Burford Capital announces the preliminary findings of its analysis of the trading of its shares last week. We believe that trading shows evidence consistent with illegal market manipulation.
While Burford continues to analyse the data, it has made regulatory authorities and criminal prosecutors aware of these preliminary conclusions and Burford is considering its own options. Burford has retained the law firms of Quinn Emanuel Urquhart & Sullivan LLP, Freshfields Bruckhaus Deringer LLP and Morrison & Foerster LLP in connection with these matters.
A forensic examination by Burford and its expert1 of the detailed trading data made generally available by the London Stock Exchange for 6 August (the day on which Muddy Waters tweeted about a forthcoming – but unidentified – short target) and 7 August (the day on which Muddy Waters released its short attack on Burford) discloses trading activity consistent with material illegal activity.
1 The analysis summarized in this release was predominantly conducted by Professor Joshua Mitts of Columbia University, who specializes in the analysis of market data surrounding short attacks such as this. See, e.g., https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3198384. Burford’s and Professor Mitts’ analysis is ongoing, and it is possible that the final results will differ, perhaps materially, from these preliminary findings. However, Burford’s view is that it is important for the market and investors to be appraised of these preliminary results even though they are preliminary and may later be refined or amended.
Spoofing and layering
Spoofing is the placement of a high volume of trading orders at a price equal to or better (i.e., lower) than the best-bid-best-offer price and subsequently cancelling these orders to move the price in a given direction without actually concluding any trades. For example, consider a stock where the current best offer is £9.99 per share. A spoofer might place a high volume of sell orders at £9.98, causing the best offer to decline to £9.98, immediately cancel those sell orders before they can execute, and then place a high volume of new sell orders at £9.97. The strategy of repeatedly placing and cancelling sell orders at or below the best offer without actually selling any shares artificially drives down the share price.
Layering is similar to spoofing, except that instead of placing and cancelling a high volume of orders at the best offer price, the manipulator places these orders deeper in the order book, at prices above the best offer. Continuing the prior example, suppose the manipulator worries that the artificially “spoofed” sell orders at £9.98 will be inadvertently executed before they can be cancelled. Instead of placing these orders at £9.98 (or £9.99, the original best offer), the manipulator may place a high volume of orders at £10.01, £10.05, or some other price slightly above £9.99. These orders are virtually certain not to be filled but they affect pricing by suggesting falsely that there is a large volume of shares for sale.
Spoofing and layering are both illegal and have resulted in criminal convictions in the past. As an example, layering led to the 2010 “flash crash” when the Dow Jones Industrial Average fell 600 points in five minutes; the perpetrator was found guilty of fraud.
6 August trading following the Muddy Waters tweet
On 6 August, in the several hours following the 13:30 release of the Muddy Waters tweet about a forthcoming short attack, almost £90 million of sell orders were placed and cancelled without being filled – for a stock whose average trading volume for an entire day was less than one-fifth that amount. As discussed above, that trading conduct is consistent with illegal market manipulation. Moreover, during five one-minute periods on 6 August (14:17, 14:30, 14:35, 14:43, 14:45), Burford’s shares fell 6%, or over £170 million in value, some of its sharpest declines of the day. During these periods, executed sell orders totaled a mere £186,000. That mismatch between price movement and executed orders is consistent with market manipulation. To show the impact we believe this behavior had on Burford’s shares, compare five other minutes of trading on 6 August with more significant levels of actually executed orders: during the one-minute periods of 14:07, 14:59, 15:14, 15:15 and 15:43, executed orders totaled £1.5 million – eight times as much as the periods identified above – and the price of Burford’s shares rose by nearly 2% during those five one-minute periods.
7 August trading around and following the Muddy Waters report
On 7 August, a day on which over 28 million Burford shares traded, Burford’s share price suffered its greatest declines over just ten single minute periods with very low volumes of executed sales and very high volumes of cancelled sales orders. Indeed, Burford’s share price declined by a full 60% over those 10 one-minute periods even though only 739,724 shares were actually traded – around 0.3% of Burford’s shares.
The following table shows the ten one-minute periods during which the price of Burford’s shares fell the most on 7 August, as well as the number of sell orders2 created, cancelled and executed.
2 Figures for orders reflect number of shares underlying those orders. Rows in the table may not sum given the inherent nature of isolating single minutes of trading.
||Price Decline (approx.)
||# of Sell Orders Created
||# of Sell Orders Cancelled
||# of Sell Orders Executed
These ten minutes collectively reflect a 60% decline in the price of Burford’s shares. Yet it is striking how few shares were actually sold over these windows.
For example, at 08:53, the minute which saw a 7.6% price decline, there were only 27,885 shares actually sold – less than 10% of the number of shares underlying the orders created, 291,364. It strains credulity to believe that a decline on the order of hundreds of millions of pounds in market capitalization was driven solely by actual trading amounting to a few hundred thousand pounds absent market manipulation.
To see why in our view it is unreasonable to conclude that the decline in Burford’s share price was driven by actual sales of shares, simply compare 08:53, when 27,885 shares were sold and the share price declined 7.6%, with 11:22, when more than twice as many shares were sold but the share price decline was just over half in magnitude. What accounts for the difference between 08:53 and 11:22? The former had nearly 100,000 more orders for shares cancelled than the latter – suggesting that it is the volume of sell-side cancellations, not executions (actual trades), that drove down the price.
The Muddy Waters delayed tweet
A large wave of sell order cancellations arrived in the few minutes preceding Muddy Waters’ first tweet on 7 August identifying Burford Capital as the target of its short attack.
On 6 August, Muddy Waters posted a tweet stating that at “8 am London time” the next day “we will announce a new short position on an accounting fiasco that’s potentially insolvent and possibly facing a liquidity crunch.”
An examination of Twitter microdata shows that Muddy Waters’ first tweet on 7 August actually announcing the identity of Burford Capital as its victim was posted at 08:53:48, a delay of 53 minutes and 48 seconds from the previously promised time of 08:00.
Muddy Waters did not publicly reschedule its announcement to 08:53 or otherwise provide any public indication of the actual timing of the report’s release.
The LSE data show that an unusual flood of sell-side cancellations began to arrive in the three minutes immediately preceding this tweet. Specifically, from 8:50am-8:53:47am, sell orders totaling over 578,000 shares were cancelled – approaching Burford’s entire regular daily trading volume in less than four minutes. In fact, during those three minutes and forty-eight seconds before the tweet, there were 578,112 shares worth of cancelled sell-side orders as compared to only 36,597 shares sold in actual executed sales, a ratio of 15.8 to 1. We currently see no non-manipulative explanation for that market phenomenon.
Given that Muddy Waters made no public announcement about the actual timing of the release of the report, we do not see why a legitimate market participant without knowledge of the actual tweet’s expected release time and content would be placing and cancelling a large number of sell orders in the three minutes before the release of the tweet.
It is also worth noting that according to FCA data Muddy Waters initiated its short position in Burford on 5 August, the day before its teaser tweet about a report being forthcoming, reduced its short position by 20% on 6 August, the day of the teaser tweet, and exited a further 63% of the position on 7 August, the day of the report’s release. In other words, while Muddy Waters was suggesting that Burford was insolvent, it was at the same time buying Burford shares.
Twitter and algorithmic trading
Muddy Waters asserted in a tweet and in its report that Burford was “arguably insolvent”.
There is no factual basis for that statement and we debunked it in our rebuttal to the report. To reach it, Muddy Waters needed to engage in an entirely contrived and unrealistic analysis that we do not believe it has defended or reiterated since.
Burford is informed that posting certain phrases, such as “insolvent”, on Twitter can induce an algorithmic sell off in a stock, and that a wider analysis of short seller tweets in general shows that references to “liquidity risks” and “insolvent,” which appeared in Muddy Waters’ tweet at 08:56, lead to sharp price declines. Statistically speaking, these terms jointly increase the odds of an artificial price decline in a manner that is highly unlikely to have been caused by random chance.
Burford is continuing its analysis and considering next steps. Burford also intends to continue to monitor trading activity going forward and will refer any further evidence it believes to be indicative of manipulation to the appropriate prosecutorial authorities.
Christopher Bogart, Burford’s CEO, commented: “We are committed to working with investors to address any questions and concerns they may have about Burford and its business and we are grateful for their feedback. Burford’s market-leading business today is the same as Burford was a week ago. What has changed is that a substantial amount of market value was wiped out by activity we believe is consistent with illegal market manipulation that has nothing to do with Burford’s business. That is wrong and that is illegal.”
For further information, please contact:
|Burford Capital Limited
|Elizabeth O’Connell, CFA, Chief Financial Officer
||+1 212 235 6825
|Macquarie Capital (Europe) Limited – NOMAD and Joint Broker
||+44 (0)20 3037 2000
|Liberum Capital Limited – Joint Broker
||+44 (0)20 3100 2222
|Numis Securities Limited – Joint Broker
||+44 (0)20 7260 1000
|Montfort Communications Limited – Financial Communications
||+44 (0)20 3770 7908
|Robert Bailhache – email
About Burford Capital
Burford Capital is the leading global finance and investment management firm focused on law. Its businesses include litigation finance and risk management, asset recovery and a wide range of legal finance and advisory activities. Burford is publicly traded on the London Stock Exchange, and it works with law firms and clients around the world from its principal offices in New York, London, Chicago, Washington, Singapore and Sydney.
For more information about Burford: www.burfordcapital.com
This release does not constitute an offer of any Burford fund. Burford Capital Investment Management LLC (“BCIM”), which acts as the fund manager of all Burford funds, is registered as an investment adviser with the U.S. Securities and Exchange Commission. The information provided herein is for informational purposes only. Past performance is not indicative of future results. The information contained herein is not, and should not be construed as, an offer to sell or the solicitation of an offer to buy any securities (including, without limitation, interests or shares in the funds). Any such offer or solicitation may be made only by means of a final confidential Private Placement Memorandum and other offering documents.