Book excerpt: The thesis focuses on understanding the dynamic nature of contracts used in various economic context, specifically financial economics and industrial organization. The first chapter "A Theory of Dynamic Contracting with Financial Constraints'' draws on a large empirical literature documenting that small businesses are financially constrained, and operate at an inefficient level. In the paper, we build a theoretical model where financial constraints arise endogenously as a product of interaction between persistent agency frictions and agent's inability to raise external capital.The paper makes two general points. First, efficiency is a certainty in the long run, and it is achieved through monotone slacking of financial constraints. Second, persistence makes the path towards efficiency much more constrained in comparison to the model with the iid technology. In particular, we show that dynamic agency models with persistence predict a larger cross section of firms in the economy to be financially constrained.At a technical level, we invoke the recursive approach of \citet{aps}, using a two-dimensional vector of promised utilities as a state variable. We show that the optimal contract always stays in a strict subset of the recursive domain termed the shell, and the optimal contract is monotone within this set. We also verify that the results continue to hold in continuous time.The second chapter "Dynamic Contracts with Unequal Discounting'' looks at dynamic screening with soft financial constraints. In contrast to the first paper, the agent can raise money but at a different rate than the principal.We solve for the optimal contract and show that efficiency is not attainable with soft financial constraints. Therefore, the predictions of dynamic models of mechanism design are not robust to the assumption of equal discounting. For the large set of parameters, the optimal contract has the restart property- dynamic distortions are a function of the number of consecutive bad shocks, and once the good shock arrives the process repeats again. We also show that restricting attention to contracts which have the restart property is in general approximately optimal. The endogenous resetting aspect of restart contracts shares features of various contracts used in practice.In the third chapter "On Dynamic Pricing'', we explore dynamic price discrimination, extending a canonical model of monopolistic screening to repeated sales, where a seller uses timing of purchases as a screening instrument. The importance of time as an instrument for price discrimination has been understood since Varian [1989].In the paper, we are aiming to provide a formal analysis of pricing strategies to discriminate amongst consumers based on the timing of information arrival and/or the timing of purchase.A seller repeatedly trades with a buyer. Buyer's valuations for the trade follow a renewal process; that is, they change infrequently at random dates. For the model with two periods, We show that selling the first period good for a spot price and selling the second period good by optioning a sequence of forwards is the optimal pricing strategy. Specifically, at the outset, the seller offers an American option which can be exercised in each of the two periods. Exercising the option grants the buyer with a forward- an obligation to purchase the second period good for a specific price, and a strike price- a right to buy (or not) the good in the second period after learning his value. The buyer with a high valuation exercises the option in the first period, whereas one with a low valuation waits until the second period and then takes a call.We extend the analysis to the general continuous time renewal processes and assess the performance of price discrimination based on American options on forwards:i.optioning forwards is shown to be the deterministic optimum for the sequential screening problem- when the seller makes a sale in a single fixed period;ii.optioning forwards is shown to be the exact optimum for the repeated sales problem in the restricted class of strongly monotone contracts- when allocative distortions are monotone in a whole vector of buyer's valuations;iii.the optimum for the repeated sales problem in the unrestricted class of contracts is shown to be backloaded and a theoretical bound is provided for the fraction of optimal revenue that can be extracted by optioning forwards.Finally, the construction of dynamic pricing mechanism and bounds is ported to study repeated auctions.