Introduction
- The use of financial market indicators by authorities has been rising steadily in recent years (see for example Mylonas and Schinch, 1999). This is particularly the case in countries, which have adopted inflation target (IT) regimes as their monetary policy framework. The forward-looking orientation of IT coupled with the fact that the monetary transmission process is becoming more complex, makes data extracted from financial markets especially valuable.
- In Israel such data play an important role in the monthly monetary policy decisions. Thus, market-based inflation expectations derived respectively from nominal and CPI-indexed bonds, extending over various maturities, are an important input in monetary policy decisions.
- Because Israel is a small open economy the exchange rate is an important part of the monetary transmission process, and information about the distribution probabilities of the exchange rate can be very useful for policy purposes. FX currency options are a natural source of such information.
The purpose of this paper is to describe how the Bank of Israel (BOI) extracts and uses information from two types of FX currency options. The first type is FX options offered in weekly auctions by the BOI. Since 1993 the BOI has offered At-The-Money-Forward (ATMF) options for three and six months respectively, and since the end of 1998 it also offers ATMF put options for three months. Since these options have no intrinsic value, their price reflects only the uncertainty regarding the FX rate.
The amount offered in these auctions is constant, 25 million dollars in notional value per week.
The weekly auctions for the three months call and put are 10 million dollars and these auctions take place twice a week. The six-month call option is auctioned off once week and its notional value is 5 million dollars per week.
In fact the BOI started these auctions as early as 1989 as part of a financial market nurturing strategy. At that time FX options were not used in Israel and the BOI options were the ‘only player in town.’ Today these options are a very small part of the FX options offered by banks and others. Although the parameters of the BOI options have been changed since then (strike prices, maturities etc.), the principle of maintaining a fixed supply, regardless of macro and exchange-rate developments, has not changed.
The second type of FX options is those traded daily on the Tel-Aviv Stock Exchange (TASE). These are European call and put options with maturities of up to six months. In section 1 we describe the usefulness of the information extracted from the BOI options for the assessment of the monetary policy stance. In section 2 the same is done for the TASE options. Section 3 combines the information from the two types of auction in order to point to a conflict between inflation targeting and FX intervention. Section 4 provides brief concluding remarks.
1) The Value of the Information Extracted from the BOI Options
Monetary policy in Israel aims at attaining the inflation target set by the government, which is 2–3 percent for 2002 and 1–3 percent from 2003 onwards, a range which is defined as price stability. The monthly interest-rate decisions by the BOI are based on its assessment of inflation one and two years ahead. This assessment is based on the Bank’s own model (Elkayam, 2001) and is complemented by information on inflation expectations derived from the bond market (regular and CP-indexed), as well as on data from professional forecasters (Figure 1). Future exchange -rate developments obviously play an important part in the assessment of future inflation. The econometric model provides information on the interaction between the exchange rate and inflation (the passthrough) but does not say anything about the uncertainty which surrounds future exchange- rate developments. In this respect the implied volatilities derived from the BOI options are useful in three important aspects. First, following their development over time (Figure 2) may provide answers to such questions as what is the probability of a 10% devaluation in the next three months and how does this probability evolve over time (Figure 3). Second, the ratio of the implied volatilities of the six to three month options, and its evolution over time, is an indicator of the term structure profile of FX uncertainty (Figure 4). Third, comparing the implied and historical volatilities may be useful in detecting peso-type problems (Figure 5).
Combining this information against, say, the backdrop of the government’s difficulties in maintaining fiscal discipline may bias the BOI’s decision towards monetary tightening even if future inflation is not expected to deviate from the inflation target.
2) The Importance of the Information Extracted from the Option Traded on the TASE
The implied volatilities extracted from the BOI options are based on the Black and Scholes (1973) formula, which assumes that changes in the exchange rate are normally distributed. It is very likely, however, that this assumption is invalid in the Israeli context. Israel has still an official exchange- rate band and although it is over 40 percent wide at present, for most of the last decade the exchange rate has been close to the lower limit of the band (appreciation, see below). Thus, it stands to reason that not only is the distribution asymmetric but that asymmetry itself changes over time. Detecting such changes, particularly in periods of large local-currency depreciation (appreciation), may provide the central bank with important insights. It may, for example be a useful indicator as to whether the behavior of participants in the FX market is stabilizing or destabilizing.
Such information may be derived from the two months to maturity options ‘smiles’ which are traded on TASE. For example, Figures 6 and 7 show respectively the ‘smiles’ before and after the events of September 11 and a recent sharp unexpected reduction of the key interest rate by the BOI (see also Figure 7a, 7b). The fact that in both cases the smile tilts in the direction of local-currency appreciation (large appreciations are more likely than large depreciations, following a sizeable depreciation) indicates stabilizing behavior by FX market players. This information is obviously valuable to a central bank.
3) Using FX Options Information to Detect Problems in the Monetary Policy Framework
In addition to the types of information described above, information derived from FX options can also be used to detect inconsistent and unsustainable policy frameworks. Israel is a case in point in this respect. Since 1992 Israel has adopted inflation targeting as well as a crawling band - rate regime. Although the band is very wide and getting wider over time (the slope of the upper limit is six percent annually, while that of the lower one was recently reduced from two percent to zero) and is currently over 45 percent wide, it continues to be a potential problem for monetary policy. There are two reasons for this: a) Throughout most of the last decade the exchange rate has been very near to and at times right on the lower (appreciating) limit of the band. On those occasions the BOI has had to intervene to defend the lower limit by sterilized intervention. Such interventions reduce the effectiveness of monetary policy. b) Although the official upper limit of the band has been far above its lower limit, the effective upper limit, as perceived by FX market participants, is much lower. In fact, because the public is aware of the passthrough from the exchange rate to prices, the more credible IT becomes the narrower is the effective band. As the effective band becomes narrower (measured by the distance between the effective upper limit and the official lower one), the probability that the BOI will again have to engage in FX sterilized intervention to defend the lower limit grows. This in turn uthe effectiveness of monetary policy, pointing to the inconsistency of the IT regime and a crawling band.
First, we use the TASE options to construct the effective upper limit. Second, we use the premium of the BOI options to measure exchange-rate uncertainty. Third, FX uncertainty is shown to be perversely related to the BOI key interest rate and the width of the effective band, when the exchange rate is on the lower edge. These perverse results are evidence of inconsistency in the monetary policy framework.
To construct the ‘effective’ upper limit we used FX currency options traded on the TASE. On any given trading day we searched for an out-of-the-money call option with the highest available strike price for which a minimal positive premium has been paid2) and which has at least five transactions. We mark the exchange rate corresponding to the strike price of this option as a point on the effective upper limit. This procedure is repeated for each trading day, from February 1996 to June 2001. The result is an effective upper limit, which is depicted in Figure 8.
__________________
2) We now run the following simple regression:
The formula for deciding on the minimum
where Premium = P and Strike Price = K
PRM is the premium on the six months BOI ATMF call option. As stated above, this variable represents exchange-rate uncertainty.
KR is the BOI key rate.
VK is the width of the effective band.
This regression was run for the period from February 1996 to June 2001, and for two sub-periods, July 1997 to June 2000, when the exchange rate was near but above the lower limit, and February 1996 to June 1997, when the exchange rate adhered to the lower limit (for much of the period).
The results, presented in the table are interesting in several respects. First, we find that a rise in the key rate results in greater FX uncertainty. This is true when the exchange rate was inside the band (first row, second column) and particularly when the exchange rate was on the lower limit of the band (second row second column).
* These regressions, for the total period and period I, are estimated using the RA procedure. The coefficients of AR were 0.57 and 0.6 respectively. This procedure was applied because of high serial correlation in the OLS version. The numbers in parentheses are t-values.
Concluding Remarks
Information derived from FX options have the potential to provide central banks with timely forward-looking information about the expected volatility of the exchange rate.
In addition, information extracted from FX options also points to inconsistencies in the monetary policy framework. Evidence on this point was provided for Israel, where inflation targeting and a crawling exchange-rate band coexist unpeacefully. The Israeli experience may, however, also be relevant for other IT countries which have no official band but which intervene in the FX market. If FX options exist in those countries their story might be that even an implicit exchange -rate band leads to conflict, as is the case in Israel.
Bibliography
Black, F. and Scholes, M.S. “The Pricing of Options and Corporate Liabilities”, Journal of Political Economy, vol81, 1973, 637-654.
Brenner, M. and Sokoler, M. “Inflation Targeting and Exchange-Rate Regimes; Evidence From the Financial Markets”, Unpublished Manuscript, November 2001.
Elkayam David, ‘A Model for Monetary Policy Under Inflation Targeting: The Case of Israel”, Bank of Israel Working Paper, July 2001.j
Mylonas, P. and Schinch, S. “The Use of Financial Market Indicators by Monetary Authorities”, Financial Market Trends, vol. 74, October 1999, 161-189.