The forces underlying the debt cycle of the 1920s were set in motion by World War I. Between 1914 and 1919, largely as a result of loans floated on behalf of the French and British governments, America's net debtor position was extinguished and replaced by a net creditor position of comparable magnitude. This surge of lending, which continued into the 1920s, reflected a combination of factors: continued rapid growth of the US economy, the wartime rise in saving, and the demand for capital to reconstruct the devastated European economies. Yet in the immediate aftermath of World War I, the international capital market remained becalmed. Changing rates of return played some role in stimulating US foreign lending: returns on foreign medium-grade bonds rose steadily (relative to yields on domestic medium-grade bonds) from the early 1920s until 1928. Rates of return alone, however, can explain little of the variation in the volume of lending abroad. The role of other factors, specifically risk, is especially evident before 1924, when US investors were unwilling to lend to foreigners at virtually any price.
The risks of lending were most evident in the case of Central Europe. So long as the size of their reparations obligations remained uncertain, it was not clear that the nations of this region possessed the resources to service additional external debt. The US perception that foreign lending was risky was not limited to Central Europe, however: at the beginning of the 1920s lending to Latin America was equally depressed. The dominant factor in this region was the depressed level of world trade and uncertain prospects for its recovery. The re-establishment mentof large-scale capital inflows required League of Nations intervention in the form of stabilization loans and substantial steps to reconstruct international trade and monetary institutions. If a lesson is to be drawn from the initial phases of this debt cycle, it is that an existing debt overhang and threats to an open trading system can dam the flow of resources to potential borrowers, and that outside intervention by governments or international institutions may be required to restart it.
In the 1920s as in the 1970s, financial innovation stimulated the surge in foreign lending. American investors acquired familiarity with the merits of foreign bonds through the Liberty Loan campaign of World War I. Banks enlisted in that campaign established or expanded their bond departments, while others established security affiliates to engage in the entire range of bond market activities. Once the Federal Reserve Act relaxed restrictions on the establishment of foreign branches, member banks began to move abroad. The spread of investment trusts enabled the small investor to participate in the market. Together, the rapid development of retailing and underwriting activities and the proliferation of investment vehicles provided organizations and individuals with both the incentive and the opportunity to increase their participation in foreign bond markets.
How did capital markets operate once foreign lending resumed? A standard criticism of the international capital market in the 1920s is that it failed to discriminate adequately among borrowers. The same criticism has been levelled at US lenders in the 1970s, and has provided the motivation for studies of the pricing of foreign bonds. These studies provide a benchmark for my analysis of the bond market in the 1920s. I analyse the determinants of the yield to maturity on a cross-section of some 200 categories of foreign bonds issued between 1920 and 1929. I find a positively sloped yield curve and a relatively high risk premium on foreign corporate bonds. While both results are consistent with standard models, they contrast with the findings of other investigators for the 1970s. I also find that the lowest risk premia were charged to Scandinavian countries, members of the British Commonwealth, small Western European countries, and small Central American republics economically or politically dependent on the United States. There is little evidence, however, that lenders took into account current policy in the borrowing countries, or that they charged higher premia for larger loans. It would seem that reputation more than current economic developments influenced bond market participants.
The analysis provides some evidence that lenders discriminated among potential borrowers on the basis of reputation and political factors which conveyed information about the probability of default, but it does not establish that they discriminated adequately. To address this issue, I specify a simple model of ex-ante and ex-post returns and estimate it using a sample of 50 foreign dollar bonds from the 1920s for which ex- post returns have been computed. The results suggest that, while investors incorporated differential default risk into the spreads they demanded of foreign borrowers, they did so incompletely and demanded inadequate premia to hold those bonds most subject to default risk.
If default risk was imperfectly perceived at time of issue, bondholders still could have recognized and acted upon it subsequently. I therefore examine the pricing of these same foreign bonds after 1931. These prices reveal that the suspension of debt service affected not only the prices of defaulting bonds, but also the prices of continuously serviced bonds, mainly those of neighbouring countries. This suggests that default created negative externalities by casting doubt on the credit-worthiness even of countries which maintained service on their external debts.
Approximately two-thirds of foreign securities held by American investors fell into default over the course of the Depression decade. Contemporaries believed that the experience had a lingering impact on the attitudes of American investors. One way to approach this issue is to compare US foreign lending in the ten years immediately succeeding World Wars I and II. The comparison reveals that US capital exports were actually larger after 1946 than after 1918. The difference is due almost entirely, however, to unilateral transfers by government, notably through the Marshall Plan: the real value of portfolio lending was more than 80% lower after 1945 than after 1918. This is precisely what one would expect had purchasers been deterred by the interwar defaults on foreign bonds.
This decline in portfolio lending could reflect either a general disenchantment with foreign loans or a special inability to borrow by countries with a recent history of default. An examination of the lending figures and a multivariate analysis of total external borrowing by governments between 1946 and 1955 provide no evidence that countries which defaulted in the 1930s found it more difficult to borrow in the 1940s and 1950s. The same result emerges from an analysis of private portfolio lending to Latin American countries between 1948 and 1955. If there existed a default penalty of the sort posited in many recent theoretical models, it did not take the form of a differential inability to borrow. Default tended to depress the volume of private portfolio lending generally, not to divert it from defaulting countries towards faithful debt servicers.
The evidence suggests that a substantial share of the cost of default was borne by debtors other than defaulting countries. Debtors may therefore find it advantageous to coordinate their decisions, whether or not that decision is to maintain service on their external debts.