مداخله بانک مرکزی و روند معاملات طول روز در شکل گیری قیمت در بازار ارز

We propose a novel theory of the impact of sterilized spot interventions on the microstructure of currency markets that focuses on their liquidity. We analyze the effectiveness of intervention operations in a model of sequential trading in which i) a rational Central Bank faces a trade-off between policy motives and wealth maximization; ii) currency dealers' sole objective is to provide immediacy at a cost while maintaining a driftless expected foreign currency position; and iii) adverse selection, inventory, signaling, and portfolio balance considerations are absent by assumption. In this setting, and consistent with available empirical evidence, we find that i) the mere likelihood of a future intervention—even if expected, non-secret, and uninformative—is sufficient to generate endogenous effects on exchange rate levels, to increase exchange rate volatility, and to impact bid-ask spreads; and ii) these effects are exacerbated by the intensity of dealership competition, the extent of the Central Bank's policy trade-off, and the credibility of its threat of future actions.

The foreign exchange (forex) market is one of the most active financial markets in terms of volume, frequency, and intensity of trading (e.g., Bank for International Settlements, 2008). The recent availability of high-frequency data has stimulated significant interest in its microstructure. However, with few notable exceptions (e.g., Bossaerts and Hillion, 1991, Naranjo and Nimalendran, 2000, Lyons, 2001 and Evans and Lyons, 2005), the empirical analysis of this data has not been preceded by theoretical investigations of the potential impact of the many institutional features specific to the forex market on its functioning. This paper contributes to closing this gap. We focus on one of those features—the presence of a rational, but not necessarily profit-maximizing Central Bank (CB)—and derive from its inclusion novel implications for the process of price formation in currency markets. Two sets of observations about CB interventions guide our effort. First, many macroeconomics textbooks describe the exchange rate as an intermediate target of monetary policy: CBs choose levels (or bands of fluctuation) for the domestic currency compatibly with the “ultimate” trade-off of monetary policy, between sustainable economic growth (the “output gap”) and moderate inflation (e.g., Lewis, 1995 and Taylor, 1995). CBs may also serve less stringent agendas, under pressure from political power, interest lobbies, etc. In both cases, their actions may not be motivated by pure profit. There is anecdotal and empirical evidence that policy objectives and wealth maximization often collide ( Taylor, 1982 and Neely, 2000). CBs may also take positions in the currency markets for purely speculative motives, as in the frequently cited example of Bank Negara, the Malaysian CB, in the early 1990s (e.g., Brown, 2001). This potential (and often ignored) trade-off between policy goals and costs (or profits) may affect the impact of CB interventions on exchange rate dynamics and on the liquidity of the forex market. Second, while a consensus has emerged in the economic literature (e.g., Adams and Henderson, 1983) that unsterilized interventions influence the exchange rate through the traditional channels of monetary policy, both the effectiveness of sterilized interventions—i.e., those accompanied by offsetting actions on the domestic monetary base—and their impact on forex market liquidity remain controversial ( Sarno and Taylor, 2001). Within the macroeconomic approach, sterilized interventions may affect the exchange rate through either of two channels, portfolio balance and signaling. According to the portfolio channel ( Branson, 1983 and Branson, 1984), interventions altering the relative supply of foreign assets influence the exchange rate when domestic and foreign assets are imperfect substitutes. According to the second channel ( Mussa, 1981, Bhattacharya and Weller, 1997 and Vitale, 1999), interventions influence the exchange rate by conveying information on policy intentions and/or macroeconomic fundamentals. Many empirical studies of the portfolio channel (e.g., Edison, 1993, Payne and Vitale, 2003 and Pasquariello, 2007) find its effects on exchange rates either small and short-lived (particularly in the 1970s and 1980s) or economically and statistically insignificant, despite the important imperfect substitutability documented by Evans and Lyons (2005). There is stronger supporting evidence for the signaling channel, especially when interventions are secret and unannounced ( Dominguez, 1992, Kaminsky and Lewis, 1996, Payne and Vitale, 2003 and Pasquariello, 2007). Extant studies suggest that, in those circumstances, the resulting dissipation of information and adverse selection may increase exchange rate volatility and widen bid-ask spreads ( Vitale, 1999, Naranjo and Nimalendran, 2000 and Chari, 2007). Yet, this type of intervention is rather infrequent (e.g., Dominguez, 1998, Dominguez, 2003 and Fischer and Zurlinden, 1999). Recent empirical research also reports that even expected, non-secret, and announced interventions have large effects on currency returns, return volatility, bid-ask spreads, and trading intensity, albeit often inconsistent with those predicated by information theory ( Dominguez, 1998, Dominguez, 2003, Dominguez, 2006, Payne and Vitale, 2003 and Pasquariello, 2007). Motivated by these considerations, in this paper we propose an alternative theory of the impact of CB interventions on the intraday process of price formation in the currency markets that focuses on their liquidity. This theory illustrates that both the temporary and persistent impact of CB interventions on prices, volatility, and transaction costs in the presence of imperfect substitutability may be related to the special role of forex dealers as liquidity providers. To concentrate on this role, we construct a stylized currency market in which trading occurs sequentially. The market is populated by a continuum of risk-averse investors and an occasionally active CB facing a trade-off between policy and wealth-preservation motives—i.e., accounting for the expected cost of its pursuit of a short-term target level for the exchange rate (e.g., Bhattacharya and Weller, 1997 and Vitale, 1999). The likelihood of each order arrival to be from any of the investors or the CB is exogenous, but their trades are endogenously determined in equilibrium. CB interventions are also non-secret, sterilized, uninformative, and unrelated to the relative supply of foreign assets. These assumptions rule out portfolio balance and signaling effects in our framework. Finally, we model forex dealers as risk-neutral, pure market makers providing immediacy, in the spirit of Garman, 1976 and Brock and Kleidon, 1992, and Saar, 2000a and Saar, 2000b. In particular, we impose that, at each round of trading, they stand ready to buy and sell by setting quotes and spreads that maximize the expected instantaneous compensation for their services they can extract from the order flow while maintaining a driftless expected position in the foreign currency. Hence, our forex dealers are indifferent to the value of the exchange rate at the end of the economy (the investors' concern), given their sole objective of facilitating trading. Our theory generates novel macro and micro implications of the presence of an active monetary authority in the currency markets. To begin with, we show that the mere possibility of a CB intervention is sufficient to induce forex dealers to revise the quotes they post and the spread they charge to risk-averse investors. Intuitively, the positive likelihood of future CB intervention provides dealers with both an opportunity and a threat. Potential CB trades are additional order flow from which to extract rents. They also represent additional liquidity demand to satisfy, possibly creating an imbalance in the dealers' expected position in the foreign currency. At each round of trading, the dealers accommodate this potential pressure, while still profiting from their role of liquidity providers, by revising their bid and ask quotes in the direction of the intervention. The revision serves a dual purpose: to elicit the investors to take the other side of the CB trade, and to reduce the expected magnitude of potential CB trades by pushing the foreign currency closer to its policy target. In equilibrium, CB intervention—albeit non-secret and uninformative—not only is effective in influencing the level of the exchange rate but also increases its unconditional volatility. These effects are either temporary or persistent depending exclusively on the likelihood of future CB actions. We further show that sign and magnitude of the impact of CB trades on bid-ask spreads, as well as on equilibrium prices and price volatility, are crucially related to the degree of dealership competition in the forex market. A monopolist dealer is able to adjust his quotes symmetrically in the presence of an active CB, hence preserving his marginal expected rent from investors' order flow (i.e., the spread), since he is the sole provider of liquidity in the economy. Perfectly competitive dealers instead have to transfer any additional expected positive or negative rent from the potential presence of a CB onto the population of investors. In equilibrium, the ensuing more significant, and often asymmetric revisions in quotes translate into either wider or tighter spreads (depending on the existing demand for foreign assets), greater effectiveness of CB intervention, and higher unconditional price volatility. The relationship between the extent of dealers' market power and the effectiveness of CB interventions has never been explored by the economic and financial literature, and is consistent with recent empirical evidence on the importance of dealership competition for forex bid-ask spreads (Huang and Masulis, 1999 and Pasquariello, 2007). The paper is organized as follows. Section 2 describes our basic economy. We derive its equilibrium in the presence of a CB in Section 3. Section 4 concludes. Proofs are in the Appendix.

Despite much careful work in the economic and financial literature, there remains an intense theoretical and empirical debate on the channels through which sterilized Central Bank (CB) interventions may affect not only the level of exchange rates—e.g., portfolio balance or signaling—but also the microstructure of the markets where they are traded—e.g., adverse selection or interdealer trading. For instance, mounting evidence suggests that, although the portfolio balance channel performs poorly in direct tests, there is significant imperfect substitutability in the currency markets and even expected, non-secret, or uninformative interventions significantly influence daily and intraday currency prices, price volatility, and bid-ask spreads (e.g., Dominguez, 2003, Evans and Lyons, 2005, Chari, 2007 and Pasquariello, 2007). Our study contributes to this debate by proposing a novel theory that links the impact of those interventions on the process of price formation in the forex market to the important role of forex dealers as providers of liquidity. We develop a model of sequential trading in which i) prices are set by either a monopolist or competitive risk-neutral dealers; ii) those dealers' main activity is to dispense immediacy to all market participants at a cost (the bid-ask spread); iii) the demand schedule of the monetary authority results endogenously from the optimal resolution of a trade-off between policy and wealth-preservation; and iv) adverse selection, inventory, signaling, and macro and micro portfolio balance considerations are ruled out by construction. In this setting, we show that the mere likelihood of a non-secret and uninformative CB intervention is sufficient to push the equilibrium exchange rate closer to the CB's policy target, to increase its unconditional volatility, and to induce revisions in the quoted bid-ask spreads—for as long as such threat is credible and the more so the less is forex dealers' market power. These implications are consistent with the aforementioned empirical evidence on the microstructure of currency markets, as well as with anecdotal evidence on the rapid cycles of success and failure of several interventions by monetary authorities around the world over the last three decades.