Foreign Direct Investment and Foreign Portfolio Investment

Foreign Direct Investment and Foreign Portfolio Investment

Systemic Liquidity and the Composition of Foreign Investment: Theory and Empirical Evidence Theory and Empirics by Itay Goldstein, Assaf Razin, and Hui Tong February 2007 The key prediction of the model is that countries that have a high probability of an aggregate liquidity crisis will be the source of more FPI and less FDI. The intuition is that as the probability of an aggregate liquidity shock increases, agents know that they are more likely to need to sell the investment early, in which case, if they hold FDI, they will get a low price since buyers do not know whether they sell because of an individual liquidity need or because of adverse information on the productivity of the investment. As a result, the .attractiveness of FDI decreases, and the ratio of FPI to FDI increases The Efficiency Advantage Imagine a large company that has many relatively small shareholders.Then, each shareholder faces the following well-known free-rider problem:if the shareholder does something to improve the quality of management, then the benefits will be enjoyed by all shareholders. Unless the shareholder is altruistic, she will ignore this beneficial effect on other shareholders and so will under-invest in the activity of monitoring or improving management. Oliver Hart. The Disadvantage: A Premature Liquidation However, when investors want to sell their investment , prematurely, because of a liquidity shock they will get lower price if they are conceived .by the buyer to have more information Because, other investors know That the seller has information on the Fundamentals and suspect That the sales result from bad prospects of the project .Rather than liquidity shortage Liquidity Shocks and Resale Values

Three periods: 0, 1, 2; Project is initially sold in .Period 0 and matures in Period 2 R F ( K )(1 ) Production function cdf G ( ), G ( 1) 0, G (1) 1, g ( ) G ' ( ) R K (1 ) 1 AK 2 2 Distribution Function :Production Function Special Form In Period 1, after the realization of , the productivity shock The manager observes the .productivity parameter Thus, if the owner owns the asset as a Direct Investor, the chosen level :of K is 1 K * ( ) A Expected Return (1 )(1 ) 1 1 2 E (1 ) 2 E A A 2 A 2A

In Period 1, after the realization of , the productivity shock The manager observes the .productivity parameter Thus, if the owner owns the asset as a Direct Investor, the chosen level :of K is 1 K * ( ) A Expected Return (1 )(1 ) 1 1 2 E (1 ) 2 E A A 2 A 2A Liquidity Shocks and Resale Values Three periods: 0, 1, 2; Project is initially sold in .Period 0 and matures in Period 2 R F ( K )(1 ) Production function cdf G ( ), G ( 1) 0, G (1) 1, g ( ) G ' ( ) R K (1 )

1 AK 2 2 Distribution Function :Production Function Special Form Portfolio Investor will instruct the manager to maximize the expected return, absent any information on the productivity parameter. 1 K Expected return A (1 ) 1 E (1 2 ) E 2A 2A A Liquidity Shocks and Re-sales Period-1Price is equal to the expected value .of the asset from the buyers viewpoint Productivity level under threshold D which the direct owner probabilit y (1 )G ( D ) Is selling with no liquidity shock 1 2 D D (1 )

1 2 (1 ) g ( )d g ( )d 2A 2A 1 1 (1 )G ( D ) D P1, D P1, D (1 D ) 2 2A The owner sets the threshold so that she Is indifferent between the price paid by buyer And the return when continuing to hold the asset If a Portfolio Investor sells the asset, everybody knows that it does so only because of the liquidity :shock. Hence 1 1 2 1 P1, P g ( )d 2A 2A 1 Since D 0 P1, D 1 P1, P 2A

Trade-off between Direct Investment and Portfolio Investment Direct Investment P1, D (1 D ) 2 2A .Return when observing liquidity shock :If investor does not observe liquidity shock D 1 (1 D ) 2 (1 ) 2 return g ( )d g ( )d 2A 2A 1 D :Ex-Ante expected return on direct investment 1 2 D (1 ) 2 (1 D ) 2 ( 1 ) D VD (1 ) g ( )d g ( ) d 1 2A

2A 2 A D Portfolio Investment :When a liquidity shock is observed, return is P1, P 1 2A :When liquidity shock is not observed return is E (1 2 ) 1 2A 2A :Ex-ante expected return is VP 1 2A Dif ( ) VD VP C Firms sold to Direct Investor Dif ( ) VD VP C Firms sold to Portfolio Investor 1 Direct Investment Portfolio investment (C )

Dif(0) 0 Probability of midstream sales Direct Investment :Resale probability (1 )G D ( ) Portfolio Investment :Resale probability Only in a few cases, the probability Of an early sale in an industry with Direct investment is higher than for .An industry owned by portfolio investors Heterogeneous Investors Different investors face a price which Does not reflect their true liquidity-needs. This may generate An incentive to signal the true parameter .By choosing a specific investment vehicle . Suppose there is a continuum [0,1] of investors Proportion of them have high expected liquidity needs,H , and proportion , . have low expected liquidity needs L 1 H L 2 rational expectations equilibrium Assuming that rational expectations hold in the market, D has to be consistent with the equilibrium choice of investors between

FDI and FPI. thus, it is given by the H H , FDI L L , FDI following equation: D H , FDI L , FDI :There are 4 potential equilibria .All investors who acquire the firms are Direct Investors. 1 .All investors who acquire the firms are Portfolio Investors . 2 investors who acquire the firms are Direct Investors, and. H 3 L . investors who acquire the firms are Portfolio Investors investors who acquire the firms are Direct Investors, and. 4 H L . investors who acquire the firms are Portfolio Investors All firms are acquired by Direct Investors When investors resell, potential buyers assess a probability of that the investor is selling because of liquidity needs, and a Probability of that she is selling because she observed low productivity. Expected profits, ex-ante, for direct investors exceed expected profits for portfolio investors, for both high :liquidity and low liquidity investors High-Liquidity needs:Investors 1 1 2 1 2 1 D ( ) (1 D ( )) 1 (

1 ) 2 2 (1 H ) d 2 2A 2 1 2 A D( ) 2 1 (1 D ( )) 2 2 H 2A 1 1 2 1 1 2 1 P ( ) (1 P ( )) 1 ( 1 ) ( 1 2

) 2 2 (1 H ) d ( 1 ) d 1 2 A 2 2A 2 1 2 A P( ) 2 1 (1 P ( )) 2 2 H 2A :Low-Liquidity-needs Investors 1 1 2 1 D ( ) (1 D ( )) 1 1 (1 ) 2 2 2 (1 L ) d 2 2A

2 1 2A D( ) 2 1 (1 D ( )) 2 2 L 2A 1 1 2 1 1 P ( ) (1 P ( )) 1 1 (1 ) 2 (1 2 ) 2 2 (1 L ) d (1 ) d 2 2A 2 1 2A 2A 1 P( ) 2 1 (1 P ( )) 2 2

L 2A !The two conditions hold for some parameter values Interpretation The idea that we are trying to capture with this specification is that individual investors are forced to sell their investments early at times when there are aggregate liquidity problems. In those times, some individual investors have deeper pockets than others, and thus are less exposed to the liquidity issues. Thus, once an aggregate liquidity shock occurs, investors, who have deeper L pockets, are less likely to need to sell than investors. H Interpretation The reason for the existence of the pooled, only-FDI investment equilibrium is the strategic externalities . between high-liquidity-need Investors An investor of this type benefits from having more , investors of her type When attempting to resell price does not move against her that much, because the market knows with high probability that .the resale is due to liquidity needs When all high-liquidity need investors acquire the firms, a single investorof this type knows that when resale contingency arises, price will be low, and she will choose to become a direct investor, self validating the behavior of investors of this type in the equilibrium. The low-liquidity-need Investors .Care less about the resale contingency As we can see in the figure, the

equilibrium patterns of investment are determined by the parameters A andH. Since H L 1 , the value of H also determines L and thus can be interpreted as a measure for the difference in liquidity needs between the two types of investors. In the figure we can see that there are four thresholds that are important for the characterization of the equilibrium outcomes. Aggregate Liquidity Shocks There are two states of the world. In one state (which occurs with probability q) there is an aggregate shock that generates liquidity needs as described before. That is, in this state of the world a proportion of one type of investors have to liquidate their investment projects prematurely and a proportion of the other type have to do so as well. In the other state of the world (which occurs with probability 1-q) there is no aggregate shock that generates liquidity needs and no foreign investor has to liquidate her investment project prematurely. probability of an aggregate liquidity shock The intuition is that as the probability of an aggregate liquidity shock increases, agents know that they are more likely to need to sell the investment early, in which case they will get a low price since buyers do not know whether they sell because of an individual liquidity need or because of adverse information on the productivity of the investment. As a result, the attractiveness of FDI decreases. first empirical prediction Countries with a higher probability of liquidity shocks will be source of a higher ratio of FPI to FDI.

The Role of Opacity The effect of liquidity shocks on the composition of foreign investment between FDI and FPI is driven by lack of transparency about the fundamentals of the direct investment. If the fundamentals of each direct investment were publicly known, then liquidity shocks would not be that costly for direct investors, as the investors would be able to sell the investment at fair price without bearing the consequences of the lemmons problem. Suppose that the source country imposes disclosure rules on its investors that ensure the truthful revelation of investment fundamentals to the public. In such a case, FDI investors will have to reveal the realization of once it becomes known to them. Then, since potential buyers know the true value of the investment, direct investors will be able to sell their investment at (((1+))/(2A)). Thus, whether or not a direct investor sells the ))/(2A)). Thus, whether or not a direct investor sells the investment, he is able to extract the value (((1+))/(2A)). Thus, whether or not a direct investor sells the ))/(2A)), and so the expected value from investing in FDI is ((E((1+))/(2A)). Thus, whether or not a direct investor sells the )))/(2A))-C. The expected value from investing in FPI is (1/(2A)) as before. This is because the kind of disclosure requirements we describe here do not affect the value of portfolio investments. These are requirements that are imposed by the source country, and thus apply only for investments that are being controlled by source-country Analyzing the trade off between FDI and FPI under this perfect source-country transparency, we can see two things. First, with transparency, FDI becomes more attractive than before. Second, with transparency, the decision between FDI and FPI ceases to be a function of the probability of a liquidity shock. second empirical prediction The effect of the probability of a liquidity shock on the ratio of FPI and FDI increases in the level of opacity in the source country. Ratio of FPI and FDI Probit Dynamic Version Transparency Data

The theory is geared toward explaining the allocation of the shock of foreign capital between portfolio and direct foreign investors. Now we confront this hypothesis with the data. The latter consist of stocks of FPI and FDI in market value, that are compiled by Lane and MilesiFerretti (2006).See Summary Statistics. Table 1. Summary Statistics of FPI/FDI Table 1 presents the average of the log of FPI stock over FDI stock for 140 source countries for the period from 1990 to 2004. Obs is the number of non-missing observations for each source country. Countries with no observations at all during this period are not reported. Source: Lane and Milesi-Ferretti (2006). Country Name Obs Mean Country Name Obs Mean United States United Kingdom Austria Belgium Denmark France Germany Italy Luxembourg Netherlands Norway Sweden Switzerland Canada Japan Finland Greece Iceland Ireland Malta Portugal

Spain Turkey Australia New Zealand South Africa Argentina Brazil Chile Colombia Costa Rica Dominican Republic El Salvador Mexico Paraguay Peru Uruguay Venezuela, Rep. Bol. Trinidad and Tobago Bahrain Cyprus Israel Jordan Lebanon Saudi Arabia United Arab Emirates Egypt Bangladesh 15 15 15 15 15 15 15 15 5 15 15 15 15 15 15 15 15 14

15 11 15 15 14 15 15 15 15 15 15 15 10 9 4 15 15 15 15 15 10 15 6 15 8 4 13 15 8 5 -0.56 -0.14 -0.32 -0.37 -0.69 -1.57 -0.28 -0.40 -0.22 -0.58 -0.88 -1.11 -0.10 0.05 -0.52

-2.27 -0.62 -0.24 1.02 -1.39 -0.50 -1.26 0.43 -0.64 -0.72 -0.66 0.16 -2.91 -0.22 -0.91 -1.04 -0.54 0.58 -0.40 -3.11 0.73 -0.22 -1.12 -2.32 0.60 0.04 -0.27 1.79 -0.06 -0.89 5.66 -0.16 -3.17 Cambodia Taiwan Province of China Hong Kong S.A.R. of China India Indonesia Korea Malaysia Pakistan Philippines Singapore Thailand Algeria

Botswana Congo, Republic of Benin Gabon Cte d'Ivoire Kenya Libya Mali Mauritius Niger Rwanda Senegal Namibia Swaziland Togo Tunisia Burkina Faso Armenia Belarus Kazakhstan Bulgaria Moldova Russia China,P.R.: Mainland Ukraine Czech Republic Slovak Republic Estonia Latvia Hungary Lithuania Croatia Slovenia Macedonia Poland Romania 8 15 15 15 4 15 15 3 15

15 14 14 11 10 9 7 14 15 15 8 6 8 6 15 14 13 13 15 5 8 8 6 8 11 13 15 9 12 12 11 11 14 12 8 11 7 7 7 -0.09 -1.14 -1.37 -0.67 -4.51 -2.18

-2.27 -2.51 -0.17 0.05 -3.66 -7.45 -0.16 0.30 -3.63 -2.98 -1.07 -3.48 3.04 -3.66 -1.38 -5.38 -0.33 -1.27 0.65 -3.94 -1.95 2.08 -2.04 -1.58 -1.13 -0.28 -0.52 -3.99 -4.70 -2.94 -0.37 0.33 1.22 -2.00 -1.20 -1.88 -1.47 -3.11 -2.79 2.01 -1.97 -2.86 Algeria Argentina Bahrain

Belarus Benin Brazil Bulgaria Chile Colombia Costa Rica Croatia Cte d'Ivoire Denmark Dominican Republic Egypt Greece Hong Kong S.A.R. of China Hungary Iceland India Indonesia Israel Japan Kazakhstan Kenya Latvia Lebanon Libya Lithuania Macedonia Malaysia Malta Mauritius Mexico Moldova New Zealand Niger Pakistan Paraguay Peru Philippines Poland Romania Rwanda Saudi Arabia Senegal Slovak Republic Spain Swaziland

Thailand Togo Turkey Ukraine Uruguay Venezuela, Rep. Bol. 1987,1986, 2001,1989,1987,1986, 2002,2001,1995,1993,1991,1990,1987, 2003,1998,1997, 2004,2003, 1999,1997,1986, 1996, 1993,1987,1986, 2002,1998,1995, 2002,1998, 1998, 2003,1993,1992,1989,1988,1986, 1994, 2000,1996, 1999,1998, 2001,2000,1997,1995,1992,1989, 2001,1998, 1994, 1994, 1995,1990,1989,1988,1987,1986, 2001, 1988,1987, 1999, 1998, 1997,1996,1995,1994,1990,1987, 1995, 2004,2003,2002, 1993,1991,1988,1987, 1999, 2002, 1996,1995,1994, 2001,1994, 1998, 2002,2000,1994,1992,1988, 1998, 1997,1992,1991,1988, 2002,1998,1997,1996, 2004, 2002,2001,1997,1992,1988,1987,1986,

2000,1999,1998,1990,1987,1986, 2001,2000,1997,1990,1987, 1996, 1999,1998,1995, 2003, 1998,1996,1995,1994,1993,1992, 1993,1990,1987,1986, 1999,1998, 1994, 2003, 1997, 2001,1998,1993,1992,1987,1986, 2001,1994, 1998, 2002, 1995,1992,1988,1987,1986, Probit Table 3: Probit Estimation of Aggregate Liquidity Crises Table 3 estimates the probability of liquidity crises for 140 countries over the period 1985-2004. The dependent variable is the liquidation of source countrys foreign asset. Sovereign rating is from Standard and Poors, while all other variables are from the WDI. A pooled Probit regression is estimated. * indicates significance at 5%. Population (log) GDP per capita (log) U.S. real interest rate Sovereign rating Constant R-square Observations Coef. -0.10* -0.18 0.11* -0.18* 2.03 0.13 776 Std. Err. 0.05 0.10 0.04

0.06 1.24 Ratio of FPI and FDI Table 4: Determinants of the Ratio of FPI over FDI The dependent variable is the log of FPI stock over FDI stock, for 140 source countries over the period from 1985 to 2004. The estimated probability of liquidity crisis is based on the estimates from Table 3. All other explanatory variables are from the WDI. Case 1 is the panel estimation with country and year fixed effects. Case 2 adds a one-year-lagged dependent variable as an explanatory variable, and estimates a dynamic panel model. * indicates significance at 5%. Case 1 Log of FPI/FDI (one lag) Population (log) GDP per capita (log) Stock market capitalization Trade openness (log) Probability of liquidity crisis Observations Coef St. err. -3.02* 0.46 0.36* -0.31 4.67* 697 0.76 0.34 0.06 0.25 1.23 Case 2 Coef St. err. 0.47* 0.04 -2.73* 1.22 0.65 0.37

0.22* 0.06 -0.70* 0.23 3.93* 1.19 603 Levels of FPI and FDI Table 5: Determinants of FPI or FDI The dependent variable is the log of FDI stock in Case 1, and the log of FDI stock in Case 2 for 140 source countries over the period from 1985 to 2004. The estimated probability of liquidity crisis is based on the estimates from Table 3. All other explanatory variables are from the WDI. A dynamic panel model with country and time fixed effects is estimated. * indicates significance at 5%. Log of FPI(one lag) Log of FDI(one lag) Population (log) GDP per capita (log) Stock market capitalization Trade openness (log) Probability of liquidity crisis Observations Case 1 (FPI) Coef St. err. 0.32* 0.04 -0.71 0.98* 0.18* -0.20 -1.53* 613 0.84 0.26 0.04 0.17 0.74 Case 2 (FDI) Coef St. err.

0.42* 0.67 0.96* 0.03 0.36* -2.33* 0.04 0.67 0.25 0.04 0.16 0.67 Country Acc Opa Country Acc opa Country Acc opa Finland 17 13 Argentina 30 44 Taiwan

40 34 Belgium 17 23 India 30 48 Brazil 40 40 Germany 17 25 Venezuela 30 51 Poland 40 41 USA 20 21

UK 33 19 Russia 40 46 Canada 20 23 Denmark 33 19 Egypt 40 48 Chile 20 29 Hong Kong 33 20 Czech Rep 44 41

Israel 20 30 Australia 33 21 Turkey 44 43 Thailand 20 35 Austria 33 23 Lebanon 44 59 Japan 22 28 S. Africa 33

34 Singapore 50 24 Indonesia 22 59 France 33 37 Spain 50 34 Sweden 25 19 Mexico 33 44 Portugal 50 35 Switzerland

25 23 Pakistan 33 45 Hungary 50 36 Ecuador 25 42 Saudi Arabia 33 46 Greece 50 41 Colombia 29 43 Philippines 33 50

China 56 50 Malaysia 30 35 Netherlands 38 24 Italy 63 43 Korea 30 37 Ireland 38 26 Opacity Index Effect of Transparency on Ratio of FPI and FDI Table 7: Determinants of the Ratio of FPI over FDI (The Effect of Opacity) The dependent variable is the log of FPI stock over FDI stock, for 140 source countries over the period from 1985 to 2004. The estimated probability of liquidity crisis is based on the estimates from Table 3. All other explanatory variables are from the WDI. Case 1 is the panel estimation with country and year fixed effects. Case 2 adds a one-year-lagged dependent variable as an explanatory

variable, and estimates a dynamic panel model. * indicates significance at 5%. Case 1 Log of FPI/FDI (one lag) Population (log) GDP per capita (log) Stock market capitalization Trade openness (log) Probability of liquidity crisis Probability of crisis *Opacity Observations Coef St. err. -4.72* 0.69* 0.27* 0.29 -16.5* 0.60* 510 0.71 0.32 0.07 0.23 3.92 0.11 Case 2 Coef St. err. 0.80* 0.04 -2.37* 0.78 -0.12 0.24 0.004 0.04 -0.22 0.16 -8.66* 2.48

0.24* 0.07 459 Table 8: Episodes of Interest Rate Hike Table 8 reports the number of liquidity crises for 140 countries over the period from 1985 to 2004. The crisis is defined as a real interest rate rise of more than 4% a year. Source: World Development Indicators Country Albania Algeria Angola Argentina Armenia Australia Austria Azerbaijan Bahrain Bangladesh Belarus Belgium Benin Bolivia Bosnia and Herzegovina Botswana Brazil Brunei Darussalam Bulgaria Burkina Faso Cambodia Cameroon Canada Chad Chile China,P.R.: Mainland Colombia Congo, Dem. Rep. of Congo, Republic of Costa Rica Cte d'Ivoire Croatia Cyprus Czech Republic Denmark Dominican Republic Ecuador Egypt

El Salvador Equatorial Guinea Estonia Ethiopia Euro Area Fiji Finland France Gabon Georgia Freq 3 3 5 2 2 1 0 1 5 4 5 0 4 6 1 7 1 0 4 5 3 5 0 11 7 5 4 5 9 6 4 3 1 2

0 4 12 6 2 6 4 7 0 8 1 0 10 2 Country Germany Ghana Greece Guatemala Guinea Haiti Honduras Hong Kong S.A.R. of China Hungary Iceland India Indonesia Iran, Islamic Republic of Ireland Israel Italy Jamaica Japan Jordan Kazakhstan Kenya Korea Kuwait Kyrgyz Republic Lao People's Dem.Rep Latvia Lebanon Libya Lithuania

Luxembourg Macedonia Madagascar Malawi Malaysia Mali Malta Mauritius Mexico Moldova Morocco Mozambique Myanmar Namibia Nepal Netherlands New Zealand Nicaragua Niger Freq 0 5 5 5 4 2 3 3 2 4 2 2 0 2 5 2 7 1 3 0 5 2 9 3 4

0 3 0 4 0 2 3 11 2 1 4 1 2 5 2 1 0 3 3 1 1 4 6 Country Nigeria Norway Oman Pakistan Panama Papua New Guinea Paraguay Peru Philippines Poland Portugal Qatar Romania Russia Rwanda Saudi Arabia Senegal Singapore Slovak Republic Slovenia South Africa

Spain Sri Lanka Sudan Swaziland Sweden Switzerland Syrian Arab Republic Tajikistan Tanzania Thailand Togo Trinidad and Tobago Tunisia Turkey Turkmenistan Uganda Ukraine United Arab Emirates United Kingdom United States Uruguay Uzbekistan Venezuela, Rep. Bol. Vietnam Yemen, Republic of Zambia Zimbabwe Freq 15 3 7 0 2 8 6 3 4 1 3 0 0 3 3 0 1

3 2 3 4 2 4 0 10 2 0 7 2 1 2 4 8 2 0 0 8 6 3 2 0 9 0 8 0 3 12 9 Probit Table 9: Probit Estimation of Aggregate Liquidity Crises Table 9 estimates the probability of liquidity crises for 140 countries over the period 1985-2004. The dependent variable is the dummy indicator of liquidity crises defined as a real interest rate rise of more than 4% a year. Sovereign rating is from Standard and Poors, while all other variables are from the WDI. A pooled Probit regression is estimated. * indicates significance at 5%. Population (log) GDP per capita (log) U.S. real interest rate Sovereign rating

Constant R-square Observations Coef. -0.07 -0.06 0.19* -0.19* 0.17 0.11 689 Std. Err. 0.05 0.10 0.05 0.06 1.24 Ratio of FPI and FDI Table 10: Determinants of the Ratio of FPI over FDI The dependent variable is the log of FPI stock over FDI stock, for 140 source countries over the period from 1985 to 2004. The estimated probability of liquidity crisis is based on the estimates from Table 9. All other explanatory variables are from the WDI. Case 1 is the panel estimation with country and year fixed effects. Case 2 adds a one-year-lagged dependent variable as an explanatory variable, and estimates a dynamic panel model. * indicates significance at 5%. Case 1 Log of FPI/FDI (one lag) Population (log) GDP per capita (log) Stock market capitalization Trade openness (log) Probability of liquidity crisis Observations Coef St. err. -2.94* 0.29 0.36* -0.27 3.38*

697 0.76 0.34 0.07 0.25 0.88 Case 2 Coef St. err. 0.47 0.04 -2.61* 1.22 0.52 0.37 0.23* 0.06 -0.68* 0.23 3.06* 0.87 603 Interpretation The reason for the existence of the only-direct investment equilibrium is the strategic externalities . between high-liquidity-need Investors An investor of this type benefits from having more , investors of her type When attempting to resell price does not move against her that much, because the market knows with high probability that .the resale is due to liquidity needs When all high-liquidity need investors acquire the firms, a single investorof this type knows that when resale contingency arises, price will be low, and she will choose to become a direct investor, self validating the behavior of investors of this type in the equilibrium. The low-liquidity-need Investors .Care less about the resale contingency Figure 2.1: The Allocation of investors between FDI and FPI Aggregate Liquidity Shocks

Suppose now that an aggregate liquidity shock occurs in period 1 with probability q. Once it occurs, it becomes common knowledge. Conditional on the realization of the aggregate liquidity shock, individual investors may be subject to a need to sell their investment at period 1 with probabilities as in the previous section. Conditional on the realization of an aggregate liquidity shock, the realizations of individual liquidity needs are independent of each other. If an aggregate liquidity shock does not occur, then it is known that no investor needs to sell in period 1 due to liquidity needs. This implies that the only reason to sell at that time is adverse information on the profitability of the project. As a result, the market breaks down due to the wellknown lemons problem (see Akerlof (1970)). On the other hand, if a liquidity shock does happen, the expected payoffs from FDI and FPI are exactly the same as in case of idio-syncratic shocks section. Aggregate and Idiosyncratic Shocks The model discussed in the preceding section assumed effectively that q = 1. We now extend the model to allow q to be anywhere between one and zero, inclusive. Figure 2.1 was drawn for the case q = 1. When q is below 1, the lines and shift upward; see Goldstein, Razin and Tong (2007). As expected, there is less FPI in each equilibrium and the number of configurations in which there is no FPI rises. In the extreme case where q = 0, no foreign investor will choose to make FPI, because there is no longer any liquidity cost associated with FDI, and there remains only the efficiency advantage of the latter . With the predicted probability of liquidity shocks, we can now estimate the regression equation. The results are presented in Table 3.3. Column (b) differs from column (a) in that it does not include the market capitalization variable, as the latter is not available in all of our observations. As our theory predicts, indeed a higher probability of an aggregate liquidity shock (the parameter q of the preceding chapter) increases the share of FPI,

relative to FDI. The interaction term between the probability of an aggregate liquidity shock and GDP per capita is significant. This is indicative for a nonlinear effect of the aggregate liquidity shock and/or the GDP per capita on the ratio of FPI to FDI. liquidity crisis We define the liquidity crisis as episodes of negative purchase of external assets. The flow data on external assets is from the International Financial Statistics's Balance of Payments, where assets include foreign direct investment, foreign portfolio investment, other investments and foreign reserves. We thus define the liquidity crisis episodes as sales of external assets, which has a frequency of 13% in our sample of 140 countries from 1985 to 2004. Regression log(FPI / FDI ) X i ,t Pr obi ,t 1 Log (GDPpercapita) Pr obi ,t 1 i ,t The crux of our theory is that a higher probability of an aggregate liquidity shock (the variable q of the preceding chapter) increases the share of FPI, relative to FDI. Therefore we include in the regression a variable, Pi,t+))/(2A)). Thus, whether or not a direct investor sells the 1, to proxy this probability in period t+))/(2A)). Thus, whether or not a direct investor sells the 1, as perceived in period t. We measure this probability by the probability of a 10% or more hike in the real interest rate in the next period. We emphasize that we look at the probability of such a hike to occur irrespective of whether such a hike actually occurred. We also include country and time fixed effect variables. Probit To estimate the probability of a 10% or more hike of the real interest rate, we apply the following Probit model, similar to Razin and Rubinstein (2006). 1 yt1 0 0 yt1 0 I ( AggregateLiquidityShocki ,t 1 ) yt1 Z i ,t i ,t 1

Table 1: Summary Statistics of ln(FPI/FDI) from 1990 2004 Country Name Obs Mean Country Name Obs Mean United States 15 -0.56 Cambodia 8 -0.09 United Kingdom 15 -0.14 Taiwan Province of China 15 -1.14 Austria 15 -0.32 Hong Kong S.A.R. of China

15 -1.37 Belgium 15 -0.37 India 15 -0.67 Denmark 15 -0.69 Indonesia 4 -4.51 France 15 -1.57 Korea 15 -2.18 Germany 15 -0.28

Malaysia 15 -2.27 Italy 15 -0.40 Pakistan 3 -2.51 Luxembourg 5 -0.22 Philippines 15 -0.17 Netherlands 15 -0.58 Singapore 15 0.05 Norway 15

-0.88 Thailand 14 -3.66 Sweden 15 -1.11 Algeria 14 -7.45 Switzerland 15 -0.10 Botswana 11 -0.16 Canada 15 0.05 Congo, Republic of 10 0.30 Japan

15 -0.52 Benin 9 -3.63 Finland 15 -2.27 Gabon 7 -2.98 Greece 15 -0.62 Cte d'Ivoire 14 -1.07 Iceland 14 -0.24 Kenya 15 -3.48

Ireland 15 1.02 Libya 15 3.04 Malta 11 -1.39 Mali 8 -3.66 Portugal 15 -0.50 Mauritius 6 -1.38 Spain 15 -1.26 Niger 8

-5.38 Turkey 14 0.43 Rwanda 6 -0.33 Australia 15 -0.64 Senegal 15 -1.27 New Zealand 15 -0.72 Namibia 14 0.65 South Africa 15 -0.66 Swaziland

13 -3.94 Argentina 15 0.16 Togo 13 -1.95 Brazil 15 -2.91 Tunisia 15 2.08 Chile 15 -0.22 Burkina Faso 5 -2.04 Colombia 15 -0.91

Armenia 8 -1.58 Costa Rica 10 -1.04 Belarus 8 -1.13 Dominican Republic 9 -0.54 Kazakhstan 6 -0.28 El Salvador 4 0.58 Bulgaria 8 -0.52 Mexico 15

-0.40 Moldova 11 -3.99 Paraguay 15 -3.11 Russia 13 -4.70 Peru 15 0.73 China,P.R.: Mainland 15 -2.94 Uruguay 15 -0.22 Ukraine 9 -0.37 Venezuela, Rep. Bol.

15 -1.12 Czech Republic 12 0.33 Trinidad and Tobago 10 -2.32 Slovak Republic 12 1.22 Bahrain 15 0.60 Estonia 11 -2.00 Cyprus 6 0.04 Latvia 11 -1.20

Israel 15 -0.27 Hungary 14 -1.88 Jordan 8 1.79 Lithuania 12 -1.47 Lebanon 4 -0.06 Croatia 8 -3.11 Saudi Arabia 13 -0.89 Slovenia 11

-2.79 United Arab Emirates 15 5.66 Macedonia 7 2.01 Egypt 8 -0.16 Poland 7 -1.97 Bangladesh 5 -3.17 Romania 7 -2.86 Table 2. Determinants of FPI/FDI Case 1 Case 1 Case 2

Case 2 Case 3 Case 3 Case 4 Case 4 Case 5 Case 5 Coef. St. err. Coef. St. err. Coef. St. err. Coef. St. err. Coef. St. err. ln(Population) -2.94 0.81 -1.25 0.71 -1.99

0.87 -3.79 0.95 -2.84 1.15 ln(GDP per capita) -0.20 0.38 -0.65 0.34 -0.59 0.40 -0.94 0.42 -0.84 0.43 ln(Market Capitalization) 0.05 0.04 0.09 0.05 0.08 0.05

0.07 0.04 0.09 0.05 ln(Trade openness) -0.89 0.24 -0.38 0.23 -0.56 0.26 -0.45 0.25 -1.10 0.28 ln(M3/GDP) -0.49 0.19 -0.27 0.22 -0.62 0.19 -0.92

0.23 0.25 0.14 0.32 0.13 0.51 0.19 Liquidity Shock 0.25 0.13 Fixed exchange regime Control on FDI outflow Observations 831 860 721 583 414 R-squared (within) 0.10 0.10 0.10 0.17

0.24 Note: Coefficients different from zero at 5% level are highlighted in bold. Year and country fixed effects are included though not reported. Table 3: Determinants of FPI/FDI Table 3: Determinants of FPI/FDI (Distinguished by Country Type) Coef. St. Err. Coef. St. Err. ln(Population) -4.95 1.43 1.60 1.36 ln(GDP per capita) 0.28 0.63 0.45 0.47 ln(Market Capitalization) 0.10 0.08 0.14 0.05

ln(Trade openness) -1.98 0.34 -0.34 0.32 ln(M3/GDP) -0.76 0.31 -0.52 0.24 Observations 279 552 R-squared 0.37 0.12 Note: Coefficients different from zero at 5% level are highlighted in bold. Year and country fixed effects are included though not reported. Table 4a. Probit Estimation of Liquidity Shock Table 4a. Probit Estimation of Liquidity Shock Coef. St Err. ln(Population) -0.06

0.03 ln(GDP per capita) 0.01 0.04 ln(M3/GDP) -0.58 0.08 Bank liquid reserves/assets 0.006 0.003 US real interest rate 0.08 0.03 Fixed exchange regime -0.06 0.12 Constant 1.10 0.66 Observations 1665 R-squared 0.10

Note: Coefficients different from zero at 5% level are highlighted in bold. Table 4b. Determinants of FPI/FDI (With Predicted Liquidity Shock) Table 4b. Determinants of FPI/FDI (With Predicted Liquidity Shock) Case 1 Case 1 Case 2 Case 2 Coef. St. err. Coef. St. err. ln(Population) -3.11 0.81 -3.16 0.80 ln(GDFP per capita) -0.25 0.38 -0.28 0.36 ln(Market Capitalization)

0.05 0.04 0.05 0.04 ln(Trade openness) -0.93 0.24 -0.95 0.24 ln(M3/GDP) -0.11 0.29 Predicted liquidity shock 3.71 2.16 4.31 1.39 Observations 829 829 R-squared (within) 0.11 0.11

Results Probit Estimation We use pooled specification to predict the liquidity crisis, in that fixed-effect Probit regressions are not identified due to incidental parameters problem. Table 3 presents the Probit estimation for all countries from 1970 to 2004, subject to data availability. As we expected, higher US interest rate has a strong spillover effect on the domestic interest rate. Lower sovereign rating raises the chance of liquidity crisis, as risky countries need to raise interest rates to attract capital flows. Higher M3/GDP weakly reduces the likelihood of an aggregated shock, as abundant money supply tends to increase inflation rate while lowering the nominal interest rate. Since both sovereign rating and U.S. interest rate are significant in the Probit estimation, we can then identify the effect of liquidity shock on FPI/FDI through functional form as well as exclusion restrictions. According to Table 3, the predicted probability of liquidity crises in the sample lies between 0.003 and 0.38. FDI/FPI Determination With the predicted probability of liquidity crises, we can now estimate equation (15). We take the log of the FPI/FDI ratio as our dependent variable, to reduce the impact of extreme values. Table 4: Case 1 Table 4 reports the results with country and time fixed effects. As our theory predicts, a higher probability of an aggregated liquidity shock significantly increases the share of FPI, relative to FDI. Moreover, stock market capitalization increases FPI, while trade openness complements FDI. lagged FPI/FDI One might be concerned that lagged FPI/FDI could also affect current FPI/FDI. Hence we estimate, alternatively, the following dynamic panel regression. we use the Arellano-Bond dynamic GMM approach to estimate equation (17), which corrects the endogeneity problem.

Case 2 in Table 4 Case 2 in Table 4 reports the dynamic panel estimation. Dynamic estimation reduces the sample size, but reassuringly, results from fixed effect estimation still carry through. We find that higher probability of aggregated liquidity shocks increases FPI relative to FDI. Stock market capitalization and trade openness keep their signs and significance level. We also find that the one-year lagged FPI/FDI ratio is associated with current FPI/FDI ratio. But the estimated coefficient of the lagged FPI/FDI is around 0.50, which suggests that there is no panel unit root process for FPI/FDI. Additional Arellano-Bond tests strongly reject the hypothesis of no first-order autocorrelation in residuals, but fail to reject the hypothesis of no second-order autocorrelation. Hence, the estimations in Table 4 are valid and provide strong empirical support for our theory. Robustness Checks We add dummies for semi decades into out Probit estimation for interest rate hike. This helps capture unobservable global factors that may affect interest rate hike. We find that explanatory variables maintain their signs and significances in the Probit model. Then we plug this newly estimated probability into the pure fixed effect FPI/FDI model as well as the dynamic one. We find that the estimated probability still has significant explanatory powers in both models. For example, in the dynamic model, it has an estimated coefficient of 2.97 and a p-value of 0.000. Note that we cannot include in the Probit model time effects for every year, which would then perfectly predict U.S. annual interest rate. Alternative Indicator of Liquidity Crises An alternative Indicator of Liquidity Crises: the depreciation of real exchange rate as an alternative measurement of liquidity crisis. The depreciation shrinks the purchasing power of domestic currency and thus decreases the ability of domestic firms to invest abroad. We use the real exchange rate vs. U.S. dollar, instead of the trade-weighted real effective exchange rate. One can collect the data for the latter from the IMFs International Financial Statistics, but will miss quite a few countries such as Brazil and Thailand. That is why we use the real exchange rate vs. dollar. We define currency crisis as the depreciation of more than 15% a year. This amounts to top 5% of the depreciation. Table 5 presents the frequency of currency crisis

for the period from 1970 to 2004. We first apply Probit model to predict the one-year ahead currency crisis. Based on the literature on currency crisis, we use the following explanatory variables: country population size, GDP per capita, GDP growth rate, money stock, U.S. interest rate, trade openness, and foreign reserves over imports. We do not include Standard and Poors country rating here, because it shrinks sample size while having no explanatory power on currency crisis. Table 6 reports the Probit estimation from 140 countries from 1970 to 2004. We can see that higher GDP per capita, higher economic growth, higher reserves over imports and trade openness all contribute to the reduction of currency crises. U.S. interest rate, on the contrary, significantly increases the likelihood of currency crises. All these are intuitive and consistent with previous literature. Based on Table 6, we construct the probability of currency crisis, and then examine its impact on FPI/FDI for the period from 1990 to 2004. Results are reported in Table 7 . Note that Table 7 covers more countries than Table 4, in that we do not include S&Ps country rating as an predictor of currency crises. Case 1 is for the pure fixed effect model. We see that the higher the probability of currency crisis, the higher the ratio of FPI relative to FDI. Case 2 is for the dynamic panel model. Again, we can see that the past movement of FPI/FDI explains the current variation of FPI/FDI. Higher GDP per capita (proxy for labor cost) and trade openness decrease the share of FPI relative to FDI. Our key variable, the probability of currency crisis, still explains the choice between FDI and FPI, consistent with our theory as well as earlier results in Table 4. Both case1 and 2 include year dummies to capture unobservable global factors as well as potential global trends. In both cases, there seems to be a trend of growing FPI relative to FDI, judging from point estimates. The inclusion of year dummies, however, could potentially bias down our estimation, because they also capture global liquidity shock caused by higher U.S. interest rate. Hence, we use a time trend variable instead of year fixed effects in the dynamic model (Case 3). We can see that there is indeed a significant time trend. Moreover, the coefficient of crisis probability now rises to 5.8. This confirms our argument that time fixed effects bias down the effect of currency crisis.

Conclusion Theory In this paper, we examine how the liquidity shock guides international investors in choosing between FPI and FDI. According to Goldstein and Razin (2006), FDI investors control the management of the firms; whereas FPI investors delegate decisions to managers. Consequently, direct investors are more informed than portfolio investors about the prospect of projects. This information enables them to manage their projects more efficiently. However, if investors need to sell their investments before maturity because of liquidity shocks, the price they can get will be lower when buyers know that they have more information on investment projects. We extend the Goldstein and Razin (2006) model by making the assumption that liquidity shocks to individual investors are triggered by some aggregate liquidity shock. A key prediction then is that countries that have a high probability of an aggregate liquidity crisis will be the source of more FPI and less FDI. To test this hypothesis, we therefore apply a dynamic panel model to examine the variation of FPI relative to FDI for 140 source countries from 1990 to 2004. We use real interest rate hikes as a proxy for liquidity crises. Using a Probit specification, we estimate the probability of liquidity crises for each country and in every year of our sample. Then, we test the effect of this probability on the ratio between FPI and FDI generated by the source country. We find strong support for our model: a higher probability of a liquidity crisis, measured by the probability of an interest rate hike, has a significant positive effect on the ratio between FDI and FPI. We repeat this analysis using real exchange rate depreciation as an alternative indicator of a liquidity crisis, and get similar results. Hence, liquidity shocks do have strong effects on the composition of foreign investment, as predicted by our model.

Recently Viewed Presentations

  • Marine Corps Enlisted Administrative Separation (Mceas)

    Marine Corps Enlisted Administrative Separation (Mceas)

    The IPAC/admin will report the EAS CofG M on the unit diary upon notification from the unit Commander and validation that the Marine is in a limited duty status. Reporting of the entry will result in the EAS being established...
  • CHAPTER 5 Small Business and the Entrepreneur

    CHAPTER 5 Small Business and the Entrepreneur

    R. M. Belbin outlined a model of nine team roles. An effective team requires a variety of roles. Technology allows for virtual teams in which members are in different locations around the country or the world. Web casts, electronic file...
  • Chapter 8 Work and Machines - Council Rock School District

    Chapter 8 Work and Machines - Council Rock School District

    Section 1: Work and Power. Work is the transfer of energy that occurs when a _____ is applied over a _____. Work is done when an object moves in the same direction of the force. 2 things must happen for...
  • Effects of the Growth of American Industry  Embargo

    Effects of the Growth of American Industry Embargo

    After the Harford Convention, Federalists all but disappeared. 1816 is the last year the Federalists run a candidate for President. Presidential elections of 1820, and 1824 were all single party contests with the Democratic-Republicans in control
  • Health Expenditure in the Provinces and Territories

    Health Expenditure in the Provinces and Territories

    Health Expenditures in the Provinces and Territories, 2017. The following slides are provided forreuse in your own presentations. Please cite as follows: Canadian Institute for Health Information.
  • GACFall Recruitment Training for Unit leaders

    GACFall Recruitment Training for Unit leaders

    So, tonight at 7:00pm, right here in your school cafeteria you can sign up to be a Cub Scout! And I'm going to give each of you a sticker that has all the information on it. Once again it's tonight...
  • Exploring the Impact of Cognition on Prosthetic ...

    Exploring the Impact of Cognition on Prosthetic ...

    Exploring the Impact of Cognition on Prosthetic Rehabilitation for Lower Limb Amputees. A Literature Review. By . Lucy Pearce. ... Search broadened with additional keywords due to limited results: ... Exploring the Impact of Cognition on Prosthetic Rehabilitation for Lower...
  • Chapter 3

    Chapter 3

    Return On Investment (ROI) is a common measure of organizational financial success Return On Net Worth (RONW) measures profitability of funds invested in the business Return On Assets (ROA) provides insight on how well managers utilize operational assets to generate...