| This thesis develops a two period model of a bank in which size of bank, and portfolio decisions matter to the ability and willingness of banks to make small business loans. We distinguish between large and small banks. All banks face both default risk and liquidity risk in their lending and deposit holding decisions. All banks must make illiquid lending decisions prior to knowing what their draw of the state of the world is. Large banks may make small business loans, loans to large businesses, mortgage loans, and may hold deposits, and buy and sell funds on the inter-bank market. Small banks may not make loans to large businesses. Unlike other work on bank behavior, both risks are conceptualized as arising endogenously as a result of portfolio decisions managers make.; The model sets up two "Cases". In the first, all loans except mortgage loans are illiquid. In this "Case" bank managers face one of two "Situations". In the first, managers face a given probability of facing either default risk or liquidity risk singly. In the second, they face the probability of facing both risks simultaneously, or neither risk. Uncertainty regarding future interest rates on loans and deposits coupled with uncertainty over risks leads managers to choose certain portfolios. If they guess wrong, then they face costs of this mistake. Managers may choose to hold more of fewer liquid deposits or illiquid loans, depending on what they expect. We model the optimal level of small business lending.; In the second "Case" Commercial and Industrial loans to large businesses are securitizable, and hence salable. Small business loans remain illiquid. Facing the same "Situations", but now with more options, we model the optimal level of small business lending by large and small banks. Small banks cannot originate securitized loans, but can purchase them to increase the diversity and/or liquidity of their balance sheets.; The model is tested econometrically, and is shown to work banks, but not for small. |