This thesis studies the technological change in the US commercial banking market and its influence on banks' lending practices. The second chapter provides some empirical facts.

The third chapter studies the welfare consequences of the destructive creation (bank

branches replaced by internet banking) of the US commercial banking

market following the Great Recession of 2009. Using a structural model,

we find that the cleansing effect (closure of unproductive bank branches)

of the recession increases the units of internet banking by about 56\% in 2016, compared to the case where the cleansing effect is absent. The share of internet banking in the retail service market is increased from 48\% to 60\% and the price of internet banking service is decreased by a factor of 16 by the cleansing effect of the Great Recession.

The two changes lowers the price of retail banking services in 2016 by 37\%: 53\% of the price reduction is attributable

to the replacement

of branches by internet banking and 47\% is attributable to the reduction of the price of internet banking. However, this cleansing effect also

results in a 2.5\% decrease in small business services in small cities.

These findings suggest that the cleansing effect of the recession benefits

retail consumers. However, small business lending may suffer.

The fourth chapter evaluates how information technology (IT) improvements contribute to the decline of small business lending in the US commercial banking market from 2002 to 2017. This paper estimates a general equilibrium dynamic model with banks that differ in size and choose the level of the transaction (hard information intensive) and relationship (soft information intensive) lending. The model shows that banks’ costs of evaluating borrowers’ hard information declined over this period by 46\%, and small business loans fell by 7\% (12\% in the data). This paper finds that banks’ higher reliance on IT to issue transaction loans is responsible for 37\% of the decline in the data, and the consolidation caused by IT improvements caused 22\% of the decline. Contrary to previous findings, this paper finds that when general equilibrium is considered, policy protecting small banks cannot increase small business lending.

This study investigates the impact of portfolio disclosure on hedge fund performance. Using a regression discontinuity design, I investigate the effect of the disclosure requirements that take effect when an investment company's assets exceed $100 million; when that occurs, a fund is required by the SEC to submit form 13F disclosing its portfolio holdings. Consistent with the argument that portfolio disclosure reveals "trade secrets" and also raises front running costs thus harms the funds that disclose, I find that there is a drop in fund performance (about 4% annually) after a fund begins filing form 13F, as well as an increase in return correlations with other hedge funds in the same investment style. The drop in performance cannot be explained by a change in the assets under management or a mean reversion in returns. Consistent with the idea that funds with illiquid holdings tend to employ sequential trading strategies, which increase the likelihood of being taken advantage of by free riders and front runners, the drop in performance is more dramatic for funds that have more illiquid holdings. In addition, I find that the incentive fees paid to fund managers are 1% higher when portfolio disclosure is required, which supports the hypothesis that investors' monitoring of portfolio holdings disciplines adverse risk-taking by fund managers and allows for higher convexity in the optimal compensation structure. Finally, there is a drop in flows into funds that file 13F, which suggests that hedge fund investors negatively value 13F disclosure. Overall, this study suggests that the cost of portfolio disclosure is economically large. It contributes to the policy debate over what constitutes optimal disclosure.