Removing the barriers that keep capital inside national borders would raise welfare in the typical country by about 5.5% of labor income — and would concentrate economic activity among the world’s largest firms while pushing smaller ones to contract, according to a working paper released by the Federal Reserve Bank of Minneapolis.
The paper, by Loukas Karabarbounis of the University of Minnesota and the Minneapolis Fed, Bruno Pellegrino of Columbia Business School and Juliana Salomao of the University of Minnesota, builds a general-equilibrium model covering roughly 23,000 publicly listed firms across 48 countries, some six billion consumers and about $200 trillion in household wealth as of 2023. The views are the authors’ own and don’t necessarily reflect those of the Minneapolis Fed or the Federal Reserve System.
Its central empirical claim is that cross-border equity investment remains heavily constrained. Moving from complete financial autarky to today’s observed allocation of global capital produces only small welfare changes — a median gain of about 0.3% — because the frictions the model infers from actual portfolio holdings leave the world close to autarky already. The gains from removing what remains are an order of magnitude larger.
Those gains are unevenly distributed. Among large economies, the model puts the welfare improvement from full financial liberalization at roughly 3.1% of labor income in the U.S. and 10.5% in India, with emerging markets generally gaining more than advanced ones. Eliminating tariffs and non-tariff measures — leaving shipping costs in place — yields a median gain of about 3.8%, concentrated in mid-sized open economies rather than the largest ones. Canada is the outlier at roughly 22%, reflecting both high initial trade barriers and heavy reliance on imported consumption.
The distributional finding may prove more contentious than the aggregate one. In both liberalization scenarios, larger firms expand as financing costs or export costs fall, while smaller firms — which also face lower frictions — nonetheless contract under intensified competition. The share of national equity held by the top 1% of firms rises in every country under financial liberalization, by 23.7 percentage points in Russia, 18.5 in Canada and 17.8 in Korea.
That result depends on model features often left out of comparable work. When the authors switch off the matrix governing product substitution between firms, nearly all the losers disappear and almost every firm expands. When they switch off the correlation structure of equity returns, the welfare gains from financial liberalization all but vanish. Both are modeling choices rather than measured facts, and the paper’s conclusions about who wins and who loses rest on them.
The methodology combines a hedonic demand system on the goods side with mean-variance portfolio choice on the asset side — an approach the authors describe as “Markowitz-Sharpe meets Lancaster-Rosen” — chosen partly because it yields near-closed-form solutions at this scale. Product similarity between firms is inferred from textual analysis of business descriptions. Bilateral financial frictions are not observed but backed out from equity holdings, under the assumption that investors face no friction investing at home.
That assumption drives an unusual result: the model infers that foreign investors face a roughly 2-percentage-point return penalty on U.S. equities, because they hold less than the risk-return profile of American firms would imply. The authors flag it as an exception to otherwise intuitive patterns, alongside Russia, where the inferred friction exceeds 5 percentage points.
The paper’s broadest methodological argument is that studying trade and financial frictions separately understates both. Liberalizing one market reduces the gains from liberalizing the other, an interaction the authors find is far stronger in lower-income countries, where home bias in goods and assets reinforce one another. For those countries, the estimated gain from removing capital-market barriers is about 30% larger when trade frictions remain in place.
Several limits bound the results. The model is static, omitting the capital accumulation and growth channels that drive much larger estimates elsewhere in the literature — a related paper cited in the text finds median gains near 15% moving from full closure to the observed equilibrium. It also abstracts from collateral constraints and credit frictions that other work has found can make financial integration welfare-reducing. The estimated trade elasticity of about −4 falls at the low end of published estimates.