The Hong Kong Government's Review Report:
An Interpretation and a Response

4 May 1998

Tsang Shu-ki
Department of Economics
Hong Kong Baptist University

1. Report on Financial Market Review (April 1998)

After the speculative attack on the Hong Kong dollar in October 1997, which led to very unpalatable consequences for the stock and property markets as well as the real economy, the Hong Kong government conducted a review exercise on the local financial system. It included the linked exchange rate system (the "link")----Hong Kong's idiosyncratic currency board arrangement (CBA). Proposals were collected from the academics to strengthen its defence.

When consulted, I recommended the modernized CBA of Argentina, Estonia and Lithuania (the AEL model) as an alternative, under which currency arbitrage can be performed electronically without involving cash, because the central bank guarantees the convertibility of the whole monetary base at the fixed exchange rate. Hence the spot exchange rate in the market is fully allied with the official rate (Tsang, 1996a, 1996b, 1997, 1998). While this does not remove all risks from the system, it is in my view preferable to a CBA without an effective arbitrage mechanism, as in the case of Hong Kong.

The Hong Kong government released the findings of the review exercise in its Report on Financial Market Review on 23 April 1998 (FSB, 1998). The full report is down-loadable from the web site www.info.gov.hk/fsb/finance.

The Report did not accept any of all the major proposals from the academics concerning the link, including my recommendation. It did however give specific arguments for the government's position. This short piece provides an interpretation of the Report and a quick response to the reservations that it raised about the AEL model. To make the presentation comprehensible to readers who may not have followed the story in detail, and to render the discussion coherent, I have to give some background information about currency board economics and Hong Kong's exchange rate system. I know that this puts myself in the danger of repeating what I have said elsewhere, not the least on my own web page! To redress, some new metaphors and illustrations are used.

2. CBAs: the "tripod"

An exchange rate can be fixed by: (1) foreign exchange controls and/or government interventions in the market; or (2) arrangements that activate self-interested market forces. In many developed economies, the first method has been the norm, whereas market-driven systems include the old gold standard and currency boards.

A currency board issues cash (notes and coins) with 100% foreign exchange reserves backing at a fixed exchange rate against a designated currency (Schwartz, 1993; Williamson, 1995). This supposedly fosters "economic discipline" in monetary and fiscal policies, which would instill confidence and lead to exchange rate stability.

It is far-fetched to argue that an exchange rate can be "fixed" by discipline-generated confidence alone, without practical mechanisms that bind the exchange rate. Technically, a CBA differs from a fixed rate regime based on market intervention. Like the gold standard, it depends on two "automatic stabilizers" to anchor the exchange rate: (1) specie flow and (2) arbitrage.

Under the specie flow process, an outflow of capital, as a result of doubts about the exchange rate, would lead to a contraction of the money supply. The interest rates then go up, and a counter-flow of funds is induced. The series of event would take place automatically and speedily, so that the exchange rate can be "fixed" without government intervention. Such logic seems a bit shaky. Under normal circumstances, interest rate hikes may contribute towards the stabilization of a currency. But if the exchange rate is itself fluctuating and looks insecure, higher interest rates will not necessarily induce a counterflow of capital. In this sense, the specie flow process is not a reliable mechanism in fixing an exchange rate.

Therefore, there is the need for the second mechanism of the CBA: currency arbitrage (alternatively known as exchange rate arbitrage) that directly binds the exchange rate. Given the board's 100% foreign reserves for cash in circulation, cash arbitrage can be carried out. In case the market exchange rate weakens from the official rate, people can convert their bank deposits into cash, go to the currency board to exchange the cash into foreign currency at the stronger official rate, and then sell the foreign currency in the market. This arbitrage activity will yield a riskless profit, and the selling pressure on the foreign currency will bring the market exchange rate back to the official level.

In general, there are three anchors for a CBA: (1) economic discipline because of the 100% foreign reserves requirement for the issuance of currency; (2) specie flow in the form of interest arbitrage; and (3) exchange rate (cash) arbitrage that binds the spot exchange rate. As illustrated in Figure 1, these three anchors reinforce one another.

In reality, institutional, policy and macroeconomic drawbacks exit in different CBAs and have prevented the effective functioning of particular anchors. While the Hong Kong CBA has scored well regarding the first two anchors, at least in the 1990s, it is still lacking in arbitrage efficiency, even up to now.

Figure 1

The “tripod” for a classical CBA to fix the exchange Rate

economic discipline given

full reserves backing for currency issues

specie flow: interest arbitrage
			currency (cash) arbitrage

3. The reality of Hong Kong's CBA

Hong Kong's CBA has evolved over the years since its inception in October 1983. What is interesting is that it has not functioned in accordance with the theory. There is actually no currency board. While coins are made by the government, notes are issued by a few designated notes issuing banks (NIBs), which alone can deal with the monetary authority at the fixed exchange rate of HK$7.80/US$. Notes-based arbitrage opportunities have therefore been highly restricted, rendering one of the two stabilizers almost totally inoperative. Hence the market exchange rate has strayed from the official rate of 7.80 by an average of about 1%. In the 1990s, the market rate has been on the strong side of 7.80, which is not necessarily a "good" thing for a "fixed" exchange rate regime.

In the initial period (late 1983-1987), neither economic discipline nor specie flow were depended upon, given the shaky economic and political situation. The presumed bank notes arbitrage process also did not work. The linked rate of 7.80 was held imperfectly, thanks to a combination of government intervention in the foreign exchange market, manipulation of interest rates, and administrative measures (including the legal incorporation of "negative interest rates", when the HK dollar faced the speculative pressure of revaluation in 1987-88). See Nugée (1995) for an official admission.

The problem then was that the government could not even define the monetary base as banks did not settle transactions through it. In 1988, the Accounting Arrangements were imposed. That gave the government an indirect handle on the monetary base (notes and coins in circulation plus the clearing balance of the banking system) through the Hong Kong and Shanghai Banking Corporation as the ultimate clearing bank. In the early 1990s, the launching of the Exchange Fund Bills and Notes, as well as the setting up of the "Liquidity Adjustment Facility" (LAF) as a discount window, strengthened the ability of the authority in affecting interbank liquidity in the two-tier system. (Nugée, 1995; Tsang, 1996a, 1996b).

On 1 April 1993, a central bank, the Hong Kong Monetary Authority (HKMA) was formally established by putting all the pieces of reforms under one roof and managed by one powerful institution. Moreover, a parallel development since the 1980s had been the accumulation of huge fiscal and foreign exchange reserves by the government. These developments and evolving mechanisms have enabled the HKMA to modify its stance. Officials were proud to present the link as a currency board system (Latter, 1993: HKMA, 1994). Adequate reserves and economic discipline were emphasized.

Nevertheless, "automaticity" was not on the agenda yet. After its establishment, the HKMA made it known that it would defend the Hong Kong dollar by having flexible ways to manipulate the monetary base and to influence interest rates. Deputy Chief Executive of the HKMA, Andrew Sheng, said on the heel of the Mexican crisis, " .... in recent years the HKMA has introduced various reforms to its monetary management tools, or more aptly, our monetary armoury, to maintain exchange rate stability. ...... As was seen in January (1995), our determination to use the interest rate tool was sufficient to deter further speculation against the HK dollar. In fact, currently, the HK dollar is at a stronger level than it was at 1994 year end." (Sheng, 1995, p.60) "To the extent that the HKMA intervenes through the use of foreign exchange swaps, any increase in the monetary base is fully backed by foreign exchange. We use a whole range of instruments in influencing the level of interbank liquidity to manage interbank interest rates, and consequently, maintain exchange rate stability." (p.61)

However, under the two-tier Accounting Arrangements, the clearing balance of the banks showed wide fluctuations "because banks on the odd occasion miscalculate their own liquidity position. That is why we need and are developing a new RTGS payment system to manage funds flow more efficiently." (Sheng, 1995, p.61)

The RTGS (Real Time Gross Settlement) system was installed in December 1996, replacing the previous two-tier structure. The government could then directly manage the clearing balance of the whole banking system (HKMA, 1995, 1997). The HKMA did not make any pronouncements that the link's mode had undergone any significant changes.

4 Auto-pilot? Developments after the October 1997 attack

In October 1997, the Hong Kong dollar suffered a strong speculative attack, as a result of the "contagion effect" of the East Asian financial turmoil. Doubts were cast on the nature and the robustness of Hong Kong's CBA. A controversy arose concerning whether and to what extent the HKMA did intervene in the markets on 23 October 1997, and was therefore responsible for the unprecedented high interest rates---with overnight interbank rates going up to 280% at one point.

The HKMA later argued that the Authority was just "sitting there passively", allowing the system to go on "auto-pilot". But critics pointed out that the HKMA openly warned banks in the morning of 23 October that those who repeatedly borrowed HK dollars from the LAF would be penalized. This presumably touched off a strong "announcement effect" and banks just scrambled for funds.

In reaction, the Hong Kong government took a major step in defining the link as an automatic CBA. The Chief Executive of the HKMA, Joseph Yam, made an important speech on 3 March 1998 in Japan (Yam, 1998). On 23 April 1998, the Hong Kong government published its report on the October 1997 financial storm: Report on Financial Market Review (FSB, 1998).

The key change was that the HKMA announced a commitment not to actively manage the clearing balance of the banking system to defend the exchange rate. As discussed above, before the October 1997 turbulence, the HKMA did try to influence the clearing balance to stabilize the linked rate. According to the Report (FSB, 1998, paras. 3.36-3.41; Annex 3.5), the HKMA would keep to the following rule of automatic adjustment.

"....... In line with the discipline of the currency board system, the clearing balance will be affected by the flow of funds into or out of the Hong Kong dollar. Specifically, when there is an inflow of funds involving the HKMA passively buying US dollars sold to it by the banks and providing the Hong Kong dollars, the clearing balance of the banking system will rise. ...... Conversely, when there is an outflow of funds involving the HKMA passively selling US dollars and buying Hong Kong dollars from the banks, the clearing balance of the banking system will fall. ......

The HKMA adheres strictly to this discipline which in effect involves the clearing balance of the banking system varying with the amount of US dollars sold to or brought from the HKMA at the initiative of the banks. ....." (original emphases) (FSB, 1998, paras. 3.36-3.37)

Although the government did not use the term: it is the "specie flow" mechanism---one of the three anchors of the classical currency board system. So the HKMA has chosen to abandon the pro-active manipulation of interbank liquidity (the clearing balance) and interbank interest rates as a means of defending the link, and the specie flow process is allowed unfold naturally. This in is a major policy shift from past practices. Nevertheless, the HKMA will maintain the option to sterilize the monetary effect of several types of "exceptional circumstances", including

1. Occasions when Initial Public Offerings (IPOs) of shares and other large scale Hong Kong dollar transactions risk creating extreme conditions in the interbank market;

2. The necessity of entering into intraday Repos and overnight Repos (through the LAF) to "smooth the settlement of interbank transactions";

3. Activities which may have the unintended effects of affecting the clearing balance, such as a transfer of fiscal surpluses by the government to the HKMA.

The HKMA will undertake to neutralize their effects on the clearing balance by recycling or offsetting interbank liquidity through appropriate actions. Note that these are sterilization measures to contain domestic shocks rather than international shocks.

In terms of the "tripod" on which a classical CBA relies in fixing its exchange rate (see Figure 1), the HKMA has arrived at a situation where two of three anchors can be effectively used: (1) economic discipline on the basis of adequate reserves; and (2) automatic specie flow. The problem lies with the third anchor---exchange rate arbitrage.

In the Report on Financial Market Review, the Hong Kong government admitted the implausibility of bank notes arbitrage---a key pillar of the classical CBA---as an effective mechanism to bind the exchange rate (FSB, 1998, para. 3.34). I have tried to drive home this point for quite some time (Tsang, 1997, 1998). To compensate for this deficiency, however, the government did not put in place an alternative arbitrage mechanism, and rejected my proposal (Tsang, 1996a, 1996b, 1997) of adopting the AEL (Argentina, Estonia, Lithuania) model of convertible reserves under which arbitrage can be performed without involving bank notes or cash (FSB, paras. 3.64-3.65).

Instead, the HKMA has opted for a tactic of "constructive ambiguity" (Yam, 1998, p.24), under which it would manipulate a "surprise element" and choose the level of exchange rate at which it intervenes directly in the foreign exchange market. At the end of 1997, Hong Kong's international reserves covered more than six times of the currency in circulation. In fact, including the Land Fund, which was transferred to the management of the HKMA in September 1997, Hong Kong had the third largest foreign exchange reserves in the world, which represented over 40% of HK$M3. Hence, the HKMA can afford discretionary market intervention, under which

 "....the HKMA will need to decide at which particular level to enter the foreign exchange market to support the exchange rate. This involves judgement by the HKMA as to whether or not the circumstances have become abnormal, for example, when there is speculation.......Furthermore, the intervention level may not be exactly at 7.80, although very close to it. For instance, when there are signs of speculative pressure, the HKMA may establish its presence in the foreign exchange market even though the exchange rate is on the strong side of the link. " (FSB, 1998, para. 3.43)

As the Hong Kong regime now stands, instead of the traditional tripod of currency board anchorage (Figure 1), the HKMA is relying on a new tripod, as depicted in Figure 2. It is a unique CBA, with the arbitrage anchor replaced by discretionary foreign exchange market intervention to underpin an officially sanctioned exchange rate.

Figure 2

The tripod for Hong Kong’s CBA in April 1998

specie flow			discretionary exchange rate intervention near 7.80

5. Why did I propose the AEL model?

As far as currency board economics is concerned, a clear lesson for all from Hong Kong is that cash arbitrage is hopeless in defending a CBA in the modern context. Greenwood and Gressel (1988) detected the problem some time ago; Tsang (1996a, 1996b) tried to tackle it, and then gave up (Tsang, 1997, 1998). The Report on Financial Market Review put paid to its practicability in Hong Kong (FSB, 1998, para. 3.34). In a banking system with fractional cash reserves, allowing depositors to covert deposits into bank notes for the sake of doing arbitrage is a very hazardous business.

Is there an alternative? One way out is to modernize the CBA and adopt the "convertible reserves mechanism" of Argentina, Estonia and Lithuania (the AEL model), under which interbank exchange rate arbitrage can be performed without involving cash.

I singled out Argentina, Estonia and Lithuania for detailed investigation (Tsang, 1996a, 1996b), and found that their model (the AEL model) overcame the problems of cash-based arbitrage. These three countries began a currency-board type system in 1991, 1992 and 1994 respectively (Baliño and Enoch, 1997). Though latecomers compared with Hong Kong, their improved arrangement has shown a much higher degree of arbitrage efficiency and exchange rate stability, with the spot exchange rate invariably quoted around the official rate, despite political and economic turbulence.

Under the AEL model, banks have an account with the central bank, in which deposit reserves as well as other balances are kept. The central bank guarantees the full convertibility of these bank balances, at the fixed exchange rate. This setup bypasses the problem of moving cash around for arbitrage.

Assume that we are in a country adopting the model and the domestic currency "peso" is pegged to the US dollar at parity. Banks are forced to quote the official rate of 1 peso per US dollar. If Bank X deviates by quoting the exchange rate of 1.1 peso, any Bank Y can sell US$1 million to it for 1.1 million pesos, asking X to transfer the pesos to Y's account at the central bank. At the same time, Y would of course transfer US$1 million to X's account there. On demand, the central bank would convert the pesos into US$1.1 million for Bank Y, which then obtains an arbitrage profit of US$100,000. Bank X, on the other hand, suffers a loss of 100,000 pesos because its US$1 million at the central bank can only be exchanged into 1 million pesos. If it still does not surrender, it would be vulnerable to much greater losses in the interbank market.

No cash movements are involved, as the central bank plays the role of clearing the arbitrage transactions between the two banks. After settlement, the central bank's foreign reserves will be reduced by US$100,000. In other words, the central bank risks losing reserves if a commercial bank like X rebels against the CBA. Nevertheless, that loss is matched by a correspondent shrinkage (100,000 pesos) in Bank X's balance sheet. Since the deal is sealed and settled by telephone calls and electronic means, the transaction cost is reduced to a minimum.

In reality, under the convertible reserves system, no banks would dare to deviate in exchange rate quotations. All banks are bound by the rule of the game to quote the official exchange rate, within a very narrow buying and selling spread that truly reflects petty transaction cost, or they will be hit by their market rivals. Hence no actual arbitrage needs to take place, and the central bank will not suffer any loss in reserves. With this improved form of CBA, Argentina, Estonia and Lithuania have been able to literally fix their spot exchange rates despite serious economic or political instability (Tsang, 1996a, 1996b).

Hong Kong can adopt the convertible reserves system to rein in the market exchange rate. With the RTGS system in place, the HKMA may impose a deposit reserve requirement on the banks. To overcome possible resistance from the banking sector, the ratio, which could be interpreted as a "financial tax", should be as small as possible, and near-market interest rates need to be paid on those reserve deposits. The idea is not to tax the banks, but to ensure that there is suitable liquidity in the reserve account to minimize any possible impact on the interest rate, perhaps in the initial stage. As explained above, if the system works, no actual arbitrage occurs and the spot exchange rate can still be fixed. In that situation of benign equilibrium, the deposit reserve ratio might be very small, even approaching zero. A practical suggestion that I put to the local Chinese press was to start with a modest 5% ratio (which would already yield reserves much greater than the average of the existing clearing balance) and then revise it downward quickly in the light of the evolving situation.

With huge reserves, the HKMA should have little problem in settling electronic arbitrage among banks at the fixed exchange rate. Adopting the AEL model requires an even lower degree of activism on the part of the HKMA. The Hong Kong economy should face less unpalatable results if the speculators return, as they can see that all the banks are bound by self-interest to quote around the rate of 7.80. In other words, speculators have to fight the whole banking system, rather than just the HKMA.

I am afraid that not many people appreciate this point, even up to now. Moreover, I have to emphasize that my proposal of the AEL model is based entirely on the consideration of improving the technical robustness of a CBA. Economic optimality of currency boards, and indeed of the fixed exchange rate system as a whole, is a different issue. Interesting as it may be, it is not the subject matter of this article.

6. The Hong Kong Government's Reservations about the AEL model

I proposed the AEL model in the consultation exercise of the Hong Kong government after the October 1997 turbulence. The government acknowledged but did not accept the proposal in its Report on Financial Market Review, as it regarded the distinction between the monetary arrangements in Hong Kong and those under the AEL model as "relatively minor"---a point on which it did not elaborate (FSB, 1998, para.3.65). All relevant sections of the Report are extracted in the Appendix to this article. In short, a number of concerns was raised (FSB, 1998, para. 3.44 and para. 3.65):

1. The AEL model "will not protect the economy from the interest rate pain", as a result of speculative attack. Argentina in 1995 was cited as an example of dwindling reserves and high interest rates. Moreover, "since there is no scope for the exchange rate to move, the impact of the flow of funds will fall entirely upon interest rates", thus leading to greater interest rate volatility.

2. There are "transitional problems of moving the exchange rate from the present (strong) level to 7.80".

3. The "proposed statutory reserve requirement will be very unpopular among banks". Moreover, the LAF mechanism has already provided a cushion to sharp interest rate movements.

7. My response

These legitimate concerns should be analyzed from the appropriate perspective. Of course, when there is an exodus of funds, the government's reserves will contract, even if the exchange rate remains fixed. Moreover, the AEL model cannot eliminate the "interest rate pain". Which system can? Although the spot exchange rate is stable, the forward rate may not necessarily be so if people do not have sufficient confidence in the system. According to the arbitrage equilibrium equation, local interest rates could still be higher than those of the foreign counterpart, co-existing with weak forward exchange rates for the domestic currency. Nevertheless, such imperfections have logical roots, and it is important to distinguish between two different types of perceived risk: (1) efficiency risk; and (2) systemic risk.

One risk is that market participants are not sure whether the CBA could really "fix" the spot exchange rate. In other words, there is an "efficiency risk" regarding the exchange rate and they demand an interest rate risk premium, inflicting the "interest rate pain". In an international financial centre like Hong Kong, the efficiency risk cannot be under-estimated as there is no recourse to any form of exchange controls. Anyway, the fixity of the spot exchange rate under an effective arbitrage mechanism over an extended period should lead to a consolidation of confidence, and convergence in domestic-foreign interest rates and spot-forward exchange rates would take place as people dare to engage in interest arbitrage. In the case of the AEL model, such a phenomenon has occurred in Argentina, Estonia and Lithuania, as analyzed by Baliño and Enoch (1997, Appendix I).

 The convergence process has however not been perfect in the three countries: local interest rates have still been higher than those of the US dollar (to which the Argentine peso and the Lithuanian litas are pegged) and the German Mark (to which the Estonian kroon is linked). This is due to the existence of "systemic risk". Although market participants observe the fixity of the spot exchange rate, they are not sure that such a "perfect" system that is working so well will not be abandoned in the future, not because it is defective, but as a result of other economic or political factors. No matter how good it is in anchoring the exchange rate, whether a fixed rate regime is optimal for the economy is always a controversial issue.

Analysts familiar with the situations in the three "AEL" countries understand why some of their people might be nervous, justifiably or otherwise, about the possibility of political instability. Argentina in 1995 (in the aftermath of the Mexican crisis and in the midst of presidential election politics) must be a "perfect" counter-example for the Hong Kong government to cite: it was a time when the systemic risk was so huge. I do not want to appear rhetorical: but the question can surely be reversed: without the AEL model effectively fixing the exchange rate of the peso against the US$ at parity, would not the Argentine situation have been even more "disastrous"?

If Hong Kong adopts the AEL model, the efficiency risk should be eliminated rather quickly. Interest rate convergence could unfold at a faster pace than that in those three countries. As to systemic risk, Hong Kong's political and economic situations are far more stable, although re-pegging or re-floating out of optimality considerations can never be ruled out.

Let me return to the specifics of applying the AEL model to Hong Kong. It has already been pointed out that the required reserve ratio could be set at low levels and banks should be paid near-market interest rates, so as to reduce their resistance. With the spot exchange rate locked at 7.80 and quoted by all banks, speculators will have to think very carefully before launching any attack, as they will be playing game with the whole banking system, not just the HKMA. Such hesitation should significantly reduce the pressure on the link, and hence on interest rates through the automatic specie flow process.

The present deviation of the market exchange rate on the strong side of 7.80 is seen by the HKMA as providing a "scope of adjustment" on top of interest rate movements. Nonetheless, it is small in magnitude (less than 1%) and it is not clear whether it generates a net benefit. Alongside with "constructive ambiguity" (Yam, 1998, p.24) or "the surprise element" (FSB, 1998, para. 3.44), the deviation may actually introduce uncertainty into the system. Whenever the market rate breaches 7.75 now, people will, justifiably or otherwise, regard it as a sign of "weakness" and expect the HKMA to intervene. There is a danger for 7.75 to become a self-imposed defence line if the authority is excessively concerned about the "transitional problems" of moving the market rate to 7.80. Any "surprise element" will then be lost.

The non-alliance of the market exchange rate with the official rate could also have been the source of the problem: in the heat of the East Asian crisis, speculators might have regarded the "non-fixity" as a sign of "insecurity" of the link, and therefore decided to have a go at it. If there had been no such deviations, because of say a more robust arbitrage mechanism, speculators might not have come, or might not have been so aggressive.

The key question is whether the combination of automatic specie flow and discretionary foreign exchange market intervention (as depicted in Figure 2) constitutes an effective defence of the Hong Kong CBA. A related issue is what the least-cost option is.

I think that these are still open questions that economists can debate about.

REFERENCES

Baliño, Tomas and Enoch, Charles (1997) "Currency Board Arrangements: Issues and Experiences," IMF Occasional Paper, No. 151, August.

Financial Services Bureau (FSB) (1998), Report on Financial Market Review, Hong Kong Government.

Greenwood, John, and Daniel Gressel (1988) "How to Tighten the Linked Rate Mechanism," Asian Monetary Monitor, January-February, pp.2-13.

Hong Kong Monetary Authority (HKMA) (1994) The Practice of Central Banking in Hong Kong.

Hong Kong Monetary Authority (HKMA) (1995) "Hong Kong's Payment System," Quarterly Bulletin, August, pp.1-6.

Hong Kong Monetary Authority (HKMA) (1997) "Hong Kong's Real Time Gross Settlement System," Quarterly Bulletin, February, pp.30-37.

Latter, Anthony (1993) "The Currency Board Approach to Monetary Policy-- from Africa to Argentina and Estonia, via Hong Kong," in Hong Kong Monetary Authority, Monetary Management in Hong Kong, proceedings of the Seminar on Monetary Management, pp.26-43.

Nugée, John (1995) "A Brief History of the Exchange Fund," Quarterly Bulletin, Hong Kong Monetary Authority, May, pp.1-17.

Schwartz, Anna J. (1993) "Currency Boards: Their Past, Present and Possible Future Role," Carnegie-Rochester Conference on Public Policy, 39, pp.147-193.

Sheng, Andrew (1995) "The Linked Exchange Rate System: Review and Prospects", Quarterly Bulletin, Hong Kong Monetary Authority, May, pp.54-61.

Tsang, Shu- ki (1984) "On the Cash-based Fixed Exchange Rate System," in The Pearl in the Mouth of the Dragon: Collected Essays (in Chinese), Hong Kong: Wide Angle Press, pp.179-201.

Tsang, Shu-ki (1996a) "The Linked Rate System: through 1997 and into the 21st Century," in Ngaw Mee-kau and Li Si-ming (ed.), The Other Hong Kong Report 1996, Hong Kong: The Chinese University Press, chapter 11.

Tsang, Shu-ki (1996b) A Study of the Linked Exchange Rate System and Policy Options for Hong Kong, a report commissioned by the Hong Kong Policy Research Institute, October.

Tsang Shu-ki (1997) "Currency Board the Answer to Rate Stability," Hong Kong Standard, 31 October 1997.

Tsang Shu-ki (1998) "The Case for Adopting the Convertible Reserves System in Hong Kong," Pacific Economic Review, forthcoming.

Williamson, John (1995) What Role for Currency Boards? US: Institute for International Economics.

Yam, Joseph (1998), "Hong Kong: Financing Asia's Development," Keynote Address, Hong Kong Development Council Financial Roadshow in Tokyo, 3 March.

Appendix: Extracts from the Hong Kong Government's Report on Financial Market Review (April 1998)

"Exchange Rate Level

3.42 Rather than being entirely transparent and passive by always adhering to the fixed level of the exchange rate (7.80) at which it buys or sells US dollars against Hong Kong dollars, the HKMA under normal circumstances leaves the foreign exchange market pretty well alone. This approach is considered appropriate as it allows the foreign exchange market as much freedom as is desirable without interference from the HKMA.

3.43 The consequence of adopting such an approach is that the HKMA will need to decide at which particular level to enter the foreign exchange market to support the exchange rate. This involves judgment by the HKMA as to whether or not the circumstances have become abnormal, for example, when there is speculation, although in practice the judgment is never difficult to make. Furthermore, the intervention level may not be exactly at 7.80, although very close to it. For instance, when there are signs of speculative pressure, the HKMA may establish its presence in the foreign exchange market even though the exchange rate is still on the strong side of the link.

3.44 An important issue to consider is whether the HKMA should continue the present mode of operation or should instead be entirely passive and undertake to buy and sell US dollars at the fixed exchange rate of 7.80. We believe a change to the latter would entail the following downside risks:

(a) if the exchange rate is fixed at a single level, the foreign exchange market involving the Hong Kong dollar will largely disappear as trading among banks would be displaced by trading with the HKMA;

(b) since there is no scope for the exchange rate to move, the impact of the flows of funds will fall entirely upon interest rates, even though these flows are not related to speculative activities. The resultant greater volatility in interest rates will have important ramifications for financial and other economic activities;

(c) greater transparency will mean giving up the surprise element which is often quite helpful in managing markets effectively as transparency and predictability often make it easy for speculators; and

(d) a sudden move to the 7.80 level engineered by the HKMA may be misinterpreted by the market as a weakening of resolve on the part of the Government to accept the interest rate pain.

3.45 In relation to the problems set out in (a) and (b) above, a possible alternative is to allow the exchange rate to move within a pre-determined band rather than fixing it at a single level. Once the exchange rate hits the limits defined by the "band", the HKMA would trigger the auto-pilot adjustment mechanism by passively buying and selling US dollars in the foreign exchange market and allow the full money market effects to be felt. However, a significant drawback of this system is that the exchange rate band, regardless of whether it is presented in the form of a spread or a commission rate, would provide clear target points for speculators to attack. The present practice provides no clear target for speculators.

..............

AEL (Argentina, Estonia and Lithuania) model

3.64 In contrast to the US$ LAF/HK$ put option scheme which aims at modifying the currency board arrangements to ameliorate the interest rate pain, Professor Tsang Shu-ki of Baptist University suggests that Hong Kong should follow the classical currency board system as practiced in Argentina, Estonia and Lithuania (see Annex 3.11 for details). Specifically:

 (a) Professor Tsang proposes that the convertibility at the fixed rate of 7.80 should be extended from banknotes to the entire monetary base. In other words, the exchange rate at which the HKMA undertakes to buy and sell US dollars against the Hong Kong dollar clearing balances of licensed banks will be entirely transparent, passive and be fixed at 7.80; and

(b) as the aggregate clearing balance maintained by licensed banks with the HKMA is very small, given in particular an efficient real time interbank payment system, even small flows of funds into or out of Hong Kong dollar, which are likely all to be conducted with the HKMA as there will be no deviation in the exchange rate from 7.80, would cause sharp volatility in interest rates if the measure set out in (a) is implemented (see also paragraph 3.44). Recognizing this, Professor Tsang also proposes creating a cushion for the management of interbank liquidity, by the introduction of statutory reserve requirements with averaging provisions over a period of time.

3.65 The HKMA's view, which is supported by Professor Goodhart and experts in the IMF, is that the distinction between monetary arrangements in Hong Kong and the "classical" currency board system are relatively minor (see, for example, the letter from Mr. David Goldsbrough of IMF at Annex 3.9). In relation to the two measures suggested above, the following are worthy of note:

(a) the proposed arrangements will not protect the economy from the interest rate pain - as evidenced in the case of Argentina in 1995, the AEL model would not protect the local currency from speculative attack; nor would it minimize the economy from the pain of high interest rates when there is a large-scale outflow. Foreign currency reserves in Argentina dropped by 43% from around US$14 bn in December 1994 to a low of US$8 bn by the end of March 1995. Money supply contracted by more than 13% during this period. Short-term interbank interest rates shot up to 70% in early March and the prime rate almost doubled from 12.4% to 23%;

(b) transitional problems in moving the exchange rate from the present level to 7.80 should not be underestimated - see paragraph 3.44 for a detailed discussion on this issue;

(c) the proposed statutory reserve requirement will be very unpopular among banks - the statutory reserve requirement amounts to a tax on the banking system (unless market interest rate is paid on the reserves). In recent years, a number of economies (e.g. Switzerland and Canada) have abolished such a requirement. In some other economies, it has been retained mainly for the purpose of controlling the money supply. However, under a fixed exchange rate system, money supply is endogenously determined by balance of payments situation, rather than controlled by the central bank; and

(d) the LAF has already provided a cushion to prevent sharp interest rate movements - the concern over an overshooting of interest rate due to the small size of the aggregate clearing balance has been addressed by the provision of liquidity through the LAF. Presently, the outstanding amount of Exchange Fund paper and other eligible private sector debt securities amounted to HK$101 bn and HK$100 bn respectively. About 70% of these papers are held by the banking sector. With the clarification on the definition of "repeated borrowings" in the use of LAF, uncertainty regarding the access to LAF has been removed.

..............."