Expert witnesses testifying on behalf of the prosecution told the court there were signs that the trading of the three penny stocks was subjected to manipulation and that the prices they were trading at just before the crash of October 2013 were as much as more than 3,000% overvalued.
“If a market is not fair, it can be rigged and if that happens, people will shy away from the market,” said Australia-based Michael Aitken, who was engaged by the Monetary Authority of Singapore (MAS) in September 2016 to analyse the trading pattern of Blumont Group, Asiasons Capital and LionGold Corp shares before the crash.
“And if investors shy away from the market, then effectively you have lower liquidity that, in turn, leads to higher transaction costs for investors and higher cost of capital for companies. So, it is very important that markets are fair,” said Aitken, founder and chief financial architect of the market-trading surveillance system, SMARTS, which is used by bourses such as Nasdaq and the Singapore Exchange. He was previously CEO and chief scientist at the Capital Markets Cooperative Research Centre but has retired from that job and is now working three days a week at Macquarie University.
Aitken was testifying at the long-running trial of John Soh Chee Wen and co-accused Quah Su-Ling, who are facing charges of masterminding the manipulation of the three penny stocks collectively known as BAL. The share prices of these three counters had soared for months before crashing spectacularly in early October 2013, destroying some $8 billion in market value.
Aitken explained SMARTS uses a series of automated algorithms to scan all the orders and trades of a stock exchange to identify potentially problematic behaviour like insider trading or market manipulation.
From his analysis, Aitken believes that there was a very heavy proportion of so-called “wash trades” where the participants were essentially trading with themselves and are not a reflection of genuine demand and supply.
According to his analysis, the volume of wash trading, as a percentage of the total trading volume of all participants in the market over the whole relevant 13-month period, was 42% for Asiasons, 49% for Liongold, and 18% for Blumont.
Soh and Quah allegedly made use of a network of dozens of accounts to manipulate the trading. The court had heard earlier from various brokers and trading representatives on how they have had taken trading instructions from the duo or their close associates.
According to Aitken, these so-called “controlled accounts” comprised 60% of the trading volume for Blumont and 88% and 90% respectively for Asiasons and Liongold during the period in question.
“The sheer volume of the wash trades was such that the markets for all three counters were not subject to the genuine forces of demand and supply, throughout the relevant periods. This created a market that was false and misleading in each of the three counters, throughout the Relevant Periods,” said Aitken.
In addition to detecting the prevalence of wash trades, Aitken discovered other forms of unusual trading behaviour. One of the algorithms described by Aitken looks for stocks whose prices were being pushed up by buyers in a very short time. Typically, an ordinary buyer would key in a price and wait for sellers. If this fails to attract buyers, he might then put in a higher bid to entice sellers to bite.
“Typically, you might expect them to wait five or 10 minutes, particularly if it is early in the day. But if it was, say, getting close to the end of the trading session, you might expect them to be a little bit more aggressive,” said Aitken.
What Aitken found was that around a third of BAL trades could be classified as such. “What we are looking for here is people who appear to be picking up BAL stocks at higher and higher prices, or appearing very impatient. And we look for repeated patterns of this behaviour,” he told the court.
Another algorithm searched for so-called “uneconomic trading”, where instead of buying low and selling high, the opposite seemed to be happening. “We calculate the volume-weighted average price of buys and sells. When we see the buys appear to be greater than the sells, the surveillance people start asking questions,” said Aitken.
According to his analysis, a very high proportion of Asiasons trades could be considered “wash trades” where the stock changed hands multiple times to send out misleading information to the market. The trades are often executed by parties in collusion.
Aitken notes that wash trades accounted for 21% of the total Asiasons trades in August 2012 and up to 68% in June 2013. “On the assumption that these accounts are associated, we have a significant and unusual amount of wash-trading, which makes this a false and misleading marketplace,” he told the court.
According to Aitken, between Aug 1, 2012, and Oct 3, 2013, the volume of LionGold trades that could be classified as wash trades ranged from 46% to as high as 78%. “So again, my conclusion is the same, based on the fact that I think they controlled demand and supply. And if you can control demand and supply, you affect prices,” said Aitken.
As for Blumont, there were “a significant number of alerts going off”, including 170 days where wash-trading constituted between 12% and 24% of the trades in months leading up to and including September 2013. “There is not quite the control of demand and supply there but it is still very, very significant,” said Aitken, adding also that there were also 15 days of pre-arranged trading, 13 days when “constraining the spread” was observed and 58 days of “uneconomic trading”.
“That is an indication of active trading and therefore the market is false and misleading,” he added.
Rumours at cocktail parties
In his cross-examination, Soh’s defence lawyer N Sreenivasan tried to get Aitken to agree that the SMARTS surveillance system did not take into account many other factors that affect the price movements of a stock.
These included macroeconomic events as well as market sentiment, rumours and reports by analysts.
“Some CEOs or major shareholders also talk up their stock, meaning they’ll make sure they are giving out news of expansions and saying how good their share is at every cocktail party. That is also an occurrence in the market, am I right?” asked Sreenivasan.
“It used to be but it is becoming less common these days because of the problem of providing information to one or two parties rather than a whole group of parties at the same time,” replied Aitken.
“Have you come across analyst briefings where a company calls a group of stockbrokers together and tells them what it plans to do?” asked Sreenivasan.
“Yes, we used to have those but they are less common these days,” replied Aitken.
“Well, they may be less common in Australia, Professor Aitken but are you aware they continue to still exist in Singapore?” asked Sreenivasan.
“No,” said Aitken.
“So, that too will have an impact on price and volume, am I right?” asked Sreenivasan.
“Yes, it could,” replied Aitken.
Sreenivasan also argued that Aitken’s analysis did not take into account the actions of company insiders. “If I am the non-executive chairman of a listed company and I tell all my friends I am buying more shares in that company. That would generally raise confidence in the company, am I right?” asked Sreenivasan, who, outside the courtroom, is also the non-executive chairman of Q&M Dental Group.
“It could, yes, but it could also depend on the reputation of the company and the reputation of the person. But it could, yes,” replied Aitken.
“Okay, it depends whether people believe I’m genuinely increasing my investment or people think I’m just talking up the price for some nefarious purpose, right?”
“Yes,” said Aitken.
“On the other hand, if senior members of the company, major shareholders, and directors and board members, sell huge amounts of shares, that would undermine confidence in the company, am I right?”
“Yes, if that is known publicly, they would,” said Aitken.
“So just to run through, in addition to not considering macroeconomic factors or rumours, you have not considered the activities of people influential in the affairs of these three companies, am I right?” asked Sreenivasan.
“Well, I did look for financial analyst statements similar to what you are talking about but I couldn’t find any,” said Aitken.
“Would I be correct to say that apart from these factors, sometimes the sentiment just depends on who the people involved are? For example, Elon Musk said he is buying bitcoin. Just because it is Elon Musk, everybody thought it was a big deal. Agreed?” asked Sreenivasan. “Yes,” said Aitken.
“Even though his expertise or business activity is in electric cars and spaceships?”
“Well, it’s a bit broader than that but I get the point, yes,” said Aitken.
Based on experience, not on research
In his cross-examination, Quah’s lawyer Phillip Fong similarly pointed out potential blind spots in Aitken’s analysis. For example, he got Aitken to agree that the 10 algorithms applied to analyse the trades were created based solely on his own experience and not on any scientific research.
“That is correct. There is very little of such work,” said Aitken.
“So, your algorithms are based on your subjective view of how the markets work, am I right?” asked Fong.
“Well, yeah, like how you would go about trying to detect unusual situations. Yes,” said Aitken.
“And your algorithms, in this case, are based on your subjective view of how the Singapore stock exchange works, correct?” asked Fong.
“The algorithms are not necessarily based on how the SGX works but they are based on my collective experience of over 30 cases of looking at these types of activities over time,” said Aitken.
“But would you say that you do understand how the Singapore exchange works?” asked Fong.
“Yes,” said Aitken.
Blumont overvalued by 3,112% on eve of crash
The share prices of the three penny stocks, Blumont Group, Asiasons Capital and LionGold Corp at the eve of the crash in October 2013 were “wholly out of proportion to true value on any basis”, according to expert witness John MH Ellison, who is a senior managing director in FTI’s Economic and Financial Consulting team.
As of Oct 1, 2013, Blumont was trading at 3,122% of Ellison’s estimated fair value, 1,514% for Asiasons, 1,514%; and 464% for LionGold.
Ellison, who has around 50 years of valuation and accounting experience, including 13 years as chair of KPMG’s forensic accounting division, based his valuation on a sum-of-the-parts approach, given how the three companies then all held various stakes and assets in different businesses and industries.
For example, Blumont held shares in both quoted and unquoted minority shareholdings without engaging in significant operations itself. It also owned investment properties classified as held for sale and it operated a sterilisation business via its 55.4% owned Indonesian subsidiary PT Rel-ion Sterilization Services.
On the other hand, Asiasons owned subsidiaries in fund management, shares of other companies, a financial advisory business and 70% in an entity called Hub Media Group, which is a content distribution company selling to broadcasters.
Meanwhile, LionGold, besides owning numerous stakes in gold-mining companies, was still holding on to its legacy office equipment-making business with various entities incorporated across China, Hong Kong and Macau.
According to Ellison, his valuations of the three stocks showed they may be “overstated”, given how LionGold was generally loss-making while Blumont and Asiasons became loss-making in the quarter to June 30, 2013; Asiasons’ segments in financial advisory and media were already loss-making in 2012 as was its investment in US energy company Black Elk in 2013. LionGold’s core segments in office equipment and mining were also facing challenging environments, states Ellison.
“Given the vast differences between my valuations and the actual stock market prices, any justifiable minor differences of valuation opinion would anyway not affect my broad conclusions,” he added.
