We propose a model of limited arbitrage based on financial relationships. Financially constrained arbitrageurs may choose to seek additional financing from banks that have the technology to profit from the strategies themselves. A holdup problem arises because banks cannot commit to providing capital. To minimize competition, arbitrageurs will choose to stay constrained and underinvest in the arbitrage unless banks have sufficient reputational capital. This problem arises when mispricing is largest. More competition among financiers, higher arbitrageur wealth, and allowing for explicit contracts can worsen the holdup problem. When arbitrage is risky, financial relationships are more valuable, mitigating the problem

We analyze strategic speculators’ incentives to trade on information in a model where firm value is endogenous to trading, due to feedback from the financial market to corporate decisions. Trading reveals private information to managers and improves their real decisions, enhancing fundamental value. This feedback effect has an asymmetric effect on trading behavior: it increases (reduces) the profitability of buying (selling) on good (bad) news. This gives rise to an endogenous limit to arbitrage, whereby investors may refrain from trading on negative information. Thus, bad news is incorporated more slowly into prices than good news, potentially leading to overinvestment

We study an economy populated by three groups of myopic agents: constrained agents subject to a portfolio constraint that limits their risk taking, unconstrained agents subject to a standard nonnegative wealth constraint, and arbitrageurs with access to a credit facility. Such credit is valuable as it allows arbitrageurs to exploit the limited arbitrage opportunities that emerge endogenously in reaction to the demand imbalance generated by the portfolio constraint. The model is solved in closed-form, and we show that, in contrast to existing models with frictions and logarithmic agents, arbitrage activity has an impact on the price level and generates both excess volatility and the leverage effect. We show that these results are due to the fact that arbitrageurs amplify fundamental shocks by levering up in good times and deleveraging in bad times