Chief Justice Li
I agree with the judgment of Sir Anthony Mason NPJ.
Justice Bokhary PJ
I agree with the judgment of Sir Anthony Mason NPJ.
Justice Chan PJ
I agree with the judgment of Sir Anthony Mason NPJ.
Justice Ribeiro PJ
I agree with the judgment of Sir Anthony Mason NPJ.
Sir Anthony Mason NPJ
The important question in this appeal from the Court of Appeal (Stock JA, Yeung JA and Barma J) is “whether the offence of conspiracy to defraud is formulated with sufficient precision to be an offence ‘prescribed by law’ in the Basic Law or ‘according to law’ in the Bill of Rights Ordinance” (“BORO”). These expressions are used in art.39 of the Basic Law and art.11(1) of the Hong Kong Bill of Rights (“BOR”).
On 9 January 2007, the Court of Appeal certified the point of law stated above after it had delivered a judgment by which the Court allowed the appellant’s appeal against her conviction (before HH Judge Wright in the District Court) on charge 1, as well as an appeal by another appellant against her conviction on that charge, but dismissed the appellant’s appeal against conviction on charge 2 and affirmed her conviction on charge 3. Each conviction was for an offence of conspiracy, the first two offences being conspiracy to defraud, charge 3 being for conspiracy to pervert the course of justice.
On 5 February 2007, the Appeal Committee granted leave to the appellant to appeal on the certified point only and leave was refused on other grounds. The consequence is that the certified point of law arises for consideration in the context of the conviction on the second charge of conspiracy to defraud which was added by leave at the trial following a question raised by the trial judge.
THE PROSECUTION CASE AT TRIAL
The prosecution alleged two agreements, the subject of conspiracy charges 1 and 2 respectively, to create a false picture of active trading in the shares of a publicly listed company, Shanghai Land Holdings Limited (“SL”) whereby the appellant, who was the first defendant at the trial, arranged for transactions in the shares of SL through the trading accounts of other persons. These transactions were paid for by the appellant or persons connected with her, the named parties to the transactions having no beneficial interest in them.
The first agreement (charge 1) commenced to operate in about June 2002 and the second (charge 2) in about March 2003. The second agreement differed from the first in that the occasion for the second agreement was the need to maintain the market price of SL shares at a specific level in order to avoid payments falling due under a loan agreement between Global Town Limited (“Global Town”), a company incorporated in the British Virgin Islands, and the Bank of China (Hong Kong) Limited (“BOC”), by which shares in SL were pledged to the Bank.
THE CONSPIRACY TO DEFRAUD CHARGE
The second conspiracy charge was against the appellant alone. It alleged that between 1 March 2003 and 31 May 2003 she conspired with Lui Ching-yee, Joanne (“Lui”) and Chau Ching-ngai (“Chau”) and others “to defraud the Bank of China by dishonestly creating a false or misleading appearance with respect to the market for or the price for dealing in shares in [SL] ....”
The particulars of this charge were as follows:
[As to the economic right that was said to be prejudiced] The Bank of China could be put at risk of economic loss arising from the loan arrangement between themselves and [SL].
[As to the object of the alleged fraud] The charge involves an agreement the object of which is to maintain turnover and/or the price of [SL] shares.
Chau was the appellant’s husband and sole director and shareholder of Global Town.
GLOBAL TOWN, ITS LOAN AGREEMENT WITH BOC AND THE ACQUISITION OF SL
On 23 April 2002, BOC approved a loan facility in favour of Global Town of up to $2,156m to enable Global Town to acquire SL, then called imGO Limited (“imGO”). It was a term of the loan agreement that the shares in that company be pledged to BOC as security for the indebtedness.
Global Town drew on the facility to acquire the share capital of imGO. In all, under the facility, it borrowed $1,777,122,079 from BOC. To comply with the 25% public float requirement under the relevant Listing Rules, Global Town made three placements in the autumn of 2002 to reduce its shareholding to 75%.
By the loan agreement, Global Town agreed that the value of the pledged shares on any given date would not be less than the outstanding balance of the loan on that date. If the value of the pledged shares fell below that balance, BOC was entitled under the agreement to call upon Global Town to make good the difference. The agreement also provided, however, that, as from 31 March 2003, which was the date stipulated for a particular repayment, 60% rather than 100% of the market value of the shares had to be at least equal to the amount of the outstanding loan. At 31 March 2003, the outstanding loan was $791,425,720. The effect of the provision was that, if the share price fell below 58¢, a differential would be payable. This meant that, if the share price fell to 57¢ on that date, 60% of the market value of the shares would be $782,673,840 and a differential of $8,751,880 would become payable immediately.
THE OPENING OF BANK ACCOUNTS IN THE NAMES OF THE TRADERS
Between September 2000 and February 2003, more than 42 securities trading accounts were opened in the names of at least 12 persons with 13 different securities companies. Many of the accounts were opened without the knowledge or consent of the traders in whose names they were opened. In August and September 2000, bank accounts were opened for each of the 12 traders at the behest of the appellant. Blank cheque books pre-signed by the traders were given to the appellant. Those accounts were used by Lui, the appellant’s personal assistant, for share trading.
From about June or July 2002, on the appellant’s instructions, Lui traded in SL shares, using the accounts already opened and additional accounts opened for the same purpose in the names of friends and relatives of the appellant and Lui. The trading was margin trading. After Lui purchased shares, she would sell them, or approximately the same number, the next day. Only SL shares were traded. Payments to the securities companies through whom the trades were effected were made with the pre-signed cheques. Lui regularly informed the appellant how much was required to meet these payments so that arrangements could be made for the deposit of the required sums into the relevant bank trading accounts. On receipt of this information from Lui, the appellant gave instructions to Chung Sau-ling (“Chung”) who arranged for the deposits to be made.
THE 58¢ THRESHOLD
After the Securities and Futures Commission made inquiries of 14 securities companies seeking information in respect of dealings in SL shares, the appellant instructed Lui that trading had to be reduced and it was reduced. In late March 2003, the financial controller of SL told Lui that the closing price of SL shares should not be allowed to fall below 58¢, otherwise interest would be payable to BOC. Next day, near to closing time, the appellant told Lui to maintain the share price at 58¢. This she did. The attempt eventually collapsed in May 2003, when it became known publicly that Chau had been arrested in Shanghai for separate matters. In order to try to maintain the share price at 58¢, Lui was required to purchase all shares in SL for which the asking price was below 58¢. Sometimes, if the quantity for sale was large, she would be instructed by the appellant not to buy and the price would then drop below the target figure. She did not consult the account holders before using their trading accounts. The trading pattern was one in which there was a similarity between purchase price and selling price. A loss frequently resulted. Payments were made from the trading accounts to repay margin, interest and the loss resulting from the sale of the shares. Payments were made by cheque drawn on the bank accounts of the traders, although they themselves did not make the funds available. It was Lui’s role, as before, to tell the appellant what cheques needed to be issued and Chung was instructed to deposit funds into the relevant bank accounts.
The first time the share price fell below 58¢ per share was on 9 April 2003 and thereafter that happened often. As a result, BOC telephoned Global Town from time to time, requesting payment of the differential. The Bank would usually wait until the close of the trading day to see whether the price had risen back to 58¢ and, if it had, the Bank would not insist upon payment of the difference.
Between 26 July 2002 and 24 April 2003 the traders’ purchases and sales of SL shares took place each trading day, their activity accounting for 86% of buying and 81% of selling. Market turnover of SL shares averaged $13m a day and the traders bought and sold on average $11.1m and $10.4m per day, respectively. Between 25 and 28 April 2003 the traders only purchased shares and from 29 April 2003 onwards, the Group only sold, save for a small amount on 9 May. At about the end of April 2003, the appellant instructed Lui to use Shun Loong Securities, a company of which the appellant and Chau were directors, to acquire SL shares for Chu Lai Ding (“Chu”). He paid $2,600,000 for SL shares, when their price was falling. He was later approached for further monies but did not agree to buy. Nonetheless a huge amount was paid, ostensibly on his behalf but in fact without his authority, on 28 May 2003. The judge was satisfied that his account was used in an attempt to support the price of SL shares.
The price continued to fall and when margin calls were made by various securities companies, payments were made by the appellant. On 30 May 2003 the share price closed at 34.5¢ and on 6 June 2003 the Board of Directors appointed a receiver.
Of the $62.598m deposited into 39 of the 43 trading accounts for the purpose of settling SL share purchases, $17.7m was attributable “on a conservative basis”, as the judge put it, to the bank account of the appellant, whereas $7.646m was attributable, again on conservative basis, to the account of Li Yung, a bank account used by the appellant, aggregating 40.52% of the total deposits.
In relation to the second charge, the judge referred to the evidence that BOC made telephone calls to request payment of the differential and to the Bank’s practice to wait until the close of the trading day to see whether the price had reverted to 58¢, in which event the Bank did not insist on payment. The judge went on to say:
In my judgment this, alone, provides sufficient reason for the conspirators to have sought to support the price of [SL] shares at 58 cents per share – even ignoring the additional incentive of avoiding actual payment of the differential ....
At no stage was a differential payment made to BOC, although very substantial amounts became payable from 9 April onwards as a result of the drop in the share price.
The judge also found that:
In so far as the second charge is concerned, the essence of the offence is to prejudice or to take the risk of prejudicing the rights of BOC, the offender/s knowing that he or she has no right to do so. Those rights in question are also economic rights. There might well have been no intention to cause loss to BOC but it is the economic right of BOC to assume that the information in respect of the stock which it holds as security for amounts due to it has not been adulterated by an artificially increased trading volume, that third-party interest in the shares of [SL] are genuinely represented by the share trading volume. Thus the artificial trading volume acts to defraud BOC by the creation of a false or misleading impression of interest in the shares of [SL].
It was also found that the conspirators entered upon an agreement designed specifically for the purpose pleaded. This purpose included the purpose of avoiding the effect of the differential clause of the loan agreement. It followed that they appreciated that their conduct, if effective, would deprive the Bank of monies to which it would be otherwise entitled.
THE COURT OF APPEAL'S JUDGMENT ON THE CONSTITUTIONAL POINT
Although the Court of Appeal acknowledged that there was no precise definition of the offence, the Court concluded that it was formulated with sufficient precision to satisfy the requirements in the expressions “prescribed by law” and “according to law” in the Basic Law and the BORO respectively. Their Lordships considered three classes of case:
where the agreement relates to action which will or may cause economic loss (the present case);
where the agreement relates to action which is to the prejudice of another person’s right; and
where the agreement relates to action whereby a person or body under a public duty is induced to act contrary to that duty.
Their Lordships held that the offence, as it applied to classes (1) and (3) above, is sufficiently certain, while class (2) is a narrow class of non-economic frauds, rarely prosecuted, which, even if uncertain, should not affect the constitutionality of the offence in its application to classes (1) and (3).
THE CASE FOR THE APPELLANT
Ms Montgomery QC argues that the Court of Appeal’s analysis of the offence significantly underestimated the width of the offence and thus misjudged its constitutionality. Her argument consists of four separate but inter-related submissions:
that there is a lack of precision in the definition or description of the elements of the offence, notably dishonesty and the lack of any requirement for deception or deceit;
that it is uncertain whether the offence extends to the deprivation of non-economic loss outside the “contrary to public duty” situations;
that it is uncertain whether the offence extends to cases where a person or authority is induced to act or not act “contrary to private duty” as well as contrary to his or its public duty; and
that the offence suffers from overbreadth in the sense that, as formulated, it embraces, or is capable of embracing, acts that either are not criminal in their nature or go beyond the range of acts intended to be penalized.
These general submissions are supported by specific submissions which will be considered in the course of these reasons.
CRITICISMS OF THE WIDTH AND OBJECTIONABLE CHARACTERISTICS OF THE OFFENCE
The appellant relies on strong criticisms of the offence of conspiracy to defraud made in the reports of law reform authorities in Hong Kong, the United Kingdom and Australia and in the writings of leading academic commentators. From these reports and writings the following points emerge:
that the offence is not susceptible of precise definition and is of great width, being expressed in very general terms that are capable of applying to a broad range of transactions;
that it criminalizes conduct by two or more persons which, if undertaken by an individual, would not be criminal or tortious;
that it penalizes conduct which involves no deception;
that it penalizes conduct which is the subject of specific statutory offences;
that it is formulated in vague and imprecise terms so that it offers insufficient guidance as to what can and cannot be lawfully done;
that it infringes the principle that the criminal law should be knowable in advance regarding the conduct to be penalized; and
that it violates the principles of legality and legal certainty.
The cogency of the first four points have been acknowledged in judgments of the highest authority - see, for example, R v Withers  AC 842 at 862E (per Lord Diplock), 874A–F (per Lord Kilbrandon). The same comment cannot be made about the remaining three points. Ms Montgomery QC seeks to make them good.
CONSPIRACY TO DEFRAUD: THE MODERN AUTHORITIES
The elucidation over the last fifty years of the common law offence of conspiracy to defraud has resulted in the recognition that deceit is not an essential element of the offence and that, instead, the emphasis is on dishonesty. Whether dishonesty goes to intent (or purpose) or the means by which the agreement is to be effected or both is a critical matter for consideration, as is the scope of the offence. On both these matters, it is necessary to look at three decisions of the House of Lords and a decision of the Privy Council.
In Welham v Director of Public Prosecutions  AC 103, the House of Lords held for the first time that the expression “intent to defraud” in s.4(1) of the Forgery Act 1913 could not be confined to the idea of depriving a person by deceit of some economic advantage or of inflicting upon him some economic loss, there being a difference between “an intention to deceive” and “an intention to defraud”, “defraud” having a wider meaning than “deceive”. The House also held that intent to defraud existed when a false document was brought into existence for no other purpose than of deceiving a person subject to a public duty into doing or not doing something contrary to that public duty.
Lord Denning rejected the argument that an intention to defraud involves an intention to cause economic loss and said - Ibid at 133:
Put shortly, ‘with intent to defraud’ means ‘with intent to practise a fraud’ on someone or other. It need not be anyone in particular. Someone in general will suffice. If anyone may be prejudiced in any way by the fraud, that is enough.
Lord Radcliffe agreed with Lord Denning and rejected the proposition that in ordinary speech “to defraud” is confined to the idea of depriving a man by deceit of some economic advantage or inflicting upon him some economic loss - Ibid at 124. Other members of the Appellate Committee agreed with Lord Radcliffe and Lord Denning. Although the decision related to a statutory provision, it has always been considered to provide guidance in cases of conspiracy to defraud.
The effect of Welham was, as the appellant submits, to broaden the offence of conspiracy to defraud by removing a limitation which had previously been thought to apply, namely an intention to deceive. R v Scott  AC 819 confirmed that this was the effect of Welham. In Scott, the appellant agreed with employees of a cinema that in return for payment they would, without their employers’ consent, temporarily abstract films so as to enable the appellant to make and distribute copies on a commercial scale, without the consent of the owners of the copyright and distribution rights in the films. The appellant was charged with conspiracy to defraud and other charges. He was convicted of conspiracy to defraud.
Scott, Ibid at 836B–C, 838B–C, G–839A held that the common law offence of conspiracy to defraud was not limited to an agreement between two or more persons to deceive the victim and by such deceit to defraud him. As deceit was not an essential element of the offence, the appellant was rightly convicted. Viscount Dilhorne, with whose speech other members of the Appellate Committee concurred, stated - Ibid at 839B:
One must not confuse the object of a conspiracy with the means by which it is intended to be carried out.
His Lordship went on to say - Ibid at 839C:
‘to defraud’ ordinarily means .... to deprive a person dishonestly of something which is his or of something to which he is or would or might but for the perpetration of the fraud be entitled.
Lord Diplock in his concurring speech said that where the victim of a conspiracy to defraud is a private individual, the purpose of the conspirators must be to cause the victim economic loss by depriving him of some property or right to which he is or might become entitled. He went on to treat the “contrary to public duty” cases as a special category - Ibid at 841C. His Lordship said that to constitute fraud - Ibid at 841B:
the intended means by which the purpose is to be achieved must be dishonest .... Dishonesty of any kind is enough.
In Wai Yu-tsang v The Queen  1 AC 269 (a decision of the Privy Council on appeal from the Hong Kong Court of Appeal), the defendant, the chief accountant of a bank, dishonestly agreed with others to conceal the dishonour of cheques purchased by the bank. He was charged with others with conspiracy to defraud. The judge directed the jury that the defendant had been a party to an agreement with others with the common intention to defraud one or more of the persons named in the indictment, and that imperilling the economic or proprietary interests of such person or persons was sufficient to constitute fraud even if no loss was suffered and the defendant did not desire that any loss should be caused. The Privy Council upheld the conviction. Lord Goff of Chieveley, who delivered the advice of the Judicial Committee, treated Welham as establishing that - Ibid at 276E–F:
‘intent to defraud’ is not to be given a narrow meaning, involving an intention to cause economic loss to another. In broad terms, it means simply an intention to practise a fraud on another, or an intention to act to the prejudice of another man’s right.
His Lordship then turned to the short concurring speech of Lord Diplock in Scott. Lord Goff considered Lord Diplock’s categorization of the “contrary to public duty” cases as too narrow and affirmed the general principle that conspiracies to defraud are not restricted to cases of intention to cause the victim economic loss ( 1 AC at 277), the general principle having been enunciated in Welham and favoured by Viscount Dilhorne in Scott.
Finally Lord Goff turned to the mental element in conspiracy to defraud. On this question his Lordship observed -  1 AC at 279H – 280B:
The question whether particular facts reveal a conspiracy to defraud depends upon what the conspirators have dishonestly agreed to do, and in particular whether they have agreed to practise a fraud on somebody. For this purpose it is enough for example that, as in Reg v Allsop (1976) 64 Cr. App. R. 29 and in the present case, the conspirators have dishonestly agreed to bring about a state of affairs which they realise will or may deceive the victim into so acting, or failing to act, that he will suffer economic loss or his economic interests will be put at risk. It is however important in such a case, as the Court of Appeal stressed in Reg v Allsop, to distinguish a conspirator’s intention (or immediate purpose) dishonestly to bring about such a state of affairs from his motive (or underlying purpose). The latter may be benign to the extent that he does not wish the victim .... to suffer harm; but the mere fact that it is benign will not of itself prevent the agreement from constituting a conspiracy to defraud.
With respect to Lord Goff’s general principle that conspiracies to defraud are not restricted to cases of intention to cause economic loss, it should be noted that, in Wai, no reference was made to R v Withers,  AC 842 a decision of the House of Lords delivered on the same day as Scott. Nor was Withers cited in argument. In Withers, the defendants conducted an investigation agency in the course of which they made reports to clients about the status and standing of third parties. The defendants agreed to make false representations to public officers and officers of banks and building societies to obtain confidential information by pretending to be acting in an official capacity. They were convicted on charges of conspiracy to effect a public mischief. The House of Lords held that the law knew no such generalised offence as conspiracy to effect a public mischief, although certain cases so classified in the past fell within established heads of conspiracy, such as conspiracy to defraud or conspiracy to pervert the course of justice. The issue therefore arose in Withers whether the appeal should be dismissed on the ground that no miscarriage of justice had actually occurred.
In this respect, Viscount Dilhorne (with whom Lord Reid agreed) expressed no firm opinion on the question whether, on the facts, which involved no possibility of economic loss, the defendants might have been charged and convicted on conspiracy to defraud - Ibid at 860H. Lord Diplock thought that the particulars under count 2 of the indictment might have supported a conspiracy to defraud charge on the “contrary to public duty” footing (Ibid at 862G), but, in the light of his views in Scott, not otherwise. Lord Simon of Glaisdale accepted that the law of conspiracy to defraud had developed to the point of recognizing that it applied to the contrary to public duty cases. His Lordship was, however, opposed to the view that conspiracy to defraud otherwise extended to cases of non-economic loss, as his comments (Ibid at 872H – 873B) about R v Bassey (1931) 22 Cr. App. R. 160 demonstrate. He agreed with Lord Kilbrandon’s view that count 2 should be quashed. Lord Kilbrandon also accepted that conspiracy to defraud extended to the contrary to public duty cases but was not in favour of it extending otherwise to cases of non-economic loss -  AC at 877H – 878A. His Lordship quashed the conviction on count 2 because the facts did not support a conviction on that count.
It follows that three members of the Appellate Committee were of the opinion that, apart from the contrary to public duty cases, conspiracy to defraud does not apply to cases of non-economic loss.
CONSPIRACY TO DEFRAUD: THE ESSENTIAL ELEMENTS
What emerges from this discussion of the cases?
First, that there is no formulation which amounts to a comprehensive definition of the common law offence.
Secondly, that the offence is constituted by becoming a party to an agreement with another or others to use dishonest means
with the purpose of causing economic loss to, or putting at risk the economic interests of, another; or
with the realization that the use of those means may cause such loss or put such interests at risk.
The offence extends also to cases in which the dishonest means cause a person to act contrary to his public duty, this being a class of case which requires no further consideration by the Court in the circumstances of this case.
Thirdly, that the identification of the elements of the offence, may leave or leaves still to be resolved –
what is meant by the requirement of “dishonesty”;
whether the offence extends to non-economic loss; and
whether the offence extends to cases where the victim is induced to act contrary his or her private duty.
The appellant argues that the existence of these questions demonstrates that there is uncertainty which is at the very core of the offence.
The importance of the first question stated above is illustrated by the different views considered, and indeed expressed, by the High Court of Australia in Peters v The Queen (1998) 192 CLR 493, although the disagreement expressed did not result in any substantial variation in the directions to be given to juries in cases of conspiracy to defraud. In any event, the High Court’s later decision, which was not cited in argument, Macleod v The Queen (2003) 214 CLR 230 resolved these differences.
In considering the first question posed in para.40 above, no assistance is to be gained by attempting to draw distinctions between the use in the decided cases of the different expressions “dishonesty”, “dishonestly” and “by dishonest means”. Assistance is, however, to be gained from recalling that the emphasis on the element of dishonesty accompanied the rejection of deceit as an indispensable element in the offence. In other words, “by dishonest means” replaced the earlier reliance on deceit and to that extent broadened the offence. That this is so appears from the speeches of Viscount Dilhorne and Lord Diplock in Scott and is confirmed by Wai where there was no rejection of Lord Diplock’s view on this point.
In the context of conspiracy to defraud, it is becoming a party to the agreement to use dishonest means that constitutes the material element of dishonesty. Dishonesty involves a state of mind. Most cases of agreement to use dishonest means are cases of deceit. In other cases of dishonest means a defendant may agree to do something which he knows he has no right to do or believes that he has no right to do or he agrees to do something which is dishonest and he knows or believes that it is dishonest.
Dishonesty apart, conspiracy to defraud involves another state of mind – the intention or purpose of inflicting economic loss on another or the realization that the action agreed upon will or may put that person’s economic interests at risk, the element to which Lord Goff referred in Wai  1 AC at 280. This state of mind may well involve dishonesty. It is preferable, however, to treat it as a separate element. In this respect, it was not suggested in argument that R v Woollin  1 AC 82, a decision on intent in murder, has any application to the mental element in conspiracy to defraud.
That “dishonesty” describes a state of mind was accepted by the English Court of Appeal in R v Ghosh  1 QB 1053. It was a decision on s.1 of the Theft Act 1968 (UK) but the Court of Appeal, reflecting the view earlier taken in Scott, considered that there was no difference between the element of dishonesty in theft and the element of dishonesty in conspiracy to defraud - Ibid at 1058F – 1059E. The Court concluded that “dishonestly” in s.1 of the Theft Act described a state of mind and was not a characterization of a course of conduct - Ibid at 1063F.
Although the Court of Appeal’s conclusion on dishonesty was that it involved a mental state, the Court formulated a two-stage test of dishonesty consisting of both objective and subjective elements. The test was expressed in these terms - Ibid at 1064E–F:
In determining whether the prosecution has proved that the defendant was acting dishonestly, a jury must first of all decide whether according to the ordinary standards of reasonable and honest people what was done was dishonest. If it was not dishonest by those standards, that is the end of the matter and the prosecution fails.
If it was dishonest by those standards, then the jury must consider whether the defendant himself must have realised that what he was doing was by those standards dishonest. In most cases, where the actions are obviously dishonest by ordinary standards, there will be no doubt about it. It will be obvious that the defendant himself knew that he was acting dishonestly. It is dishonest for a defendant to act in a way which he knows ordinary people consider to be dishonest, even if he asserts or genuinely believes that he is morally justified in acting as he did ....
Two comments should be made about the two-stage test. First, if dishonesty involves a mental state why do we need the objective test in stage 1? The answer to this question is that the dishonesty of the means agreed upon is, in the first instance, a question of fact for the jury. The meaning of an ordinary word is a question of fact, not of law, unless the word is used in some special sense. Here the word “dishonest” in its association with “means” is used in its ordinary sense, namely by reference to the ordinary standards of reasonable and honest people. It is therefore a question for the jury to decide whether, in the circumstances, the words do or do not cover or apply to the proved facts. This approach was taken in Brutus v Cozens  AC 854 where the words “insulting .... behaviour” were made an essential element of a statutory offence triable before justices.
Although dishonesty in its ordinary sense does not lend itself to definition, it is not an arbitrary standard and a citizen should have little difficulty in appreciating what is dishonest judged by the ordinary standards of reasonable and honest people. This was recognized in Shum Kwok Sher v HKSAR (2002) 5 HKCFAR 381 at 412D–J where it was pointed out that the search for definition can be taken too far and that a higher degree of definition of some standards is often unattainable. Dishonesty in its ordinary sense is one of them.
The second comment to be made is that the focus of the Ghosh stage 2 test is on the defendant’s belief or absence of belief in the honesty of the means agreed upon. This is in contrast to the former position when conspiracy to defraud was confined to deceit. Then, the relevant mental element was based on the defendant’s knowledge or belief in the falsity of the means agreed upon such as the representations agreed to be made. The justification for the Ghosh focus arises from the shift from the former insistence on deceit as an indispensable element to the requirement “by dishonest means”. And the Ghosh stage 2 test has the important practical advantage that it provides a uniform test capable of general application in cases of conspiracy to defraud.
In those cases in which the dishonest means involve deceit, it might be possible to consider applying the former approach based on knowledge or belief of the falsity of the deceitful action agreed upon. This would involve
identifying the knowledge, belief or intent which is said to render the defendant’s conduct dishonest; and
deciding whether the defendant had that knowledge, belief or intent and, if so, whether on that account the act was dishonest by the standards of ordinary standards of honest and reasonable people -
see Macleod v The Queen (2003) 214 CLR at 230.
But this alternative approach has not been argued and in the absence of argument it would not be right to embark upon the question. As things stand, Ghosh represents the law in Hong Kong. The Ghosh two-stage test has been consistently applied in Hong Kong since it was enunciated by the English Court of Appeal. It is the test to be applied in Hong Kong for determining whether the means agreed upon by the alleged conspirators are dishonest. Despite academic and other criticisms of the test, there is nothing before this Court to indicate that the application of the test has generated problems or difficulties. In Shum Kwok Sher it was held to be sufficiently certain.
DOES THE OFFENCE EXTEND TO NON-ECONOMIC LOSS?
There is now a strong body of English authority to support the proposition that conspiracy to defraud extends to causing non-economic loss or prejudicing non-economic interests. This proposition was accepted by Lord Denning and Lord Radcliffe with the assent of other members of the Appellate Committee in Welham. It was confirmed by Scott, with only Lord Diplock dissenting, and it was accepted in Wai where Lord Goff rejected Lord Diplock’s narrow approach. These decisions have been applied in Hong Kong, notably by the Court of Appeal - see HKSAR v Law Kam Fai  2 HKLRD 879; HKSAR v Chan Chun Hong  1 HKLRD 346; see also the judgment of the Court of Appeal which is under appeal in this case, para.60. The proposition has support in Australia - see Peters v The Queen (1998) 192 CLR 493 at 525 (where McHugh J said it may be sufficient that the object of the agreement concerns “a non-economic right or interest – such as private reputation or personal status”). On the other hand, in Wai, the attention of the Privy Council was not drawn to Withers where, as already pointed out, three members of the Appellate Committee seemed to be opposed to extension of the offence to non-economic loss cases outside “contrary to public duty” situations.
In these circumstances, it is preferable that this Court refrain from resolving this question until it is the subject of a concrete case between contending parties in which arguments are put for and against the application of the offence to non-economic loss.
CONTRARY TO PRIVATE DUTY
We have not been referred to any authority which supports the application of the offence to a situation in which the agreement is to use dishonest means to induce a person to act contrary to his private duty. Neither principle nor policy provides support for such an application. The proposition, if accepted, could open up conspiracy to many cases of breach of contract rather than leave conspiracy in its application to contractual relationships confined to the category of economic loss and prejudice to economic interests.
CONCLUSIONS ON MATERIAL ISSUES CONCERNING THE OFFENCE
It follows from the preceding discussion:
that dishonesty is an essential element in the offence in that the means agreed upon must be dishonest;
that the test for dishonesty is the two-stage test enunciated in Ghosh;
that the offence includes cases of economic loss and the “contrary to public duty” cases;
that the essence of the offence in the former category is to use dishonest means to cause economic loss to another or to put at risk another’s economic interests and while there is strong authority to support the proposition that the offence extends to cases of non-economic loss, the correctness of this proposition remains finally to be resolved; and
the offence does not extend beyond the “contrary to public duty” cases to “contrary to private duty” situations.
Having identified these elements of the offence, I can proceed to deal with the appellant’s case so far as it was directed to the material elements and leave for later consideration the arguments that were directed to the issue of width and overbreadth.
THE REQUIREMENT FOR CERTAINTY
Article 28 of the Basic Law provides:
The freedom of the person of Hong Kong residents shall be inviolable.
No Hong Kong resident shall be subjected to arbitrary or unlawful arrest, detention or imprisonment ....
Article 39 of the Basic Law provides:
The rights and freedoms enjoyed by Hong Kong residents shall not be restricted unless as prescribed by law ....
These provisions have their counterparts in the BOR. Thus art.5(1) of the BOR provides:
Everyone has the right to liberty and security of person. No one shall be subjected to arbitrary arrest or detention. No one shall be deprived of his liberty except on such grounds and in accordance with such procedure as are established by law.
And art.11(1) provides:
Everyone charged with a criminal offence shall have the right to be presumed innocent until proved guilty according to law.
As this Court pointed out in Shum Kwok Sher v HKSAR (2002) 5 HKCFAR 381 at 402A,
.... the expression ‘prescribed by law’, when used in a context such as art.39 of the Basic Law, mandates the principle of legal certainty. This principle is likewise incorporated in the expression ‘according to law’ in art.11(1) of the [BOR].
In the same case, this Court referred to decisions of the European Court of Human Rights which equated the expression “prescribed by law” to the principle of legal certainty, set out the requirements of that principle and held that these requirements were applicable to the provisions in the Basic Law and the BOR which corresponded to the relevant provisions in the European Convention on Human Rights and Fundamental Freedoms (the European Convention). In addition to these authorities, there is the recent decision of the House of Lords in R v Rimmington  1 AC 459 at 482H where Lord Bingham of Cornhill re-stated the principles applicable at common law and applicable to the European Convention.
For present purposes, it is sufficient to summarise the principles enunciated in these decisions. A criminal offence must be so clearly defined in law that it is accessible and formulated with sufficient precision to enable the citizen to foresee, if need be with appropriate advice, whether his course of conduct is lawful or unlawful. It is, however, accepted that absolute certainty is unattainable and would entail excessive rigidity. Hence it is recognised that a prescription by law inevitably may involve some degree of vagueness in the prescription which may require clarification by the courts.
The concept of legal certainty recognizes that in a common law system, the common law, declared as it is by the judges, involves an incremental element of judicial lawmaking, whether by way of moulding the law to meet new circumstances and conditions or to correct errors of principle or doctrinal error. In any event, with the common law, as with the interpretation of statute law, it is inevitable that questions continue to arise which require clarification by judicial decision. That is one reason why absolute certainty is unattainable and why some degree of vagueness is inherent in the formulation of laws, especially laws expressed in general terms.
It is also to be expected that, in the case of a general offence expressed in broad and abstract terms, that the degree of vagueness will be perhaps greater than that to be expected in the case of a specific offence directed to a particular situation or particular situations. Many efforts have been made to frame and enact specific offences with a view to bringing certainty to this area of the criminal law. Nevertheless it has been generally accepted that there is a need to have a general offence, even if some law reform agencies and commentators believe that a newly minted general offence would be preferable to conspiracy to defraud. It has been widely recognized that there is no limit to the ingenuity of fraudsters in engineering novel means of defrauding others. This ingenuity leads to the conclusion that the enactment of specific offences is not an adequate safeguard unless they are accompanied by a general offence. It is for this reason that s.16A(4) of the Theft Ordinance, Cap. 210 preserved the common law offence of conspiracy to defraud without any qualification of it. Section 159E(2) of the Crimes Ordinance, Cap. 200 contains a similar provision.
DISHONESTY: IS IT UNCERTAIN?
With these preliminary observations in mind, I turn to the argument that there is uncertainty at the core of the offence of conspiracy to defraud rather than at the margins, the latter being immaterial for present purposes, as Ms Montgomery QC acknowledges. In considering this submission, it is necessary, first, to look to the element of dishonesty. In this respect, it is suggested that the common law does no more than criminalize dishonesty. If this suggestion means that dishonesty is left at large, undefined and open-ended, as I take it to mean, the suggestion must be rejected. The essence of the offence is an agreement to use dishonest means to cause economic loss to another or to put at risk another’s economic interests. Dishonest means are means which are dishonest according to the ordinary standards of reasonable and honest persons and known by the defendant to be dishonest by those standards.
In the vast majority of cases, there will be no difficulty in ascertaining in advance whether the means agreed upon are dishonest. The decided cases do not reveal that difficulty has been experienced in determining what is meant by “dishonest means”. Although it is argued that there is no clear dividing line between “dishonest means” on the one hand and sharp practice and legitimate competitive commercial practice on the other, there are bound to be difficulties in determining whether a given case falls within or outside a stated norm. So it is with “dishonest means”. Some cases of sharp practice may amount to “dishonest means”, others not. The point is that it is not possible to generalize in relation to questions of this kind. So much depends upon particular examples and then the particularity of the example often does no more than reflect a difficulty that arises so often in the application of a law to the facts of a given situation.
The appellant submits that it is not clear whether mere silence would constitute “dishonest means” - cf. Adams v R  2 Cr. App. R. 295 (where the agreement was to use a system to conceal or disguise the source and utilization of money from legitimate inquiry). Here again the answer to the question would turn on the particular facts. It is not an example of uncertainty in the constitutional sense.
It is convenient to consider, in this context, the appellant’s submission that action taken by a defendant pursuant to an agreement with others to seize the property of a third party under a claim of right necessarily demonstrates the uncertain operation of the law of conspiracy to defraud. The submission cannot be accepted. Whether the defendant’s action in such a case is dishonest turns primarily on the defendant’s belief. If he believes that his action is honest by the ordinary standards of reasonable and honest people, the means adopted are not dishonest in the required sense. In other words, a jury might conceivably find that the means were dishonest according to ordinary standards but not in the required sense because the defendant did not believe his action was dishonest by those standards.
NON-ECONOMIC LOSS: UNCERTAINTY?
Then there is the question about the application of the offence to non-economic loss. As the authorities presently stand, the offence extends to non-economic loss. On that footing, the law is formulated with sufficient precision to enable a citizen to regulate his conduct in a way that enables him to determine, if need be, with legal advice, that he should not enter into an agreement to use dishonest means whereby non-economic loss may be caused to another. If, on the other hand, we assume that the question whether the offence extends to non-economic loss is finally still to be resolved, the existing state of the authorities is such as to provide fair warning to the citizen that he is at risk if he enters such an agreement. And that question is of a kind which the courts are expected to resolve as an ordinary incident of the judicial process in a common law system. In any event, this case does not involve non-economic loss or prejudice to non-economic interests. The appellant’s conviction was based on the purpose of prejudicing the economic interests of the Bank.
Accordingly, and because the offence does not extend to cases of “contrary to private duty”, the argument that there is uncertainty at the core of the offence rather than at the margins is rejected.
OVERBREADTH AND WIDTH
Next, it is necessary to consider the appellant’s arguments on overbreadth and width of the offence. At times, the arguments seemed to proceed on the footing that width was but an aspect of overbreadth. At other times, it seemed that the arguments treated overbreadth as synonymous with disproportionality in the sense in which it has been used by this Court in constitutional cases. It is convenient to consider, in the first instance, the submissions relating to width.
Some examples of width are hypothetical and are designed to suggest that there is, or will be, difficulty in determining whether activities of a particular kind are within or outside the boundaries of the offence. Thus, it is suggested that there is uncertainty arising from the possible application of the offence to sharp practice (not amounting to fraud), to legitimate competitive commercial practices, to collective industrial action, the consequence of which is to inflict economic loss on employers, and to collective private action which inflicts loss on family members. Of these examples, it is sufficient to say that they are identified only in very general terms with the result that it is not possible to say on which side of the line they fall and how much difficulty, if any, would be experienced in reaching a conclusion once the facts of a given case are established. Further, it can be said that no cases of the kind mentioned have come before the courts in Hong Kong.
Ms Montgomery QC submits that, even if this be so, the very existence of the offence and its possible application to areas such as commercial activity and industrial relations will have a deterrent effect on legitimate conduct, that is, a chilling effect. For example, it may have a stultifying impact on entrepreneurial business activity. This is a matter of speculation. Such an impact may have more significance in some areas than others. If the offence were to have a chilling effect on liberty of the individual or freedom of speech, that would or might well amount to an inhibition incompatible with the ethos of a free society and could infringe the principle of legal certainty. But in other areas, particularly where a law is expressed in general terms to protect people from being defrauded and where it is not feasible to frame or enact a clearly defined rule in its place, it is reasonable to expect a citizen to act well within the parameters of what is lawful in order to avoid running the risk of committing an offence - see Noise Control Authority v Step In Ltd (2005) 8 HKCFAR 113 at 120F–I, 131G–132D; R v Secretary of State for Trade and Industry, ex p. Ford (1984) 4 Tr & L 150.
Another complaint is that of double criminality where conduct is criminalized under specific statutory offences as well as conspiracy to defraud. It is said that, at the time the alleged offences were committed, the market manipulation complained of in charges 1 and 2 was effectively co-extensive with the market manipulation prohibited by s.135 of the Securities Ordinance, Cap. 333, which is now repealed. This offence was punishable by imprisonment for 2 years, whereas under s.159C(6) of the Crimes Ordinance, Cap. 200, conspiracy to defraud is punishable by imprisonment for 14 years. The existence of overlapping offences does not in itself give rise to uncertainty. What is more, when the legislature decides that overlapping offences shall co-exist, although in most cases it would be expected that the prosecution would charge the appellant with the statutory offence, it does not follow that the prosecution cannot resort to the general offence where there are circumstances to justify that course. Here, for example, it was no doubt thought that the degree of criminality warranted, as indeed it did, recourse to conspiracy to defraud with its higher penalty in order to reflect the high degree of criminality on the part of the appellant and its potential impact on BOC which was the motivation for supporting the price of the shares at 58¢. This comment has particular application to charge 2.
A further point is made with reference to s.295(1)(b) of the Securities and Futures Ordinance, Cap. 571, which replaced the repealed s.135. Section 295 came into operation on 1 April 2003, after the course of dealing the subject of charge 2 began. Section 295(1) is expressed in terms which would cover the dealings after 1 April 2003 in which the defendant engaged. And s.295(7) provides a defence to a charge under s.295(1)(a) of doing anything or causing anything to be done, with the intention that, or being reckless as to whether, it has, or is likely to have, the effect of creating a false or misleading appearance of active trading in securities traded on a relevant recognized market. To take advantage of the defence provided for by s.295(7), a defendant must prove that he had no purpose of creating a false or misleading appearance of active trading in securities, a defence which would clearly not have been available to the appellant if she had been charged under s.295 in this case. Further, there was, under charge 2, the additional element of awareness on the part of the defendant that the course of conduct could prejudice the economic interests of BOC which was, of course, an essential element in conspiracy to defraud but not in an offence under s.295(1).
Beyond the parameters of the present case, the inter-relationship of the common law offence and the statutory offence with its defence available only where the statutory offence is charged, but not where conspiracy to defraud is charged, can give rise to important questions of prosecution policy and conduct. It goes without saying that it would be improper to resort to the common law offence simply for the purpose of depriving a defendant of the defence available under s.295(7). But there is nothing in s.295, or in its statutory predecessor, which goes to show that the common law offence infringes the principle of legal certainty.
It is argued that the width of the offence opens the way to the oppressive exercise of powers by way of arrest and prosecution by law enforcement agencies. Oppressive exercise of powers by such agencies is inconsistent with the guarantee of freedom from arbitrary arrest mandated by the Basic Law and the BOR. The answer to this argument lies not only in the formulation of appropriate policies and practices to be applied by law enforcement agencies and prosecutors, but more importantly in the due administration of the judicial trial process, including the courts’ jurisdiction to stay proceedings for abuse of process, and in judicial review.
The appellant’s case of overbreadth raises different considerations. Ms Montgomery QC acknowledges that vagueness and overbreadth are two distinct concepts – the intended breadth of a law may be clear yet its application may be overbroad. The case of overbreadth rests on the proposition that in order to justify the width of the offence it must be shown that there is a need for a law of that width and that its width is not disproportionate to the legitimate object which the law seeks to achieve.
This argument, though it is said to be based on Canadian authorities (see, for example, Committee for the Commonwealth of Canada v Canada (1991) 77 DLR (4th) 385 at 433; R v Heywood  3 SCR 761 at 790-793), seeks to bring into play “the proportionality” principle which has been applied by this Court in constitutional cases where the issue was whether a restriction on a fundamental right or freedom could be justified - Leung Kwok-hung v HKSAR (2005) 8 HKCFAR 229 at 252C – 254G.
There is no basis in either authority or principle for using the proportionality test as a criterion of legal certainty - R v Heywood at 790-793 (“Overbreadth has no autonomous value under the Charter”). That test was not applied in Shum Kwok Sher. Nor was it applied in the decisions of the European Court of Human Rights on the principle of legal certainty. And it was not applied in R v Rimmington (the public nuisance case). Nor does Committee for the Commonwealth of Canada v Canada, a case relied upon by the appellant, support the use of the proportionality test otherwise than in the context of constitutional cases where a restriction on a fundamental right or freedom is in issue.
THIS CASE IS NOT GOVERNED BY EXISTING AUTHORITY?
Although the arguments sought to draw support from the decided cases, in particular Shum Kwok Sher, Hashman v United Kingdom (2000) 30 EHRR 241 (the contra bonos mores case) and Rimmington, each of these decisions rested on the application of established principles to the characteristics of the particular laws and order in question. There is nothing in the subject matter of these cases which would enable a court to treat the application of relevant principles in those cases as requiring or influencing a decision one way or the other in the present case. On the other hand, the decision of the Divisional Court in Norris v Government of the United States of America  2 All ER 29 rejected an argument that the offence of conspiracy to defraud infringed the principle of legal certainty. The argument in that case rested on grounds more limited than those presented to this Court. Moreover, the decision is under appeal to the House of Lords.
CONCLUSION ON THE ISSUE OF UNCERTAINTY
For the foregoing reasons the appellant’s submissions directed to showing that the offence of conspiracy to defraud infringes the principle of legal certainty must be rejected.
RULES OF PRACTICE OR POLICY?
Ms Montgomery QC submits that this Court should formulate rules of practice or policy to govern prosecutions for conspiracy to defraud in cases where specific statutory offences have been enacted governing the conduct in question. The argument is that the court should adopt an approach similar to that approach stated by Lord Bingham of Cornhill in Rimmington  1 AC at 479E–G.
It is sufficient for this Court to say that, as a matter of good practice, conduct falling within the terms of a specific statutory provision should be prosecuted under that provision rather than under conspiracy to defraud unless there are circumstances justifying the latter course. Such a rule of good practice accords with what was said in Rimmington and by the High Court of Australia in The Queen v Hoar (1981) 148 CLR 32 at 38.
QUASHING THE CONVICTION?
Ms Montgomery QC submits that if the court were to arrive at a conclusion on the essential elements of fraud which are inconsistent with the basis on which the appellant was convicted, her conviction should be quashed. There is nothing in the statement of the law of conspiracy to defraud contained in these reasons for judgment which is inconsistent with the basis on which the appellant was convicted.
It was also suggested that the conviction might be quashed if it appeared that the prosecution failed to conform to appropriate standards of practice and procedure said to be applicable in cases of this kind, notably the suggested principle that a charge of conspiracy to defraud should not be brought where the overt acts relied upon in proof of the conspiracy consist of conduct in relation to which specific statutory provisions have been enacted. The rule of good practice stated in para.83 is subject to the existence of circumstances justifying a different course. In this case, there were justifying circumstances: see para.73 above.
The certified question should be answered “Yes”. The appeal should be dismissed.
Chief Justice Li
The Court unanimously dismisses the appeal.
 See Hong Kong Law Reform Commission, Report on the Creation of a Substantive Offence of Fraud, 1996, paras 3.1 to 3.29; UK Law Commission, Report on Criminal Law: Conspiracy to Defraud, 1994, Law Com No.228; UK Law Commission, Report on Fraud, 2002, Law Com No.276, Pt.3; Report of the Model Criminal Code Officers Committee (Australia), 1997, Ch.3.
 See, for example, Glanville Williams Textbook on Criminal Law, 1978, pp 637, 641 – 642; Smith & Hogan: Criminal Law, 11th ed., Ch.12; D. Ormerod, “The Fraud Act 2006 – Criminalising Lying?”  Criminal Law Review 193.
 Sunday Times v United Kingdom (No.1)(A/30)(1979-80) 2 EHRR 245; at 271 para.49; Hashman v United Kingdom (2000) 30 EHRR 241, para.31; S.W. v United Kingdom (A/335–B)(1996) 21 EHRR 363, para.36/34.
Adams v R  2 Cr. App. R. 295
Brutus v Cozens  AC 854
Committee for the Commonwealth of Canada v Canada (1991) 77 DLR (4th) 385
Hashman v United Kingdom (2000) 30 EHRR 241
HKSAR v Law Kam Fai  2 HKLRD 879
HKSAR v Chan Chun Hong  1 HKLRD 346
Leung Kwok-hung v HKSAR (2005) 8 HKCFAR 229
Macleod v The Queen (2003) 214 CLR 230
Noise Control Authority v Step In Ltd (2005) 8 HKCFAR 113
Norris v Government of the United States of America  2 All ER 29
Peters v The Queen (1998) 192 CLR 493
R v Withers,  AC 842
R v Bassey (1931) 22 Cr. App. R. 160
R v Woollin  1 AC 82
R v Scott  AC 819
R v Ghosh  1 QB 1053
R v Rimmington  1 AC 459
R v Secretary of State for Trade and Industry, ex p. Ford (1984) 4 Tr & L 150
R v Heywood  3 SCR 761
Shum Kwok Sher v HKSAR (2002) 5 HKCFAR 381
The Queen v Hoar (1981) 148 CLR 32
Welham v Director of Public Prosecutions  AC 103
Wai Yu-tsang v The Queen  1 AC 269
Sunday Times v United Kingdom (No.1)(A/30)(1979-80) 2 EHRR 245
Hashman v United Kingdom (2000) 30 EHRR 241
S.W. v United Kingdom (A/335–B)(1996) 21 EHRR 363
Basic Law: Art.28, Art.39
Hong Kong Bill of Rights: Art.5, Art.11
Forgery Act 1913: s.4
European Convention on Human Rights and Fundamental Freedoms
Theft Ordinance, Cap. 210: s.16A
Crimes Ordinance, Cap. 200: s.159E
Securities Ordinance, Cap. 333: s.135
Crimes Ordinance, Cap. 200: s.159C
Securities and Futures Ordinance, Cap. 571: s.295
Authors and other references
Hong Kong Law Reform Commission, Report on the Creation of a Substantive Offence of Fraud, 1996
UK Law Commission, Report on Criminal Law: Conspiracy to Defraud, 1994, Law Com No.228
UK Law Commission, Report on Fraud, 2002, Law Com No.276
Report of the Model Criminal Code Officers Committee (Australia), 1997, Ch.3.
Glanville Williams, Textbook on Criminal Law, 1978
Smith & Hogan: Criminal Law, 11th ed.
D. Ormerod, “The Fraud Act 2006 – Criminalising Lying?”  Criminal Law Review 193
Clare Montgomery QC, Ronny F.H. Wong SC and W.K. Po (instructed by Messrs W M Lo & Co) for the appellant
Kevin P Zervos SC and Winnie Ho (Department of Justice) for the respondent
all rights reserved