FDI : WHAT IS IT ?
Focus : Does “FDI”, like domestic direct investment, represent “gross fixed capital formation”, i.e. additions to the (value of the) physical capital stock ?
- In the general public’s eye – YES
- Officially – NO
Re official FDI : part of it may effectively be transformed into additional consumption spending and imports; turned into additional capital outflows; or remain in domestic financial markets as savings – for long periods.
From a growth perspective, FDI may also entail the adoption of new / improved technologies ; as well as additions to the stock of human capital.
Outline : 1. FDI : OECD / IMF Definition
2. Transmission Channels
3. SA Example
4. Theoretical Observations
- Physical Factor Augmentation
- Skills / Knowledge Transfer / Production
- “Crowding-in Effects”
5. Implications for Empirical Research
6. Conclusion
1. FDI : OECD / IMF DEFINITION
· When a foreign company acquires more than 10 % of the ordinary shares or voting rights of a local business entity.
+ “lasting interest” + “a significant influence on management”
(IMF; OECD)
But :
- What if it acquires 9 % and it is used for ∆ K?
- What if it acquires 51 % with no ∆ K – merely a transfer of ownership? E.g. cross-border acquisitions & mergers (A & Ms).
· Re A & Ms: they constitute a significant share of FDI flows, both in developed and developing countries (Mody, 2004).
FDI in SA in the 1990s “(mostly) involved the acquisition of existing operations rather than the setting up of new enterprises” (Gelb & Black, 2004).
Advantages: new technologies; skills transfer; risk spreading; gaining market share, both locally and globally; etc.
But :
- How should these be valued?
· In the public eye FDI is (wrongly) assumed to be “greenfield” investments, i.e. creation of new facilities or expansion of existing facilities – gross fixed capital formation by foreigners.
And yet only a small portion of official FDI may actually constitute greenfield investments.
2. TRANSMISSION CHANNELS
Equity K Outflows
K M
Inflows
(FDI)
Fixed
Assets C
Technology Transfer;
Fixed Capital + Knowledge Production
Formation Etc.
· Official FDI may directly lead to fixed capital formation (e.g. a new residence, new hotel); or they may work their way through the equity markets and through foreign purchases of fixed assets, some of which may give rise to direct investment / FDI.
· Alternatively, the money received may be (i) used for consumption purposes (C), (ii) spent on imports (M), or (iii) leave the country as capital outflows. It may also (iv) remain within the domestic financial and property markets in the relevant period (t). The latter amount may eventually (period t+1) be spent abroad or locally – as above.
Thus, let total spending or demand,
Y d = ( C ) (1+f c ) – M (1+f m ) + I (1+f i ) + X
where fc is the fraction by which official FDI boosts private and public consumption, and fm is fraction by which it boosts imports.
Likewise, fi is the fraction by which official FDI boosts investment, or “ gross fixed capital formation”; with fi(I) being greenfield investments.
3. SA EXAMPLE
· Barclays acquired ownership of ABSA and its capital stock in 2005 – without necessarily enhancing it.
· The Barclays bid came to about R 30 billion; was settled in July 2005; and was (presumably) recorded as FDI in 2005 : Q3 – see Tables 1 and 2 below:
Table 1: Financial account (2003:Q04 to 2005:Q03)
/ 2003 / 2004 / 200504 / 2003 / 01 / 02 / 03 / 04 / 2004 / 01 / 02 / 03
Direct investment
Liabilities
/ 3 021 / 5 550 / 8,630 / -1 605 / 2 007 / -3 877 / 5 155 / 1 369 / 1 371 / 32 175Assets / -1 017 / -4 275 / 237 / -9 785 / -380 / 1 207 / -8 721 / -633 / 3 161 / -1 700
Net direct investment / 2 004 / 1 275 / 8,867 / -11 390 / 1 627 / -2 670 / -3 566 / 736 / 4 532 / 30 475
Portfolio investment
Liabilities
/ -6 107 / 7 548 / 5,105 / 14 095 / 2 984 / 22 691 / 44 875 / 7 808 / 25 281 / 7 296Assets / -555 / -1 001 / -18 / -1 493 / -1 279 / -3 156 / -5 946 / 2 264 / -1 541 / -3 225
Net portfolio investment / -6 662 / 6 547 / 5,087 / 12 602 / 1 705 / 19 535 / 38 929 / 10 072 / 23 740 / 4 071
Other investment
Liabilities
/ 9 263 / 14 594 / 4 899 / 5 112 / -2 888 / 3 758 / 10 881 / 22 291 / -2 416 / 1 938Assets / -17 762 / -36 919 / -11 102 / 3 251 / 1 589 / 4 099 / -2 163 / -6 343 / 746 / -26 375
Net other investment / -8 499 / -22 325 / -6 203 / 8 363 / -1 299 / 7 857 / 8 718 / 15 948 / -1 670 / -24 437
Source: SARB 2005
Table 2: Foreign Liabilities : Direct Investment
2003 2004 2005 2006
Banking Sector: 165 1640 31 134 268
· The Barclays-ABSA transaction did not constitute “direct investment”, or “greenfield” investments – at least not in 2005.
· Of course, the acquisition could have entailed efficiency gains, etc.
4. THEORETICAL OBSERVATIONS
FDI is generally perceived to be a means of supplementing domestic savings and augmenting the physical capital stock.
· E.g. re Factor Augmentation a-la Solow-Swan:
Y = A K a N 1 - a
Or in growth terms,
y = a + a k + ( 1 - a) n
Now, allowing for FDI,
k = s s + k f
where s is the (domestic) savings ratio, s is the output :
capital ratio, and k f is the growth rate of FDI (which can be
positive or negative).
All of k f is assumed to add to the (value of) the physical
capital stock, i.e. gross fixed capital formation.
· Similarly, technology transfer (e.g. Blomstrom & Kokko, 1998; also Stokke et al, 2007)
a = f ( k f )
· Re Skills / Knowledge Transfer: e.g. Romer’s goods & knowledge production sectors:
Y = K ay ( A N y ) 1 - a
.
A = B ( K a ) b ( N a ) π A θ
So, as before :
k a = s a s a + k a f (e.g. Gorg & Greenaway, 2004)
· Also, “Crowding-in Effects” :
Let I d and I f represent capital investments undertaken by domestic institutions and foreigners, respectively. Then
I d = I d ( I f , ...... ) with I d ¢ ( I f ) 0
with S higher and C lower
( e.g. Mody & Murshid, 2002 )
Similarly for exports and imports ; e.g.
X = X ( I f , ...... ) with X ¢ ( I f ) 0
(e.g. Eichengreen, 2000; Ozturk, 2007)
Also, domestic investment may be seen as a prerequisite for FDI ; e.g.
I f = I f ( I d , ...... ) with I f ¢ ( I d ) 0
Important issues:
Causality issue : growth FDI (e.g. Chakraborty & Basu, 2002)
“Minimum threshold stock of human capital” (Borensztein, 1995)
Domestic country’s financial system (Hermes & Lensink, 2003)
5. IMPLICATIONS FOR EMPIRICAL RESEARCH
Let official or broadlydefined FDI be given by
FDI b = f i (I k + I h) + ( f c C ) + ( f m M ) + ( f w CWO ) + ( f k o KO) [1]
where f i now represents the fraction by which FDI boosts physical and human capital ( I h ); f w is the fraction by which FDI contributes to “changes in (domestic) wealth ownership”, CWO (or “saving”); and f k o is the fraction by which it boosts capital outflows, KO.
· Impact of FDI on GDP
Y = α b + β b FDI b + ….. + ε b [2]
where Y is real GDP (growth), with α b and β b being the estimated coefficients and ε b the error term.
Now, the equivalent econometric equation for that part of FDI constituting fixed and human capital formation,
Y = α i + β i [f i (I k + I h)] + ….. + ε i [3]
is likely to yield different results from those generated by equation (2) with β b ≠ β i . Clearly, this difference will depend on the extent to which FDI b varies over time relative to f i (I k + I h).
· Impact of GDP on FDI
FDI b = α b + β b Y + ….. [4]
f i (I k + I h) = α i + β i Y + ….. [5]
where, as before, β b ≠ β i
6. CONCLUSION
· Official FDI ≠ gross fixed capital formation :
- Part is transformed into additional consumption, imports and capital outflows.
- Another part remains in the financial and fixed property markets as savings – for long periods.
- A part is transformed into gross fixed capital formation.
· Can one assume that official FDI does somehow equal additions to the capital stock – broadly defined ? E.g. that
FDI = GFKF + HKF + TT + I n
with TT being technology transfers (or increases in the value of technologies used); and In represents induced investment (e.g. I d = f ( I f )) ?
How would one measure these components?
E.g. impact of FDI on GDP growth :
a GFKF
b HKF
Growth c
TT
d
I n
Similarly re impact of GDP growth on FDI
· Your problem – not mine !
3