Online trading major Interactive Brokers cannot escape a lawsuit launched by one of its ex-customers at the Connecticut District Court. On September 30, 2019, Judge Alvin W. Thompson signed an order denying the brokerage’s motion to dismiss the case brought by Robert Scott Batchelar. Batchelar has brought a claim against Interactive Brokers, LLC alleging that its trading software was negligently designed, which resulted in an automatic liquidation of the positions in his account that cost him thousands of dollars more than it should have. The plaintiff is also suing Interactive’s parent company, Interactive Brokers Group, Inc. (“IBG”), under a theory of respondent superior, and Thomas A. Frank, an officer of IBG, claiming he was personally responsible. The defendants have moved to dismiss these claims but yesterday Judge Alvin W. Thompson nixed their attempt. Batchelar was a customer of Interactive starting in August 2011. He had a margin-trading account with Interactive. Interactive’s computer software continuously and automatically determines whether there is a margin deficiency in an account. Once it determines a margin deficiency, the software locks the account, cancels all pending trades, and prohibits the customer from depositing additional collateral or ordering particular trades to cure the deficiency while it liquidates the positions. On August 24, 2015, Interactive’s software declared a margin deficiency in Batchelar’s account and began liquidating his positions. All that Batchelar held in his account at that time were positions in a security called “SPX put option”. Batchelar had short-sold these positions, so as the sale price went higher, he lost more money. After declaring a margin deficiency, the software began liquidating Batchelar’s positions. It started at 10:11:15 A.M. and ended at 10:31:37 A.M. In that time, Interactive’s software made fifty-one trades at prices ranging from $5.00 to $83.40 per unit. At one point, during a nineteen-second period, the software executed eight trades at prices ranging from $7.00 to $83.40 per unit. This was higher than the going market price for the securities at the time of the sale. Batchelar claims that those transactions disproportionate to the market cost him somewhere between $95,145 and $113,807. In his Second Amended Complaint, Batchelar alleges that the auto-liquidation was “the result of negligent design, coding, testing and maintenance.” He alleges that the programming flaws were the result of Interactive’s failure to meet industry standards in its design and testing of the software and its failure to include certain instructions in the algorithm. The defendants assert that Interactive’s duties are defined exclusively by the Customer Agreement, and that the Customer Agreement expressly authorized Interactive to liquidate the positions in a margin-deficient account, as well as that under the Customer Agreement Interactive has “sole discretion to determine the assets to be liquidated and the order/manner of liquidation.” The court concludes that the expectations of the parties favor finding that a duty to use care exists in this case. Although Interactive is correct that the regulations and the Customer Agreement permit it to liquidate the positions in a margin-deficient account in its sole discretion, that does not mean that the parties would normally expect that Interactive has the right to liquidate the positions in the account in a negligent manner. Moreover, the fact that Interactive has been found liable, in a FINRA arbitration, to at least one customer for negligence suggests that the parties would expect that Interactive would be liable for the negligent execution of trades. Further, after assessing a number of public policy factors, the Court concluded that these factors weigh in favor of finding that Interactive had a duty to use care as to Batchelar. The Court determined that Connecticut courts would recognize an independent extracontractual duty owed by Interactive to Batchelar to use care in designing and using the auto-liquidation software. Thus, the motion to dismiss was denied with respect to the claim against Interactive. Batchelar seeks to hold Frank, who is an officer of IBG, personally liable for negligence in the design, testing, programming, and maintenance of the software. After weighing a set of public policy factors, the court concludes that they weigh in favor of finding that Frank owed a duty to Batchelar to use care in designing, testing, programming, and maintaining the software. Hence, the motion to dismiss was denied with respect to the claim against Frank. Finally, Batchelar claims that Interactive Brokers Group’s liability is derivative of Frank’s liability, as he is IBG’s employee. The court concluded that since Batchelar has adequately pled a claim of negligence against Frank, the motion to dismiss has to be denied with respect to the claim against IBG.
I don't quite understand the rational of the court in respect to Frank's alleged liability. Why would that individual sign responsible for designing, testing, and maintaining of order algorithms and the liquidation of the positions. Even if he is overseeing a certain technical group why would a lawsuit point to an individual and not at wrongdoing of the entity as a whole? In regards to liquidation, the only problem for IB I see here is if the liquidations caused market impact and if the fills indeed were away from market prices at the time of liquidation. A strange case. Sounds more than anything like the overreaching of a clueless judge who does not understand the issue at hand and needs to make an example of a broker. It should be incredibly easy to determine how far the liquidation prices were from the bid offer and whether the orders caused market impact. If that was the case then it should also be a matter of a few hours to calculate the damage caused. Given this suit takes this long when all the facts and trades are well documented I assume the judge here has zero clue. Welcome to America's awesome legal system.
I would be curious to know what algo was used. Maybe Frank was the one actually liquidating using his choice of algos available and didn't try to minimize the impact. I don't understand how IB can't simply call the client and force him to liquidate himself when significant amount is involved.
I actually feel for Frank. As you may know, i flew too close to the sun a few weeks ago. It costed me a full weekend sans sleep: https://www.elitetrader.com/et/threads/ib-close-out-advisory.335975/ I have subsequently looked into partitioning my account to constrain wipe-outs of this magnitude only to learn that section 11D of the customer agreement cross collateralised all of one's accounts: https://www.elitetrader.com/et/thre...t-operational-risk.336032/page-2#post-4928200 Operational risk is on the top of my mind and am spending a lot of time looking for improvements on this front. For every Frank there are probably 1000s more waiting to see where this suit goes before launching a class action.
How can you control the liquidation algo? from what I understand it is a VWAP that kicks in 15 minutes after market open. In the hope this suit will corner IB into forcing the liquidation of the most liquid securities by order of priority....
IB does not have to and should not have to call the client. It's for the best of all other clients and also its agreed upon in the contract. I had one forced liquidation with in in over 15 years with them and if I recall correctly then ib only liquidated a small portion of the positions until margin requirements were satisfied. Only would liquidation occur again if the account dipped again below margin requirements. The only thing that I believe might have happened is if an algo went nuts and dumped the entire position and that this in turn impacted the market. Should be easy to figure out in court.
What have client order types to do with this issue? IB runs its own proprietary liquidation algos that they surely don't disclose how they work. But I like the idea on IB's side of liquidating the most liquid positions first
and how do you do that? I thought that Adaptive Algos were for execution only and liquidation algos were 100% discretionary
As many of you know, options can get super wide in a fast market. I looked at the data for 8/24/15 and the SP was down 100 points (about 5%) at one point, so obviously his short puts were getting hammered. When these automated liquidation algorithms start running, they just dump market orders out there. It's not hard to watch a B/A spread go from $4.50/$6.00 to $0.15/$85.00 in a millisecond as the computers add/remove orders. I've always felt like that was the unfair part of the entire process - that you could have a contract worth maybe $5.25 and it gets liquidated at $64.00 or something unbelievable. I'm not saying that happened here, but it seems possible.