity proportional to the ratio of aggregate coverage of cellular networks and WLANs. The results provide guidelines for business decision in multi-operator network management.