Welcome on board, Hong Kong!

The Hong Kong SAR Government finally decided to board the ship of the AEL model, which I had persistently recommended for more than a year (see Tsang, 1997 and other pieces of mine on this web site). The Government actually rejected my proposal in its April 1998 Review Report (FSB, 1998), for reasons that I did not find convincing (Tsang, 1998a). Four months later, after the historical intervention in the stock and futures markets (Tsang, 1998d), the government changed its mind.

On 5 September 1998, the Hong Kong Monetary Authority (HKMA) announced seven "technical measures" to strengthen the linked exchange rate system (HKMA, 1998). Information is down-loadable from www.info.gov.hk/hkma. These measures can actually be grouped into two categories: (1) extending the scope of convertibility (at the fixed exchange rate) from the cash base to the whole monetary base; and (2) the replacement of the Liquidity Adjustment Facility (LAF) by a Discount Window.

The first is what the modern currency board arrangement (CBA) of the AEL (Argentina, Estonia and Lithuania) model is all about (for a quick summary, see Tsang 1997; for a more elaborate treatise, see Tsang, 1998a). The second move is more controversial. Many would agree that a modern CBA should provide lender of last resort (LOLR) facility to prevent banking crisis arising from an unpredictable liquidity strain (Caprio, Jr. et. al, 1996; Enoch and Gulde, 1997; Santiprabhob, 1997), but a regular discount window could be regarded as normal central banking in disguise. After all, changing the discount rate is one of the most common "transmission mechanisms" through which central banks implement "monetary policy". As the most rigid fixed exchange rate regime now existing, a CBA is not supposed to have any monetary policy. In any case, in most central banking systems, the discount window is also used to perform the role of the LOLR. Sometimes it is difficult to separate the "policy" function from the "rescue" function.

In Hong Kong's case, idiosyncrasy still exists on both facets. This short piece provides a quick analysis.

The "Convertibility Undertaking" at 7.75: Asymmetrical arbitrage and "devaluation risk"

Regarding the convertibility coverage, the first technical measure announced on 5 September spelt out a "convertibility undertaking" from the HKMA to the licensed banks "to convert Hong Kong dollars in their clearing accounts into US dollars at the fixed exchange rate of HK$7.75". There are two issues here: (1) Convertibility is only "one sided": the HKMA has not undertaken to convert US dollars into Hong Kong dollars at the fixed rate. (2) The convertibility rate is 7.75 instead of linked rate of 7.80.

One-sided convertibility at the fixed rate implies that the powerful "electronic arbitrage mechanism" of the AEL model works only when the market exchange rate weakens above 7.75, but would not operate if it stays on the strong side of the rate. While "revaluation" is obviously not at all a concern at the moment, such an asymmetrical treatment tends to undermine the robustness and transparency of the system. On the plus side, it is said that at least the spot HK$/US$ market is not entirely "killed" and some variability on the strong side of 7.75 still exists. So not all dealers in the original market would lose their job! Cynics would also say that the HKMA still wants to preserve the right to "punish" short sellers of the Hong Kong dollar.

As to the decision to fix the convertibility rate at 7.75, instead of 7.80, it is really the unpalatable legacy of resorting to "discretionary intervention" to defend the linked rate in the past (Tsang, 1998a). The government admitted that "(t)he rate of 7.75 is the current intervention rate of the HKMA". Moreover, the reason given by the HKMA for not moving to the rate of 7.80 in one go was that some parties had hedged against the risk of the devaluation of the Hong Kong dollar by locking in derivatives that would trigger a selling order should the exchange rate weaken beyond 7.75. So the authorities wanted to provide them with time to close those positions.

The trouble is that the HKMA made a strange announcement on 5 September 1998. It is worthwhile to quote it in detail:

"However, it is the clear intention of the HKMA, when market circumstances permit, to move the rate of the Convertibility Undertaking to 7.80, which is the fixed exchange rate of our linked exchange rate system applicable to the issue and redemption of Certificates of Indebtedness backing the bank notes. This will be done at a time when the market exchange rate is trading consistently at a level significantly stronger than 7.75."

So the HKMA plans to "devalue" the Hong Kong dollar (from the rate of 7.75 to 7.80), i.e. by about 0.64%, when the market exchange rate is "trading consistently at a level significantly stronger than 7.75"! However, given the "devaluation" pledge, how on earth would the market exchange rate be "trading consistently at a level significantly stronger than 7.75"? Unless the HKMA resort to "discretionary intervention" again, which it has promised not to do so?

More seriously, it introduces a "devaluation risk" into the whole picture. When will the HKMA move the rate to 7.80? If the HKMA makes a pre-announcement of the date of the "move", everyone will convert all their liquifiable Hong Kong dollar assets into US$ deposits before the date, and then re-convert back to HK$ assets after the date. About one third of the HK$ deposits in the banking system are current account or "callable" deposits. Pre-announcement will create chaos in the system.

If the HKMA does not make a pre-announcement, as it probably will not, then the "devaluation risk" will be reflected in interest rates. They will show a "devaluation risk premium", on top of other risk premia (systemic risk premium, the East Asian risk premium, etc.). A simple example can illustrate the problem. Suppose an investor wants to do interest arbitrage by borrowing low-cost US dollars and converting them into high-return HK dollar deposits, he has to be sure that the spot rate is safely locked by the electronic currency arbitrage mechanism of the AEL model. Let say that one-month US$ interest rate is 5%. How high should the interest rate of one month HK$ deposits be, so that the investor is interested in doing interest arbitrage in the face of the "devaluation risk"?

The answer is 13% or above! For any interest rate levels lower than that, since the extent of the devaluation (from 7.75 to 7.80), 0.64%, translates into an annualized rate of about 8%, any investor doing interest arbitrage will stand to lose if he locks into a position of one month, only to witness helplessly the "move" by the HKMA before his HK$ deposit matures!

Needless to say, this is a strong disincentive to interest arbitrage, which undermines the effort by the HKMA to bring down interest rates through adopting partially the AEL model. To rectify the problem, I think that the HKMA should either move to 7.80 immediately, before the market fully digests the implications, or stay put with the rate of 7.75. Let us have a linked exchange rate system at HK$7.75/US$! Better still, the HKMA should guarantee "bilateral convertibility" (two-way convertibility of the monetary base: from HK$ to US$, and vice versa) at the fixed rate. This will make the system entirely transparent.

The discount window: Speculative attack on a CBA that is also a financial centre

The six other "technical measures" announced on 5 September 1998 were all related to the establishment of the "discount window"(DW), which replaced LAF. The major difference is that penalty on "repeated borrowings" (from the LAF) is now largely (but not entirely) removed. Holdings of Exchange Fund paper can be used as collateral to obtain funds from the DW: the first 50% at the "base rate", and the remaining 50% at "base rate plus 5% or overnight HIBOR for the day, whichever is higher". At the moment, holdings of Exchange Fund paper by banks amount to about HK$60 billion. That implies that HK$30 billion worth of collateral is readily usable at the base rate.

Since all issues of Exchange Fund paper were and will be backed by foreign exchange reserves, the HKMA argues that more convenient borrowing by banks from the DW "does not involve any departure from the discipline of the Currency Board system". Previous eligible instruments for LAF other than Exchange Fund paper, e.g. notes issued by pseudo-public institutions such as the Hong Kong Mortgage Corporation, the Airport Authority and the Mass Transit Railing Corporation will be phased out as they mature.

The DW will help to solve the problem of the aggregate clearing balance of banks at the HKMA being too small. As I argued earlier (Tsang, 1998b), the aggregate balance should theoretically be zero, but in practice amounts to a few billion Hong Kong dollars and may even turn negative. Hence I proposed a deposit reserves system under which banks would keep 1% to 2% "excess reserves" (assuming that reserves are interest bearing). If it is 2%, the "excess" would also be about HK$30 billion, given that the total amount of HK$ deposits is now about HK$1,500 billion.

The HKMA did not adopt my proposal and go for the DW instead. Both systems have similar end results: they provide a "reservoir" beside the small pool of the aggregate clearing balance. In case of a speculative attack such as the "double market play" which aimed at pushing up interest rates and selling down stock prices, and which led to the Hong Kong government intervening into the stock and futures markets in August 1998 (Yam, 1998; Tsang, 1998d), the "reservoir" can provide convenient liquidity to relieve interest rate pressure (Tsang, 1998b; 1998c).

Some may argue that the HKMA is deviating from classical currency board principles. After all, the DW is not only playing the LOLR function, whose sole objective is to prevent financial panics or banking collapse. It is intended to relieve interest rate pressures. However, we need to appreciate Hong Kong's peculiar predicament: it is the only CBA in the world that is also an international financial centre! Cross-border flows of funds are huge, and have strong spillover effects far beyond the foreign exchange market. This is something that Argentina, Estonia and Lithuania do not face.

There is however one important difference between a DW and a deposit reserves system. In the former, banks borrow money from the central bank, and the central bank has to set a discount rate. The HKMA now finds itself having the headache of deciding on the base rate. Because punishment on "repeated borrowing" is largely abolished, the rate cannot be too high, or too low. To remove arbitrage elements and the accusation that it is practicing "monetary policy", the HKMA is trying the formula of using the average of overnight and one-month HIBORs in the preceding day and announces the base rate daily. The difficulty about a formula is that it must not be manipulable by speculators. Moreover, the base rate should not fluctuate widely lest interest rate stability will be jeopardized.

In the case of deposit reserves system, this problem is largely by-passed. The HKMA can set the interest rate that it pays the reserves with it at some long-term trend rate (slightly below the market level) and then stay put. Since the banks do not borrow any money from the HKMA when it becomes necessary to transfer funds from the reserves account to the settlement account (which constitutes a part of the aggregate clearing balance), the HKMA does not have to face the tricky problem of setting a lending rate such as the base rate.

A few concluding words

The HKMA finally decided to adopt the electronic arbitrage mechanism of the AEL model. This is certainly a move to be welcome, but one question in the mind of nearly all commentators is: why so late? I am afraid that it belongs to the realm of political economy, rather than monetary economics, and I have no answers to offer. I can only say that given Hong Kong's huge foreign exchange reserves, and the stock of existing Exchange Fund paper, those seven "technical measures" could actually have been implemented at any time, with very little preparation.

As far as the AEL model is concerned, I should emphasize that I am just a "discoverer", an "interpreter", and certainly not an inventor. I am also fully aware that not many economists in these three countries "interpret" their own system in the way that I do. Living in a regime whether the spot exchange rate is dead fixed and observing no actual arbitrage (because arbitrage efficiency is "perfect") must make them wonder what kind of fuss that Hong Kong officials and economists have been engaging in. They probably do not know that their breakthrough lesson in currency board economics could have saved Hong Kong tens of billions of dollars, if it had been learnt earlier.

Now that Hong Kong has boarded the ship of the AEL model, there are still lingering problems and new difficulties that the HKMA has to tackle. And they better be tackled wisely.

References

Caprio, Jr. G., M. Dooley, D. Leipziger, C. Walsh (1996), "The Lender of Last Resort Function Under a Currency Board: The Case of Argentina", Policy Research Working Paper 1648, The World Bank.

Enoch, Charles and Anne-Marie Gulde (1997), "Making a Currency Board Operational", Paper on Policy Analysis and Assessment PPAA/97/10, The International Monetary Fund.

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

Hong Kong Monetary Authority (HKMA) (1998), Strengthening of Currency Board Arrangements in Hong Kong, A Technical Note, 5 September.

Santiprabhob, Veerathai (1997), "Bank Soundness and Currency Board Arrangements: Issues and Experience", IMF Papers on Policy Analysis and Assessment, PPAA/97/11.

Tsang Shu-ki (1997), "Currency Board the Answer to Rate Stability," article at web site: www.sktsang.com/ArchiveI/web981.html.

Tsang Shu-ki (1998a), "The Hong Kong Government's Financial Market Review Report: An Interpretation and A Response," article at web site: www.sktsang.com/ArchiveI/web985.htm.

Tsang Shu-ki (1998b), "The Unfinished Story: Analyzing the Options of Defending the Exchange Rate and Reducing Interest Rates in the Midst of a Crisis," Ming Pao, Hong Kong, 19 June (in Chinese).

Tsang Shu-ki (1998c), "Handling Credit Crunch under Hong Kong's Currency Board System," article at web site: www.sktsang.com/ArchiveI/web987.html.

Tsang Shu-ki (1998d), "Why I Support the Hong Kong Government's Stock Market Intervention", article at web site: www.sktsang.com/ArchiveI/web988.html.

Yam, Joseph (1998), "Why We Intervened", The Asian Wall Street Journal, 20 August.