This dissertation consists of two chapters. Chapter 1 focuses on the barriers that diversifying companies could face and explore how barriers to entry differ across different types of entry. Chapter 2 turns the attention to the market competition among insurance companies that are already in a market and examines how product bundling impact insurers' market power. Chapter 1 proposes and estimates a multi-agent model of entry. The prior literature often treats the number of companies in a market as an exogenous measure of market structure. However, the number of companies is endogenously decided by the market structure and other participants. Thus, I propose a structural model of entry to address the endogenous entry decision. In addition, the estimations are conducted at each market-year level, therefore, it provides an opportunity to delineate the relative importance of barriers to entry across three dimensions: geographic, product, and time. I find that barriers to entry exist in the financial services industry, and can be quite substantial to the \textit{de novo} entrants. Overall, I find \textit{de novo} entrants are the ones most subject to barriers to entry across all markets. Expanding within a state is as costly as expanding within a product line. Upon further examination, I discover that product-specific knowledge, such as underwriting expertise, pricing schemes, and coverage designs, plays a critical role in a successful expansion. This information is also relatively more important than state-specific connections, such as how well the company knows its customers and connections with distribution channels. Among all product lines, I find that expertise in mortgage guaranty insurance creates the most barriers, and these barriers are most subject to impacts of the financial crisis. In Chapter 2, I turn the focus to the market competition \emph{within} a market and explore the impact of product bundling on market power. Product bundling is a popular way for companies to retain their customers and keep up with fast-changing market demand. In this chapter, I will specifically examine the impact of bundling on price elasticity for personal lines of insurance. Insurance demand estimation is well-explored in the literature because it is difficult to obtain individual-level data. I overcome this hurdle by using a random coefficients logit model, which incorporates flexible consumer preferences over companies' characteristics. The second difficulty in insurance demand estimation is that it is hard to find a good instrument for the endogenous price. Therefore, I propose a novel instrument, which exploits an idiosyncrasy in insurance tax laws for identification. I find that bundling, on average, can reduce consumers' price sensitivity. Thus, companies that can offer bundle-able products experience a less elastic demand and achieve market power. However, product bundling has differential impacts on the auto insurance and homeowners' insurance markets. Auto insurers that offer bundled packages experience less elastic demand in response to price increases. However, we do not observe similar patterns in the homeowners' insurance market, where doing so intensifies price elasticity. With a closer examination, we discover that the different valuation in homeowners is not driven by the financial ratings of insurers. This indicates that homeowners tend to value other characteristics, such as claims management and the quality of service, more than just price of the contract.