How do IMF announcements affect financial markets in crisis: Evidence from forward exchange markets

1
How do IMF announcements affect financial markets
in crisisEvidence from forward exchange markets
versus bond spreads
ACADEMY OF ECONOMIC STUDIES DOCTORAL SCHOOL OF
FINANCE AND BANKING

• MSc Student Raluca Turchina
• Coordinator Professor MOISA ALTAR

Bucharest, July 2009
2
Dissertation paper outline

• The aims of the present paper
• The importance of International Monetary Fund
• Brief review of recent literature
• The data
• The Component GARCH model
• The Exponential GARCH model
• The Longstaff and Schwartz model
• Concluding remarks
• References

3
The aims of the paper

• To demonstrate the difficulty of
  providing unambiguous interpretations on the
  impact of IMF events on private financial
  markets. The liquidity or the moral hazard
  interpretation has been achied till now only in
  theoretical studies. We employ a theoretical
  model developed by Corsetti and Roubini(2006) in
  which liquidity and moral hazard effects of IMF
  support are developed, and interpret our
  empirical results accordingly
• We investigate the response of forward
  exchange markets to IMF-related news
• We examine weather the results
  associated with forward exchange markets are
  consistent with the response of Eurobond spreads
  to IMF-related news

4
The importance of International Monetary Fund

• The International Monetary Fund (IMF) is
  an organization of 186 countries, created in
  1944, working to foster global monetary
  cooperation, secure financial stability,
  facilitate international trade, promote
  sustainable economic growth and reduce poverty
  around the world.
• As the Second World War ends, the job of
  rebuilding national economies begins. The IMF is
  charged with overseeing the international
  monetary system to ensure exchange rate stability
  and encouraging members to eliminate exchange
  restrictions that obstruct trade.
• After the system of fixed exchange rates
  collapses in 1971, countries are free to choose
  their exchange arrangement. Oil shocks occur in
  197374 and 1979, and the IMF steps in to help
  countries deal with the consequences.
• The oil shocks lead to an international
  debt crisis, and the IMF assists in coordinating
  the global response.
• The IMF plays a central role in helping
  the countries of the former Soviet bloc
  transition from central planning to market-driven
  economies.

5
The importance of International Monetary Fund

• The implications of the continued rise
  of capital flows for economic policy and the
  stability of the international financial system
  are still not entirely clear. The current credit
  crisis and the food and oil price shock are clear
  signs that new challenges for the IMF are waiting
  just around the corner.
• The IMF's fundamental mission is to
  help ensure stability in the international
  system. It does so in three ways keeping track
  of the global economy and the economies of member
  countries lending to countries with balance of
  payments difficulties and giving practical help
  to members

6
Brief literature review

• The Component GARCH model introduced by Engle
  and Lee (1993), used in recent papers such as
  Evrensel and Kutan (2008) regarding the forward
  exchange rate shows consistently results with the
  response of stock and bnd markets.
• Respect to studies on bond spreads Kamin and
  Kleist(1999), Eichengreen(2005), Evrensel and
  Kutan(2006,2007) Hayo and Kutan(2005). Using the
  model GARCH for analysis, program negotiations
  and approval seem to increase bond yield in
  emerging countries (Hayo and Kutan, 2005). But
  when country specific data are used, Korean
  negotiations as well as Indonesian and Korean
  approval decrease bond yields in the mentioned
  countries (Evrensel and Kutan(2006,2007)).
• The framework developed by Longstaff and Schwartz
  (1995) which incorporates both default and
  interest rate risk in the valuation of risky
  corporate fixed and floating rate debt.

7
Corsetti and Roubini model

• Giancarlo Corsetti and Nouriel Roubini
  (2006) have developed a model regarding the
  trade-off between official liquidity provision
  and debtor moral hazard. In this model
  international financial crises are caused by the
  interaction of bad fundamentals, self-fulfilling
  runs and policies by three classes of optimizing
  agents international investors, the local
  government and the IMF. The model shows how an
  international financial institution helps prevent
  liquidity runs using coordination of agents'
  expectations, by raising the number of investors
  willing to lend to the country for any given
  level of the fundamental.
• The results of the model point out
  that the influence of such an institution is
  increasing in the size of its interventions and
  the precision of its information more liquidity
  support and better information make agents more
  willing to roll over their debt and reduces the
  probability of a crisis. Different from the
  conventional view stressing debtor moral hazard,
  the model shows that official lending may
  actually strengthen a government motivation to
  implement desirable but costly policies. By
  worsening the expected return on these policies,
  destructive liquidity runs may well discourage
  governments from undertaking them, unless they
  can count on contingent liquidity assistance.
• The model show that limited
  contingent liquidity support can motivate fund
  managers or private investors to rollover their
  exposure to the country. The model also suggests
  that liquidity support always induces moral
  hazard distortions is incorrect.

8
The Data

• We examine the impact of IMF related
  events on financial markets during financial
  crises in two directions by using forward
  exchange rate volatility patterns of Romania and
  Hungary and on the other hand the Eurobond
  spreads in the same countries.
• Hungary and Romania were two of the
  first CEE countries affected by the global
  financial turbulence, government securities
  market and some other keys markets experiencing
  stress over a period of time. These pressures
  emerged despite the countries' improved
  macroeconomic policies of the past years such as
  strengthening the fiscal position, narrowing the
  current account deficit and a cautious
  implementation of monetary and exchange rate
  policies.
• Daily nominal forward exchange rates
  of the above CEE currencies against the euro,
  that is the Romanian Leu (RON) and the Hungarian
  forint (HUF). Each exchange rate is quoted as
  number of national currency units per euro.

9

• The sampling period for forward
  exchange rate covers 08th of May 2005 to 5th of
  May 2009 for the bond spreads we use the sample
  period starting 20th of March 2007 to the end of
  June 2009
• All series in levels display a unit
  root, as evident from the ADF test results. Hence
  the series are transformed into log-differences
  between the forward exchange rate and the spot
  rate multiplied by 100, respectively the
  difference between the benchmark and the Eurobond
  over the same maturity, and we obtain the
  continuously compounded forward exchange rate
  returns, respectively a continously bond spreads
  series
• As to the method of estimation, we use
  GARCH models to account for the time-varying
  volatility displayed in daily financial data. We
  experimented both with standard GARCH as well as
  asymmetric Exponential GARCH models. We found
  that the standard GARCH(1,1) fits the data much
  better than the asymmetric models.

10
The GARCH Model
The conditional variance in the GARCH(1,1) model
can be written as
Eq.(1)
Eq.(2)

• Rt indicates financial sector returns in period
  t.
• Eq. (1) is the mean equation, which is written
  as a function of a constant and an error term.
  The error term in Eq. (1), t, is assumed to have
  a time-varying variance given by Eq. (2).
• The conditional variance of returns at time t
  is predicted based on the persistence in the last
  periods shocks (t-1) and the last
  periodsconditional variance (t-1 ).
• ß1 and ß2 indicate the short term dynamic of the
  resulted volatilty series. ß2 shows the
  persistence of the volatility and ß1 the
  reactivity of the volatility.

11
GARCH Estimates
12
GARCH Estimates
13
The EGARCH Model

• The EGARCH model or Exponential GARCH has the
  following specific
• conditional variance equation

The equation implies that leverage effect is
exponential and that the predictions of the
conditional volatility is guaranteed to be
nonnegative. The presence of the leverage effect
is tested by the following condition
If the impact is asymmetric. The
error distribution of the model is supposed to be
normal.
14
EGARCH Estimates
15
The Longstaff Schwartz Model

• The basic valuation framework of
  Longstaff and Schwartz (1995) assumes that the
  dynamics of the asset value (V) implies the
  addition of the changes in V both over time and
  due to a standard Wiener process (Z1)
• (1)
• - µ and s are constants. Similarly, the
  dynamics of the risk-free interest rate (r)
  involves the changes in the rate over time plus a
  standard Wiener process (Z2 ).
• (2)
• - ?, ß, and ? are constants.
• It is also assumed that
• Financial stress occurs, when VK,
  where K is the threshold value of the firm.
  Therefore, as long as VgtK, the firm is meeting
  its contractual obligations. If default occurs
  during the life of a security, the security
  holder receives (1 -?) times the face value of
  the security at maturity, where represents the
  percentage reduction in payoff on a security,
  which is assumed to be constant and ?lt1.

16

• Assuming perfect, frictionless markets,
  in which securities trade in continuous time,
  Longstaff and Schwartz (1995) derive the price of
  a security with payoff time T, contingent on the
  value of V and r in the form of the following
  partial differential equation
• (3)
• Where
• The market price of risk, v, can be
  derived assuming logarithmic investor preferences
  within a general equilibrium framework. The value
  of the security is obtained by solving Eq.(3)
  subject to appropriate maturity condition. Using
  the above framework, the value of a risk-free
  discount bond D (r, T) is
• (4)
• Where

17

• The above valuation framework for the
  risk-free discount bond can be used to derive the
  valuation for risky discount and coupon bonds.
  Suppose P(V, r, T) is the price of a risky bond
  with maturity T. If payoff equals to unity in the
  absence of default, it becomes (1 -?) in its
  presence.
• The payoff in two states of the world
  can be expressed as
• where I is the indicator function that
  takes on the value of 1 if VK and zero
  otherwise.
• If I1, if the passage of time,?, during
  which the firms value (V) approaches the
  threshold value (K) is less than or equal to T.
• The next step eliminates V and K, X
  V/K. The value of a risky bond becomes
• (6)
• where

18

• Eq. (6) indicates that the value of the risky
  discount bond depends on V and K through their
  ratio X.
• X captures the default risk of the firm, which is
  a proxy variable for the firms credit rating.
• Eq. (6) also indicates that the value of the
  risky discount bond consists of two parts
• 1 - the first term treats its value as
  if the bond were risk-free
• 2 - the second term represents a
  discount for the default risk of the bond, which
  has two components
• D(r, T)- the present value of the
  value loss on the bond in the event
• of a default
• Q(X, r, T), implies the probability
  of default.

19

• The above model forecasts the
  following relations between the price of a risky
  bond and default risk, reduction in payoff,
  maturity, and the risk-free interest rate
• the price of a risky bond is an increasing
  function of the default-risk variable, X, because
  the higher values of X are associated with the
  situation, in which the companys value is far
  away from the default threshold
• the bond values are decreasing functions of the
  reduction in payoff, ?, because an increase in ?
  implies that the reduction of payoff on a bond in
  the event of financial distress is larger
• the value of the risky bond is a decreasing
  function of maturity, T, because as T increases,
  the value of D (r, T) decreases, and the
  risk-neutral probability of a default Q(X, r, T)
  increases
• the price of the risky bond is a decreasing
  function of the risk-free interest rate, r.
• The Longstaff and Schwartz model
  implies that credit spreads are negatively
  related to the level of interest rates, which
  pressure the fact that not only the default risk
  but also the interest rate risk are necessary
  components for a valuation model of risky debt.
• The Longstaff and Schwartz model
  motivates the inclusion of the yield curve in our
  spread estimations.

20

• We have used similar conventions
  of Batten (2005) that also takes the Longstaff
  and Schwartz model as their theoretical basis and
  we have estimate the following GARCH (1,1)
  specification that reacts the Longstaff and
  Schwartz framework
• - is the change in Eurobond at time t, S the
  arithmetic difference between the risky and the
  default free (benchmark) Eurobond
• - is the change in the benchmark risk free
  interest rate , which is equivalent to EURIBOR3M
  risk free interest rate used in the spread
  calculation
• and - is the change in interest rate
  swaps(IRS), long term product used for hedging
  the interest rate risk and credit default risk of
  the bonds, the slope yield curve serving as a
  proxy for the expected interest rate
• - indicates the conditional variance term
  and is a function of three components
• Long term mean
• News form the previous period, ARCH term
  
• Conditional variance from the previous period,
  GARCH term
• N, D, A indicate IMF-related news dummies
  regarding the start of negotiations, the
  negotiations duration and the program approval.

21
GARCH Estimations
22
EGARCH Estimations
23
Concluding remarks

• By comparing our results regarding of the
  forward exchange rates and the eurobond spreads
  in Romania vs Hungary, the outcome indicate that
  the response of various financial markets
  reaction to IMF-related news are different with
  each other, taking into consideration the
  fragility or the strengths of the economic
  environment before the IMF intervention of each
  country.
• In Romania, the impact of IMF related
  news on the forward exchange rate is more evident
  in the beginning of the negotiations. As it
  regards the Eurobond spreads, the optimistic
  sentiment given by the loan is captured better by
  the negotiations period. This could be explained
  by the fact that the investors have reacted
  positive only in the beginning of the
  negotiations, due to the confidence of receiving
  a loan from IMF, but on the other hand an IMF
  agreement can not improve the economical
  environment on longer term, even if the
  conditions implied by the lendig are very
  restrictive.
• In Hungary, especially the duration of
  negotiations with the IMF is associated with a
  forward premium on the forward exchange rate and
  is consistent with an increasing spreads over the
  Eurobond spreads.
• If investors expect future devaluation
  of the currency for reasons related to expected
  economic instability, their positive response to
  a future IMF program may indicate moral hazard.

24
Bibliography

• Corsetti, G., Guimaraes, B., Roubini, N., (2006),
  International lending of last resort and moral
  hazard a model of IMFs catalytic finance, J.
  Monetary Econ. 53, 441471.
• Brealey, R.A., Kaplanis, E.C., July(2000), The
  Impact of the IMF Programs on Asset Values,
  Working Paper, London Business School.
• Batten, J.A., Fetherston, T.A., Hoontrakul, P.,
  (2005), Factors affecting the yields of emerging
  market issuers evidence from the Asia-Pacific
  region, Journal of International Financial
  Markets, Institutions and Money 16, 5770.
• Beck, T., Demirg, Kunt, A., Levine, R.,(2000),A
  new database on financial development and
  structure, World Bank Economic Review 14,
  597605.
• DellAriccia, G., Schnabel, I., Zettelmeyer, J.,
  (2002), Moral hazard and international crisis
  lending a test, International Monetary Fund,
  Working Paper 181.
• Eichengreen, B., Kletzer, K., Mody, A., (2005),
  The IMF in a world of private capital markets,
  J.Bank. Finance 30, 13351357.
• Evrensel, A.Y., Kutan, A.M., (2006a),Creditor
  moral hazard in stock markets empirical evidence
  from Indonesia and Korea, J. Int. Money Finance
  25 (4), 640654.
• (2007b),IMF-related announcements and stock
  market returns evidence from financial and
  non-financial sectors in Indonesia, Korea, and
  Thailand, Pacific-Basin Finance J. 15 (1),
  80104.
• (2008c), Impact of IMF-related news on capital
  markets Further evidence from bond spreads in
  Indonesia and Korea, J. Int. Financial Markets
  Inst. Money 18 (2), 147160.
• (2008d), How do IMF announcements affect
  financial markets in crises? Evidence from
  forward exchange markets, Journal of Financial
  Stability, 121-134
• Ganapolsky, E.J.J., Schmukler, S.L., (1998),
  Crisis Management in Argentina during the 1994 -
  95 Mexican Crisis How did markets react? World
  Bank Policy Research, Working Paper No. 1951.
• Grenville,S., (2000), Monetary policy and the
  exchange rate during the crisis Bull. Indonesian
  Econ. Stud. 36, 4360.

25

• Hayo, B., Kutan, A.M.,(2005), IMF-related news
  and emerging financial markets. J. Int. Money
  Finance 24, 11261142.
• Kutan, A.M., Sudjana, B.C., (2003), Investor
  reaction to IMF actions in the Indonesian
  financial crisis, J. Policy Reform 6, 181190.
• Kamin, S.B., Kleist, K., (1999), The evolution
  and determinants of emerging market credit
  spreads in the 1990s,. BIS Working Papers, No.
  68.
• Kho, B.C., Stulz, R.,(1999),Banks, the IMF and
  the Asian Crisis, Working Paper, Ohio State
  University.
• Kho, B.C., Lee, D.W., Stulz, R.M., (1999),U.S.
  Banks, Crises, and Bailouts From Mexico to
  LTCM, European Corporate Government Institute
• Kaminsky, G.L., Schmukler, S.L., (1999), What
  triggers market jitters? A chronicle of the Asian
  crisis, Journal of International Money and
  Finance 18, 50 and 514
• Lane, T., Phillips, S., (2000),Does IMF
  financing result in moral hazard? , IMF Working
  Paper, No. 168.
• Longstaff, F.A.,Schwartz, E.S., (1995), A simple
  approach to valuing risky fixed and floating rate
  debt, Journal of Finance 50, 789819
• Sarno, L., Taylor, M.P., (1999), Moral hazard,
  asset price bubbles, capital flows, and the East
  Asian crisis the first tests. J.Int. Money
  Finance 18, 637657.
• Sander, H., Kleimer, S., (2003),Contagion and
  causality an empirical investigation of four
  Asian crisis episodes, Journal of International
  Financial Markets, Institutions and Money 13,
  171186.
• Tillmann, P., (2001), Switching risk-perception
  on bond markets does IMF lending induce moral
  hazard? , Working paper, University of Cologne.
• Zhang, X.A., (1999), Testing for moral hazard in
  emerging markets lending, Institute for
  International Finance, Research Papers, No. 99-1.
• Zwart, S., (2007), The mixed blessing of IMF
  intervention signaling versus liquidity
  support, J. Financial