Tyler Denboer
First presented in 1980, Martin Feldstein and Charles Horioka present a discrepancy between economic theory and what is observed within the real world. By estimating the relationship between domestic savings and domestic investment, the two economists show what is now accepted as the ‘Feldstein-Horioka Puzzle’. Standard economic theory suggests that, in the absence of frictions in financial markets, capital will flow abroad to investment opportunities with the highest rate of return until equilibrium is achieved. However, the Feldstein-Horioka puzzle shows that this is not what occurs in the real world, and that there is high correlation between domestic savings and domestic investment.
Feldstein and Horioka estimate the following equation
I/GDP = α + β · (S/GDP)
Where I is the domestic investment, S is the domestic savings and GDP is the gross domestic product. β, the saving retention coefficient, measures the extent to which an increase in domestic savings is invested domestically instead of abroad i.e. a measure of the degree of an economy’s international capital mobility.
Feldstein and Horioka present the hypothesis that β = 0 if there is perfect capital mobility and β = 1 if there is no capital mobility. Since financial markets are increasingly globalised with less restrictions than in the past, β should be close to 0.
Standard macroeconomic theory would support this hypothesis, since capital flows work in such a way to move from low to high marginal products through investment. Therefore, in the absence of regulation, domestic savings will flow to the economies where investors can earn the highest return i.e. the most productive investment opportunities. So, there should be minimal correlation between a country’s domestic savings rate and domestic investment rate.
However, this is not the case.
Feldstein and Horioka estimated the regression equation for OECD countries from 1960 - 1974, and found that β ≈ 0.9 with statistical significance from 0 but not from 1. That is, there is a high correlation between a country’s domestic savings rate and domestic investment rate.
All papers attempting to reproduce this study yield the same result. For example, Obstfeld and Taylor (2004) estimate using 5 and 10 year averages from 1870 - 2000, and find that β ≈ 0.5 and is statistically different from 0.
This outcome presents the idea that increases in savings will mostly remain within the domestic economy, which is against the trends of globalisation and does not agree with standard macroeconomic theory.
Figure 1: Regression results as from Feldstein and Horioka’s Original Paper1
Many economists have attempted to justify this discrepancy between economic theory and reality. Some possible explanations include:
Barriers to trade: Frictions in financial markets such as transaction costs, capital controls, and exchange rates, as well as a need for government to intervene in trade to prevent consistent current account deficits are a common reasoning for the high correlation between domestic savings and investment. Since current account surplus results in a capital account deficit, government intervention in trade in order to control the current account will directly impact capital mobility. However, this reasoning does not have a large support as many economists believe that the relationship between domestic savings and investment should remain low, as in a globalised economy these barriers are minimal.
Asymmetric information: Agents in an economy may lack information about financial markets and trade that prevents them from investing abroad. For example, an individual may lack knowledge on how to invest their capital into another country and therefore are unable to do it, preventing them from earning an optimal return. Information and media that supports home bias may also lead agents to invest locally. Information regarding local investment opportunities is usually more accessible and easier to verify and hence differences in availability and quality of this information may impact investment decisions.
Fallacy of composition: A more modern solution to the problem is the idea of fallacy of composition. This idea suggests that, although individuals may experience perfect capital mobility when moving their own assets, the same is not true for a country as a whole, as each transfer may be negated by that of another agent abroad. The result of this is that financial markets by themselves cannot transfer capital between countries. For example, an agent in Australia is able to freely purchase U.S. government bonds and transfer part of their capital abroad. However, this does not necessarily mean that Australia as a country can transfer part of its capital to the U.S. Without the use of the tradeable goods market, the Australian agent can only transfer their capital to the U.S. if there is another U.S. agent who is willing to exchange their U.S. dollars for AUD to purchase Australian assets. Thus, their transactions ‘cancel out’ and both countries experience no net transfer of capital. Even though this idea of a fallacy of composition is gaining popularity (even by Feldstein and Horioka themselves), it is difficult to show this concept within a model that supports the data.
But why is this issue so important?
Most of economic theory assumes that, without any frictions in financial markets and risk premiums, interest rates equalize across borders so that r∗ is the world interest rate. It is at r∗ that agents can lend and borrow. However, the Feldstein-Horioka puzzle suggests that interest rates are not fully equalizing across borders due to reasons unknown. The result of this is that there is a possibility that standard economic theory is partially incorrect, and there is more to the literature than what is currently understood. The puzzle highlights the importance for accounting the ideas of friction, information asymmetry and fallacy of composition into economic models to accurately depict the real world. Without theory that reflects and supports the data, it becomes difficult for economists to make predictions and draw conclusions. This in turn has impact on government and policy makers, which has flow on consequences for society and economic welfare.
This issue is also important for economic policy. Gaining a better understanding of these frictions that prevent interest rates to equalize would allow a country to achieve a better allocation of both goods and capital through trade. Hence, solving the Feldstein-Horioka puzzle would be a step towards achieving an increasingly optimal allocation of resources to society and hence enhance economic welfare. By understanding the puzzle, government and policy makers can use accurate economic models to create predictions about the future and make economically sound decisions. As such, they can provide a more efficient and optimal allocation of resources within society.
It is clear that solving the Feldstein-Horioka puzzle would lead to better welfare in society through an optimal allocation of resources. Understanding the frictions and fallacies that lead to a high correlation between domestic savings and investment would allow economists to better understand individual decision-making regarding capital flows and financial markets. In turn, governments can create effective policy to allow for this optimal resource allocation. However, since the problem has been persistent for over 40 years, it is unclear how close we are to solving this alluring paradox.