We take a close look at the Market Misconduct Tribunal's report on Mr Left's erroneous criticism of China Evergrande (3333), and what that says about free speech amongst participants in the HK market.

Andrew Left and the right to be wrong
23 October 2016

Is it market misconduct for unlicensed members of the public to be stupid, ignorant or just plain wrong when commenting on listed companies? We say not, but the Market Misconduct Tribunal (MMT) has taken the alternative view, finding that Andrew Edward Left (Mr Left), operating as Citron Research, disclosed false or misleading information inducing transactions and so engaged in market misconduct when criticising China Evergrande Group (Evergrande, 3333) in a report he published on 21-Jun-2012.

In the introduction web page for the report, Citron said:

"This research and analysis, compiled over several months, presents the conclusion that HK:3333 is essentially an insolvent company that has consistently presented fraudulent information to the investing public".

Citron purported to "prove this conclusion" in the report. Evergrande's response to this was swift. A denial was published at 12:48 on 21-Jun-2012, stating that the allegation was untrue, followed by a more detailed rebuttal at 11:56 the next day, which stated that it "will consider taking legal actions against Citron Research". So far, it has not sued Mr Left for libel or defamation.

The tribunal was chaired by Justice Michael J Hartmann with 2 lay members, Mr Kwok Kam Hoi and Dr Michael Wong Chak Sham. Open the MMT report dated 26-Aug-2016 and we will guide you through it in this article.

Last week, the MMT gave an oral ruling on sanctions, ordering that Mr Left be banned from trading securities in HK for the maximum period of 5 years, disgorge HK$1,596,240 in short-selling profits, pay the SFC's investigation and legal costs and cease and desist from such misconduct in future. Failure to comply with the order is an offence. A written version of the ruling is awaited (update: here it is, dated 10-Nov-2016). Mr Left has reportedly told Reuters that he plans to appeal to the Court of Appeal. This right is provided under Section 266 of the SFO.

Racy tabloid

Mr Left's style of reporting is not one that Webb-site would ever use - the MMT called it a "racy 'tabloid' format" and a "hard-hitting 'tabloid' style", using "direct, plain, Tabloid language". Tabloid is not our style, we're more of a "detailed broadsheet" in our approach. Nevertheless, there is room in this world for both the New York Post and the New York Times, the Sun and the Times of London. Anyone reading Mr Left's report would, in our opinion, be clear that Mr Left was expressing his opinions and making allegations, not stating outright facts. They would also be clear that he had no inside information - his analysis, with all its mistakes, was based on publicly available information which was extensively quoted in the report. He wasn't claiming to have some inside scoop.

Search the Citron report for the words "suggest(s)" or "indicate(s)" and you will see numerous instances in which the report was clearly expressing a view or conclusion, not a fact.

If Mr Left had been an SFC-licensed person, then he would have been held to the higher standard of the Code of Conduct in the quality of his analysis and output, but he was not - he was just a member of the investing public, not licensed and not paid by anyone for his opinions. The fact that he expressed his views on a web site under a brand (Citron Research) and has a track record of some good calls in the past, doesn't change that.

When can a statement of opinion be false?

As Mr Left is not SFC-licensed and not providing a paid service, the SFC could not take disciplinary action against him for shoddy analysis or prosecute him for operating without a license. So instead, they pursued him under Section 277(1) of the SFO, alleging disclosure of false or misleading information inducing transactions. The detailed wording of this misconduct requires that a person knows that, or is reckless or negligent as to whether:

 "the information is false or misleading as to a material fact, or is false or misleading through the omission of a material fact"

So what was the "information" that he was disclosing? In our view, the "information" was his analysis and resulting opinion. Everything else in the report was already in the public domain and not "disclosed" by him. But if it is clear that the "information" is his analysis and opinion, then even if that opinion is wrong, how can that "information" be false? After all, professional licensed analysts are often wrong in their analysis that a company's profits will improve, wrong that a stock will go up rather than down, and wrong that you should therefore buy the shares, but they don't normally get taken to the MMT for disclosing false or misleading information.

There's a reason for that. In our view, the "information" that someone holds a certain opinion can only be false if he doesn't actually hold that opinion, that is, he doesn't believe what he is saying. But the SFC, in bringing this case, did not allege that Mr Left did not believe what he was saying, and nothing in the MMT report suggests any evidence was advanced in that direction.

Contrast that with the infamous case of Henry Blodget, the former Merrill Lynch analyst who was barred from the US Securities Industry for life after internal e-mails revealed that he was privately trash-talking stocks that he and his firm were publicly promoting. In that case, the e-mails were a smoking gun that his public statements of opinion were false, because he didn't actually hold that opinion.

The MMT did briefly acknowledge the difference between facts and opinions. Paragraph 85 (ii) of the report states:

"a fact may be said to be an item of verified information, that is, the independent reality of a matter as opposed to an opinion concerning it. It is important to recognise the difference between an asserted fact and a comment concerning it"

Paragraph 116 also states:

"the information must be shown to be false and/or misleading as to a material fact, that is, a fact and not simply an expression of opinion..."

Unfortunately, for the rest of the report, the MMT does not return to this distinction, and the report does not at any point analyse whether Mr Left was simply expressing an opinion when making his allegations. The closest it comes to this is in paragraph 217, where it states:

"Opinions that are critical, however, are to be contrasted with direct accusations going to matters of fact"

Here, in our opinion, the Tribunal is drawing a distinction without a difference. If a person says about a company "I think they are cooking the books", then that is clearly both an accusation and an opinion - it is not a representation of a fact. The allegation could only be determined to be right in a successful prosecution for fraud.

The legislative intent

Section 27 of the Basic Law provides a constitutional right of free speech, and Article 16 of the HK Bill of Rights provides freedom of expression and opinion, including the right to "seek, receive and impart information and ideas of all kinds", mirroring Article 19 of the International Covenant on Civil and Political Rights. Such rights are not absolute. For example, you can be sued for libel and defamation, it is a crime to incite mutiny from the People's Liberation Army (punishable by life in jail), and if you are a corporate insider, you cannot selectively disclose price-sensitive information. If you are an SFC-licensed person, then your speech (including publication of research) is held to the standard of the Code of Conduct and you can be ordered to pay a "pecuniary penalty" (a fine) or lose your license with disciplinary action for sloppy research under Section 194.

However, in our view the clear legislative intent of Section 277(1) was to outlaw disclosure of false and misleading factual information, not incorrect but honestly-held opinions of the general unlicensed public. We have the law of libel and defamation to deal with that, and if Evergrande or companies in similar situations wish to sue, they can. The SFC should not be stepping into their shoes.

Standards of comment

Paragraph 182 of the MMT report states that Paul Anthony Phenix (Mr Phenix) was called by the SFC to give evidence as an expert witness "in matters of accountancy". Fair enough, he is a fellow of the HKICPA and a team member and former director of an audit firm, Baker Tilly Hong Kong Ltd (Baker Tilly). However, in paragraph 234, the MMT cites his opinion on something else - standards of analysis and comment, on which he has no professed expertise. The MMT writes:

"As [Mr Phenix] expressed it, if [a] person seeks to conduct an analysis of the finances of a company listed on the Hong Kong Stock Exchange, he must - for a start - have an understanding of the relevant accounting standards. If a person does not have such an understanding of these...then he should not comment on them. Mr Phenix built on this by saying that, if you are a professional analyst, then the proper thing to do is to go the company and seek clarification."

This is wrong on several levels. First of all, the market is awash with people expressing opinions on blogs, forums and in newspaper columns as well as the higher-quality Webb-site Reports. This again is free speech, part of the lifeblood of free markets. There is no prohibition on people who do not understand accounting standards from voicing their opinions, however wrong they may be. Without free speech, others will not have the chance to put them right by criticising those opinions and voicing their own.

This is particularly the case if a person thinks they have understood the accounting, even if they have misunderstood it. One does not always know when one has misunderstood something. There is no obligation on commentators to appoint accounting experts before commenting, nor to seek comment from companies before criticising or praising them, even for licensed professionals, which Mr Left is not.

There is also no prohibition on boasting (assuming it is true) about a successful track record of past reports, as Citron did. That doesn't mean they will not make mistakes and be wrong on occasion. 

Does the fact that Webb-site Reports has usually been right prohibit us from ever reaching the wrong conclusion through faulty analysis when we criticise companies? If so, we had better stop right now, and so should all other leading commentators, because we don't want to get banned from the market for making a mistake.

That's what makes a market - the constant debate in the court of public opinion, not in a tribunal. We cannot have subjective legal standards in which erroneous analysis and comment by unlicensed members of the public is judged to be "market misconduct" if you have usually been right in the past and your comments tend to move markets, but not market misconduct if you are usually wrong and nobody listens to you. Otherwise commentators will have to withdraw from the market as soon as they become successful enough for their opinions to influence it, leaving behind a market of mediocrity.

Put simply, in the words of Joss Stone, you have the right to be wrong.

The SFC's action and the MMT's finding, unless and until overturned on appeal, have cast a dark cloud over freedom of speech in the Hong Kong markets, wherever the commentator lives.

Insolvency is a matter of opinion

One area which is closer to Mr Phenix's expertise was in paragraph 188, regarding the allegation that Evergrande was "insolvent":

"Mr Phenix commented in his Issue Summary that the allegation was not, judged on the available evidence, a reasonable one to make. Evergrande, he said, was not, in his opinion, "remarkably or comparatively illiquid"." (our italics)

Two things come from this citation. First, Mr Phenix clearly accepts that insolvency is a matter of judgment and opinion, not fact. Secondly and incidentally, the Tribunal and Mr Phenix appear to be conflating solvency with liquidity. Solvency is an ability to eventually meet your obligations - an excess of assets over liabilities, but liquidity is an ability to pay short-term obligations. A person who owns a house with a 50% mortgage, for example, can be solvent but may not have enough cash to pay the electricity bill. Similarly, banks must be both solvent and maintain liquidity ratios, as the Federal Reserve explains here.

Indeed, solvency or insolvency (the central allegation of the report) must always be a matter of opinion, not fact, until the liquidators are called in and we discover whether a company's assets exceed its liabilities or not. Nobody actually knows for sure, certainly not an analyst working on public data of a company that has not gone into liquidation.

The best guide to solvency an analyst has is the last published accounts, if he believes them. Take for example, the audited accounts of EganaGoldPfeil (Holdings) Ltd at 31-May-2006, which received a clean audit report by none other than Baker Tilly. We questioned the accounts in our article on 26-Jul-2007. The stock promptly crashed, was suspended 7 weeks later and was subsequently delisted. The SFC began proceedings against 3 former directors of Egana in 2011 but nothing has been announced since then.

Footnote: probability theory

One of the requirements of Section 277(1) is that the information "is likely.. to maintain, increase, reduce or stabilize the price of securities...". Mathematicians amongst our readers will be entertained by paragraph 162. After reciting the danger of employing hindsight to determine a predictive test (in this case, whether the publication of the Citron report was likely to reduce the price of Evergrande's shares), the judge cannot resist claiming at the end:

"Nevertheless, some limited support for the Tribunal's findings can be found in the fact that the likely impact was born out by the actual impact"

First of all, he means borne out, but our point here is that the actual impact in a single test provides no support whatsoever for an estimate of likelihood. If you had made a single prediction in 2012 that Theresa May will "likely" (more than 50%) be the next the next UK prime minister, the outcome does not provide any support for your probability estimate. It might only have been a 1 in 10 chance. A prediction that she would probably not be the next prime minister might have been more accurate, even though she turned out to be.

To test the MMT's predictive power statistically you would need a large sample of research reports on which to make predictions and test outcomes, something that the tribunal does not have. The MMT did not look at a random sample of public research on stocks, decide which reports were "likely" to move a price and then test its predictive power against the outcomes in each case.

The SFC would (almost certainly) never have brought the case to the MMT if there had not been a sharp decline in the stock price after Mr Left's report was published, so the actual outcome would only have failed to match if the Tribunal had concluded that the report was not likely to affect the price.

So the MMT is tying itself in probability knots. Given that the SFC is only going to bring cases in which alleged false or misleading information was followed by a corresponding price movement, the proper and more realistic test is whether the publication of that information is the likely cause of the movement. We submit that this was (probably) the legislative intent of the words "is likely". The legislature was not concerned with hypothetical situations in which alleged false or misleading information was likely to, but failed to, have any impact.

© Webb-site.com, 2016


Organisations in this story

People in this story

Topics in this story


Sign up for our free newsletter

Recommend Webb-site to a friend

Copyright & disclaimer, Privacy policy

Back to top