Transcript Home bias and international risk sharing: Twin puzzles

Home bias and
international risk sharing:
Twin puzzles separated at
birth
Bent E. Sørensen, Yi-Tsung Wu, Oved Yosha, Yu Zhu
Presneted by Marek Hauzr, Jan Šarapatka, Daniel Vopat, Arian Zahermanesh
Outline
1.
Introduction
2.
Home bias
3.
International Risk Sharing
4.
Estimation
5.
Conclusion
Introduction

Home bias = only modest amounts of foreign debt and equity despite benefits
from international diversification

Full International risk sharing = identical consumption growth rates in all
countries

From mid of 90´s home bias ↓ and international risk sharing ↑

Is there empirical relation?

Panel-data regressions for OECD countries (1993 – 2003)


AF … stock of domestic assets owned by foreign residents
rF … rate of return on AF

AD and rD … domestically owned foreign assets and their return

CONS … total consumption

IF (rDAD – rFAF) and GDP not perf. correlated -> International Diversification ->
Smoother Income -> Smoother Consumption = Lower volatility
Home bias

Home bias = modest amounts of foreign assets despite benefits from
international diversification

Under standard assumptions CAPM predicts that countries hold identical
international portfolios of risky assets -> Home bias = deviation from this
prediction

In this paper: Debt home bias and Equity home bias
4 possible reason for home bias…?
1.
Hedging of currency risk
2.
Transaction costs
3.
Lack of information about foreign assets
4.
Costs of international trading

Ad 1) After EMU rapid decline in home bias in both EMU and non-EMU EU
counries -> hedging not a cause of home bias

Ad 2) Transaction cost not large enough to cause home bias

Ad 3)+4) cost of international trading ↓ and information
availability ↑ during 90´s -> main suspects for causing home bias
Measuring home bias

Home bias is from [0;1]

Home bias = 1 for country with solely domestical investments

Home bias = 0 if the share of country’s domestical investments equals the
share of country’s equity market in the total world equity market
International risk sharing

two alternative measures

based on consumption data (taste shocks)

based on GNI

Full/perfect consumption risk sharing = identical consumption growth rates
in all countries

Full/perfect income risk sharing = identical GNI growth rates in all countries
Individual-level data vs. country level
data

regress individual consumption on income or regress country consumption on
world consumption

Previous research -> no perfect risk sharing

This paper uses country level data (24 countries OECD)
Testing perfect income risk sharing

Estimation:

Perfect risk sharing -> left side is zero (from the definition) -> B is zero

The higher the beta the lower the income risk sharing. Sometimes used 1beta for more comprehensive results.
Percent of income risk shared over years
1993 – 2003
Percent of consumption risk shared over
years 1993 – 2003
Panel-data regressions: specification
• Mélitz and Zumer (1999) impose structure on k = K0 + K1gi „where gi is
an interaction that affects the amount of risk sharing that country i
obtains.”
• Average amount of income risk sharing by country i is 1 - K0 - K1gi1
• k was allowed to change over time
where EHB is equity home bias for country i in time
t and EHBt is the average across countries at time t
• 1 – k measures amount of risk sharing obtained in period t by
country i with equity home bias with time trend taken into account
• -K1 captures yearly increase in income risk sharing
• -K2 „measures how much higher than average EHB lowers the
amount of income risk sharing“
• Similarly debt can be used instead of equity.

They also performed similar analysis using foreign asset holdings relative to GDP.

They considered the ratio of FDI to GDP.

Liabilities can be considered instead of assets.

Where Eit = foreign equity holdingsit / GDPit ; and K2 is similar as on previous slide

Where EHB changes into EB,
which is EBit = log((foreign equity + debt holdings)it /GDPit )


They also estimated the contribution of EHB to consumption risk sharing
where
Results: Home Bias for the OECD and EU.
Panel Regression: Results

Coefficient of equity home bias and equity home bias are significant.

the coefficient to imply that declining home bias has been associated with
increasing consumption risk sharing even if the actual value is likely to not be
valid for very large changes in home bias.

There is no association between declining Debt Home Bias and consumption
risk sharing. However lack of adequate date makes it impossible to estimate
both simultaneously.
Correlation matrix
Results: Correlation matrix (continued)

GNI growth and GDP growth are highly correlated.

Income risk sharing is low and consumption risk sharing is imperfect.

Equity liabilities and FDI liabilities are highly correlated while debt liabilities
are slightly less correlated.
Results: foreign assets

Ratio of foreign asset holdings to GDP predicts income and consumption risk
sharing in the OECD.

There is a positive effect of higher foreign equity Debt Home Bias. Our
interpretation is that there is not enough data to estimate the impact of both
simultaneously.

Income risk sharing in the EU is not significantly different from zero but there
is a positive significant trend.

In the EU, income risk sharing is not significantly related to either stock or
debt home bias.

Consumption risk sharing among EU countries is lower than among the OECD
countries and neither trend nor home bias indices are significant.
Results: income risk sharing

For income risk sharing, is seen larger and more significant coefficients for debt
and FDI and, in particular, for the sum of equity and debt, and for the sum of all
three components.

All liability components are insignificant for consumption risk sharing except for
FDI.

Table 7 reports the results of multiple regressions for income risk sharing. The first
row includes interactions for all three asset categories: the point estimates for
equity and debt are of similar magnitude; equity is clearly significant and debt is
nearly significant at the 5% level while FDI has a negative significant coefficient.
Given that the re-gressors are highly correlated and that FDI has a positive sign in
a univariate regression they tend to believe that FDI is not detrimental to risk
sharing but only a larger data set can determine this with certainty. For liabilities,
see the second row, only debt holdings are nearly significant at the 10% level;
however, all variables have a positive sign consistent with Table 6.
Results: Risk sharing and foreign liability

For income risk sharing, there is a larger and more significant coefficients for
debt and FDI and, in particular, for the sum of equity and debt, and for the
sum of all three components.

All liability components are insignificant for consumption risk sharing except
for FDI.

The results of multiple regressions for income risk sharing. The regressors are
highly correlated and that FDI has a positive sign in a univariate regression
authors tend to believe that FDI is not detrimental to risk sharing but only a
larger data set can determine this with certainty.

Liabilities are only debt holdings are nearly significant at the 10% level;
however, all variables have a positive sign consistent.
Results: Risk sharing and foreign liability

Equity and debt liabilities have much smaller coefficients when they are
estimated together with assets, which probably reflects that assets are more
effective in providing risk sharing.
Results

These estimates imply that a country with large but identical amounts of
foreign equity, debt, and FDI assets and liabilities is going to achieve negative
consumption risk sharing which seems to contradict our other results.

Reasons:

Authors suspect that the large coefficients to assets together with large negative
coefficients to liabilities reflect multicollinearity in conjunction with noisy
consumption data.

Considering income and consumption risk sharing together it seems that equity
assets and FDI liabilities are conducive for risk sharing, but the empirical evidence
for such a breakdown is not strong, in particular for consumption risk sharing.

It is difficult to identify which components of international capital flows are
more beneficial for risk sharing in particular, for consumption risk sharing.
Risk sharing increases with equity
holdings within the EU

The results are similar for debt or debt plus equity holdings likely both are
important. However, there is no relation between higher FDI and income risk
sharing.

For consumption risk sharing were not found significant coefficients although
the coefficients are robustly positive and of a similar order of magnitude as
those found for the OECD sample.
Panel-data regressions: robustness

Results are not very sensitive to the inclusion of country-fixed effects.

Estimate the impact of (equity plus debt) and FDI on income and consumption
risk sharing, respectively, dropping one country at a time.

These results show that the impact of equity plus debt assets on income risk
sharing is highly robust.

The impact of FDI assets on income risk sharing is robust although the tstatistic drops to 1.83 when United States were left out.

For consumption risk sharing, the impact of equity plus debt assets is not
totally robust (t-statistic of 1.32 when Japan is left out) but the impact of FDI
assets on consumption smoothing is very robust with all t-statistics above 3.
Summary

Foreign portfolio assets is positively and robustly related to income risk
sharing.

Consumption risk sharing international asset holdings are positively related to
risk sharing.

Equity and FDI assets are more important than debt.

No detectable role for liabilities in regressions where assets are included
except that FDI appears to benefit consumption risk sharing.

Various forms of ‘‘taste’’ shocks to consumption make it harder to robustly
detect patterns of consumption risk sharing compared to those of income risk
sharing.