This dissertation analyzes micro-level data and employs a general equilibrium model to study heterogeneous firm-level responses to aggregate shocks, underlying sources of the observed difference in firm's decisions, and industry and country-level implications. The first chapter investigates how cross-sectional micro-uncertainty influences the investment of small and large firms and discusses the aggregate implications of the heterogeneity in their investment decisions. Empirically, we find that large firms show less investment decline in times of heightened uncertainty. We provide empirical evidence for the underlying driver of the observed size effect: the heterogeneous responses across firms are in fact the consequence of large firms operating in multiple markets rather than their size per se. To interpret these findings, we build a heterogeneous firm model with single- and multi-unit firms subject to (i) unit-level real frictions, i.e., fixed and convex investment adjustment costs and (ii) firm-level financial frictions, i.e., costly equity issuance. In the model with unit-level frictions, an increase in uncertainty lowers the investment of both single and multi-unit firms through a `wait-and-see' effect. For a multi-unit firm, on the other hand, firm-level financial frictions generate the interdependence of investment across units within a firm, i.e., a fall in investment in one unit enlarges internal funds and so relaxes the constraint on the amount a firm can invest in the other unit. Therefore, upon uncertainty shocks, multi-unit firms lower their investment by less than single-unit firms. This is because the `wait-and-see' effect is partially offset by the relaxation of financial constraints due to the availability of larger internal funds when investment in one unit decreases. To examine the aggregate implications due to the heterogeneity in firms' responses, we compare the benchmark economy to a counterfactual economy with only single-unit firms. The result shows that the contribution of multi-unit firms is sizable in alleviating the impact of uncertainty shocks on aggregate investment. In the second chapter, with Korean firm-level and aggregated industry-level data, we uncover a balance sheet channel through which the exchange rate shock translates into domestic prices. Exploiting the quasi-natural experiment environment during the Asian Financial Crisis, we investigate how exposure to foreign currency debt prior to the crisis leads to different price dynamics. Our empirical finding suggests that when a sector had a higher level of short-term foreign currency debt ratio prior to the crisis, the price increase is more pronounced. Based on this empirical result, we build a heterogeneous firm model to study the transition path upon an unexpected real exchange rate shock, calibrated to match the real exchange rate changes in the 1996-98 period in Korea. In our model, a currency depreciation inflates the domestic value of foreign currency debt. As a consequence, firms, with a high share of their debt in foreign currency, face tighter working capital constraint and reduce their investment more, leading to higher costs of production and higher prices. The model is able to generate qualitatively consistent and quantitatively sizeable price increase upon a large depreciation of the currency, and furthermore, explain the cross-sectional variation in the sectoral price changes across industries. We also find that the interaction of strategic complementarity in firms' price settings and heterogeneity in foreign currency debt holdings across firms within an industry play an important role in amplifying the negative balance sheet effect on industry-level price dynamics.