Mathias Hoffmann

An interpretative survey of the literature on how international capital flows affect house prices, published in the Oxford Research Encyclopedia of Economics and Finance.

Based on historical bank branch-level lending and deposit data from 1920s’ Japan, we show how branch banking integrated peripheral markets with the rest of the country, with large urban banks – those headquartered in Tokyo and Osaka – using deposit supply shocks in peripheral areas to fund lending in their core markets. Faced with high-yielding lending opportunities in central prefectures, urban banks bid up deposit rates in peripheral areas, raising local banks’ funding costs and forcing them to become more effective in their screening. This competitive reallocation of capital through funding markets allowed banks to maintain a functional specialization in different customer segments and we argue that this mechanism helps explain the continued coexistence of small relationship lenders and large integrated arm’s-length lenders in local banking markets in many countries.

U.S. state-level banking deregulation during the 1980s considerably dampened the fallout on local economies of the China trade shock a decade later. The reason: households in financially integrated areas could more easily borrow against their housing wealth to smooth consumption. This kept house prices and wages in the non-tradable sector up, facilitating labor reallocation away from manufacturing.

After the inception of the euro, the real economy in most member countries remained dependent on credit by domestic banks, which increasingly funded themselves through cross-border interbank funding. We find that this pattern of ‘double-decker’ banking integration exposed domestic banks to sharp declines in cross-border interbank lending during the eurozone crisis. As a result, domestic banks reduced lending which led to large declines in output in sectors with many small (bank-dependent) firms. We propose a quantitative small open economy model to account for these patterns and conclude that a global banking shock leading to a sudden stop in cross-border interbank lending in the eurozone is required to account for them.

The global savings glut (i.e. capital flows from China and other emerging economies into the U.S.) contributed significantly to the rise and fall of U.S. house prices prior to and during the great financial crisis. For identification, we exploit that interstate banking deregulation during the 1980s - a decade prior to the savings glut - cast a long shadow: in states that opened their banking markets to out-of-state banks earlier, house prices were more sensitive to aggregate U.S. capital inflows during 1997-2012. We use bank-level data to identify the mechanism behind this stylized fact: capital inflows relaxed the value-at-risk constraints of geographically diversified (‘integrated’) U.S. banks more than those of local banks. Hence, mortgage credit expanded more in states where integrated banks had a relatively bigger market share (due to early deregulation).