The Macroeconomic Impact of

1

• The Macroeconomic Impact of
• Sovereign Wealth Funds
• Julie Kozack
• Senior Economist, Research Department
• International Monetary Fund
• September 2008
• DISCLAIMER The views expressed are my own and
  should not be
• attributed to the staff, management, or Executive
  Board of the IMF

2
Outline of Presentation

• Rationale for SWFs
• Current and Future Size of SWFs
• Portfolio Shifts and the New SWFs
• Illustrative Scenarios Implications of Global
  Capital Flows and Key Asset Prices
• Concluding Thoughts

3
Rationale for SWFs

• Large current account surpluses in many emerging
  economies.
• Driven by manufacturing (Asia) and commodity
  (Middle East/Latin America/Africa) exports.
• Significant reserve accumulation in many emerging
  economies.

4
Emerging Economy Reserves(In billions of U.S.
dollars)
5
Macroeconomic Impact of SWFs

• Rationale for SWFs
• Current and Future Size of SWFs

6
SWFs Are Sizeable Currently

• SWFs 2-3 trillion in estimated assets under
  management.
• Examples ADIA, Norway, China, Singapore
• Some sub-sovereign SWFs also considered (Alaska,
  Alberta).
• Pension funds with explicit individual
  liabilities not considered (i.e., CalPers).
• My colleagues (Krishna Srinivasan) presentation
  discussed definitional issues in more detail.

7
SWFs By Region(Percent Share)
8
SWFs By Type(Percent share)
9
SWFs Are Projected to Grow Rapidly

• SWFs are projected to grow further.
• Projections are based on data regarding
  available external surpluses.
• Implicit assumptions that fiscal and external
  surpluses mirror each other (not true in case of
  Australia, for example).

10
Updated SWF Projection

• SWFs could reach 7-11 trillion
• in 2013.

11
SWFs Projections By Region
(Percent share)
(Billions of U.S. dollars)
12
Macroeconomic Impact of SWFs

• Rationale for SWFs
• Current and Future Size of SWFs
• Portfolio Shifts and the New SWFs

13
SWF Portfolios What Do We Know?

• SWF portfolios more diversified than traditional
  reserves holdings.
• SWFs hold greater stakes in equities and wider
  geographical dispersionprecise portfolio
  compositions are known for very few funds (Norway
  a prominent exception).
• Some market participants expect SWFs portfolios
  to evolve to resemble those of the larger public
  sector pension funds.
• Some SWFs have also been exploring alternative
  investments including hedge funds, private
  equity, and real estate.

14
Factors in Support of Financial Stability

• Long-term investment horizon and limited
  liquidity needs (except stabilization funds).
• Willing to accept short-term volatility in return
  for higher expected long-term returns.
• May be aided by less reliance on pro-cyclical
  regulatory requirements (e.g. fair value
  accounting, mark-to-market).

15
SWF Portfolio Shifts

• Established SWFs no reason to expect major,
  disruptive portfolio shifts.
• If anything, SWFs have been a source of
  stability, especially in the past year.
• New SWFs may switch from more conservative
  form of investing to more diversified investment
  strategy.

16
Macroeconomic Impact of SWFs

• Rationale for SWFs
• Current and Future Size of SWFs
• Portfolio Shifts and the New SWFs
• Illustrative Scenarios Implications of Global
  Capital Flows and Key Asset Prices

17
Implications of Portfolio Diversification for
Global Capital Flows

• Diversification of reserve assets through SWFs
  could have implications for global capital flows.
• Diversification across two spectra asset class
  and geography.
• SWFs could diversify away from low-risk,
  short-term instruments such as treasury bills
  into equities and other high(er)-risk
  instruments, such as private equity and real
  estate.
• SWFs could also focus more on emerging economies,
  with attendant implications for capital flows to
  mature markets.
• Potential impact on macroeconomic outcomes and
  asset prices across both advanced and emerging
  economies.

18
Illustrative Scenario Based on Diversification of
Reserve Assets

• Illustrative scenario based analysis of the macro
  impact of diversification of reserves through
  SWFs.
• Focus of scenarios is on new sovereign
  investments by countries that have either
  recently set up SWFs or are contemplating their
  establishment.
• Limited exercise provides a sense of the
  direction and magnitude of the possible impact on
  markets.
• Results have to be assessed with great
  cautionnotably because of their sensitivity to
  underlying assumptions.

19
Estimating SWF Flows

• Available foreign currency flows for each country
  with a newly established SWF (Brazil, China,
  Korea, Russia, Saudi Arabia).
• Lower boundassume that countries that have
  recently established SWFs channel 50 percent of
  available foreign currency inflows to their SWFs.
• Upper boundin addition, countries that are
  contemplating setting up SWFs (based on market
  reports) channel 50 percent of newly available
  foreign currency inflows to their SWFs (India,
  Japan, Thailand).
• Upper bound also assumes that 10 percent of the
  stock of existing reserves of the top 10 emerging
  market reserves holders is shifted from reserves
  to SWF holdings over the period 2008-13.

20
Stylized Portfolios

• Estimate possible change in pattern of global
  capital flows (by currency).
• Construct reserve portfolio using IMFs
  Currency Composition of Official Foreign Exchange
  Reserves (COFER) database.
• Construct two stylized SWF portfolios
  Norway-like (based on actual portfolio data) and
  a more diversified portfolio (based on media
  reports and own estimates).
• Projected changes in the pattern of global
  capital flows are derived by comparing investment
  flows based on COFER portfolio and stylized SWF
  portfolio.

21
Currency Compositionof Stylized Portfolios (in
percent)
Pound
U.S. dollar
Other
Yen
Euro
COFER
Diversified Portfolio
Norway
22
Key Simplifying Assumptions

• Currency flowscountry flows.
• Assume that reserves are invested 100 percent in
  government securities.
• Investments not in dollar, euro, pound, or yen
  (i.e., other) are assumed to be mostly emerging
  economy investments.

23
Impact on Global Capital Flows

• Results imply a range of estimates for global
  capital flows, by currency.
• Inflows into U.S. dollar assets would decline
  (relative to baseline), but by manageable
  levels.
• Exercise does not necessarily imply fall in
  absolute level of inflows to the United States,
  but inflows could decline relative to GDP.
• Capital inflows could increase (significantly) to
  emerging economies, based on diversification.
• Modest impact on euro, pound, yen.

24
Possible Change in Pattern of Capital Flows, by
Currency(In billions of U.S. dollars)
25
Range of Possible Outflows from U.S. Dollar
Assets (In billions of U.S. dollars)
Range
Range
26
Simulation Results

• Utilize IMF macro model to estimate impact of
  fall in capital flows to the United States
  (relative to baseline) on key asset prices.
• U.S. current account balance improves by ½-1
  percent of GDP relative to baseline.
• Model results
• Modest depreciation of the U.S. dollar and
  increase in U.S. interest rates.
• Appreciation of other currencies (including
  euro).
• Lower interest rates in rest of world.
• Deteriorated current accounts in rest of world.

27
Health Warning (!)

• Your are about to enter the world of a highly
  stylized dynamic stochastic general equilibrium
  (DSGE) model called GIMF.

28
Simulation Results Range of Possible Effects on
U.S. Current Account, Exchange Rates and Interest
Rates (Changes Relative to Baseline)
U.S. Real Effective Exchange Rate (In percent
positiveappreciation)
U.S. Real Interest Rate (basis points)
U.S. Current Account (percent of GDP)
29
Simulation Results Range of Possible Effects on
Euro Area Current Account, Exchange Rates and
Interest Rates(Changes Relative to Baseline)
Euro area Real Effective Exchange Rate (In
percent positiveappreciation)
Euro area Real Interest Rate (basis points)
Euro area Current Account (percent of GDP)
30
Simulation Results Range of Possible Effects on
Emerging Asia Current Account, Exchange Rates and
Interest Rates(Changes Relative to Baseline)
VERY PRELIMINARY RESULTS!
Emg. Asia Real Effective Exchange Rate (In
percent positiveappreciation)
Emg. Asia Real Interest Rate (basis points)
Emg. Asia Current Account (percent of GDP)
31
Intuition

• Short Run
• Rest of the world reduced demand for U.S.
  assets. Modeled as higher risk premium on U.S.
  assets.
• U.S. real interest rates increase ROW real
  interest rates decline.
• U.S. consumption, investment decline.
• Dollar depreciates, trade balance improves.
• U.S. current account improves.

32
Intuition (contd)

• Long run
• Improvement in U.S. current account implies
  higher (relative to baseline) accumulation of net
  foreign assets.
• Over time, the trade balance deteriorates and is
  offset by improved interest payments on NFA.
• Dollar depreciates by less in 10 years, actually
  appreciates in very long run.

33
Model Caveats

• Significant assumptions throughout on various
  model parameters.
• Main substantive caveat model does not include
  valuation effects, which could be large with a
  dollar depreciation.
• Model effects could be amplified OR
• The adjustment could be very large and abrupt if
  investors switch to holding assets in another
  currency (see reserve currency literature).

34
Conclusions

• No evidence of large (or disruptive) SWF
  portfolio shifts.
• Benign shifts aimed at diversifying asset
  allocation can have an impact on global capital
  flows and asset prices.
• Our estimates suggest that the overall impact on
  the dollar would be modest.
• The impact on emerging economies may be greater
  since those markets are smaller.