On a sunny March morning in 2017, Michelle Leder awoke in her Los Angeles bungalow, turned over in bed, and scrolled through the messages on her iPhone.
One item caught her eye. “Nicolas Elbaze,” the automated email read, “has just reached the limit of free trial views.”
The news was potentially lethal. It was an inkling that Elbaze, a researcher at quantitative hedge fund Squarepoint Capital, might have been seeking improper access to Footnoted.com, the financial website Leder had started 14 years before and had turned into a thriving news and research service.
Elbaze had asked Leder a year earlier for, first, a trial subscription, and then a flat rate for full historical access to reports.
She had refused. Experience had shown her that Footnoted data is fiendishly difficult for quants to format. Firms like Two Sigma Investments, Point72 Asset Management’s Cubist Systematic Strategies, and AQR Capital Management had queried her about subscribing. Leder had even held informal talks with two funds to buy Footnoted outright so they could do the job themselves.
Reluctantly, however, just weeks before the email, she had agreed to provide London-based Squarepoint a trial. Then Elbaze seemed to have ramped up his activity.
“I was just, ‘Holy shit, what’s going on here?’” Leder recalls asking herself at the time. She emailed her developer. “He seems to have downloaded my entire database,” she wrote. “If he did do this, it’s a big BIG problem.”
In fact, Leder estimated that Elbaze had viewed more than 17,000 pages — some of which even paid subscribers couldn’t get a hold of. A forensic investigation commissioned by Leder backed up her assessment.
Eight days after the email alert, Elbaze sent Leder a terse note. “We have investigated Footnoted content and we have come to the conclusion that we would not be able to benefit from it within our research and investment process,” he wrote. “So I’m afraid we won’t take the subscription.”
Leder immediately responded: “I appreciate you letting me know. Given the amount of information that you appear to have taken screenshots of during the trial, specifically on 3/14, I think your fund acted illegally.”
With no response from Elbaze, she escalated the matter.
Leder emailed Squarepoint’s chief compliance officer, Hung Luc, on March 20. “I have never had a legitimate fund abuse the trial access in the way that Elbaze did,” she wrote. “I do not know what Elbaze’s plans are for the data that he blatantly took from my site, but the logs of his activity on the site are crystal clear. At best, he did something highly unethical. At worst, he blatantly stole massive amounts of data from my site.”
Luc wrote back three days later denying that Squarepoint had exceeded any data limits: “Any information consumed was for the purpose of assessing the usefulness of the information to our firm and aligned with the intention of the trial access. We will not be using data from the trial going forward and have purged any data sets that we deemed connected to this trial subscription from our systems.”
Leder asked how Squarepoint could know of any internal limits the site placed on trial subscribers. “Anyone who was to review Squarepoint’s activity on my site would clearly see that what Elbaze did was highly unethical, at best, and illegal at worst,” Leder wrote. “I am reviewing my options.”
Communication between them stopped.
Leder’s anger was understandable. In Footnoted, she had built a vital investor resource that had become her life’s work. Using search algorithms that Leder had written mostly herself, Footnoted researchers scoured Securities and Exchange Commission filings looking for the leading — and often damning — telltale disclosures companies prefer to hide, often by burying them in tiny, italicized footnotes and legal jargon.
The alerts Footnoted sent out to its subscribers highlighted subpoenas, atrociously high executive compensation, and dubious accounting tricks that could make or save investors millions. Clients ranged from short-seller John Hempton of Bronte Capital to regulators to Frankfurt-based Acatis Investment, a $10 billion value-oriented stock picker.
There were accolades for Footnoted and its founder. “Michelle Leder unearths nuggets that companies bury,” wrote Kiplinger. “Nobody’s better,” crowed CNBC. And Condé Nast Portfolio magazine: “Michelle Leder is a national treasure.”
But if Footnoted’s host servers had been accessed, as Leder feared, her critical algorithms were at risk too, perhaps even worthless — an existential threat to her entire business model.
“I felt like I had been violated,” says Leder. “They were tearing apart more than a decade of my life’s work. It was my information; it was my baby.”
Today, Footnoted exists in a kind of suspended animation — like a lot of the world in this era of coronavirus. The website has not been updated since January, when Leder posted a recap of 2019 filing developments. In truth, precious little new material has been added since 2018.
But despite the withering of Footnoted that was to come, Leder’s assertion to Luc that she was reviewing her options wasn’t idle. She soon hired a U.K. law firm, Temple Bright, to represent Footnoted.
In an April 27, 2017, letter to Squarepoint, Temple Bright wrote: “Your acts are entirely disrespectful, unethical, and, as stated above, illegal.” On behalf of Footnoted, it sought an apology, details of where any Footnoted data was stored, evidence of the data’s destruction, and money — a lot more money than a Footnoted subscription would have cost.
Squarepoint’s London-based lawyers, Simmons & Simmons, responded on May 17, denying any wrongdoing by the fund. “Our client did not circumvent your client’s system as alleged (or at all),” it stated. “Your references to criminal offenses having been committed are vague and entirely unsupported.”
The firm went on to deny that Squarepoint had transferred any Footnoted data or content to a third party — something Leder had not alleged — and hinted at unspecified retaliation against her should she dare to publicize any of her allegations. It concluded: “Our client reserves all its rights to take action against your client if your client attempts to make any vexatious or false allegations about this matter publicly.”
Since the early days of the internet, Leder has been a techie and an early adopter. She jokes about being the first person fired by email — back in 1998 — and built Footnoted around search programs she wrote herself. Nevertheless, far heavier cyberpower than she and her colleagues could muster was clearly in order.
In May 2017, Leder commissioned a noted firm, LMG Security, founded by celebrated internet sleuth Sherri Davidoff, to investigate Squarepoint’s actions.
The digital investigation confirmed the timing and breadth of the suspected misappropriation of data — definitively linking the accessed data to two Squarepoint URLs. LMG found that Squarepoint had made over 17,000 page visits to more than 16,000 unique URLs on Footnoted, a formidable cache for a small site.
In a report prepared for Leder and reviewed by this reporter, but whose findings have not been independently verified, LMG claimed that Squarepoint used an automated webcrawler to access the data — the precise four- or eight-second intervals between page requests being the tipoff. The cybersecurity firm, however, did not have access to the kind of material that would have allowed it to determine precisely what data the hedge fund had accessed. So Leder couldn’t know whether her search algorithms had been compromised.
Nevertheless, she and her lawyers persisted. With the LMG report in hand, Temple Bright on November 23 sent a seven-page broadside to Simmons & Simmons, together with a USB stick containing its supporting data. “Your client’s actions constitute primary copyright infringement of our client’s copyright works” under U.K. law, it read.
Simmons & Simmons changed emphasis, responding on December 7, “The acts of which you complain took place outside the U.K. by someone other than our client. In addition, all records of the data from your client’s website were stored in servers outside the U.K.” It repeated that Squarepoint reserved the right to take action against Leder if she discussed the matter publicly.
But bills running into the tens of thousands began to mount — and would run to $200,000 or more if Footnoted were to pursue Squarepoint, Leder was told. Seemingly, no lawyer would take her case pro bono. It wasn’t even clear where she could bring a case — or, given Squarepoint’s London headquarters, in what country.
Leder had been forced to take a hiatus to deal with her ailing mother — something she disclosed to subscribers. Her absence was generating client complaints. And the extent of the damage to her business was unclear. By April 2018, she had mostly shuttered her website, at least temporarily, citing the dispute with Squarepoint as a key contributing cause, but not the only one.
“It was not knowing,” she says. “If they could do it, who else could do it? What would stop the next quant from doing it?”
“And I just didn’t know how much extra it would cost to secure the site.”
In a series of interviews, both in person and via phone, Leder described the calamitous course of events. She also provided Institutional Investor access to email threads, LMG’s forensic report, and the legal correspondence with Squarepoint and its lawyers.
A U.S. spokesman for Squarepoint, Brian Maddox of FTI Consulting, issued the following statement on his client’s behalf: “We considered and investigated Footnoted/Ms. Leder’s allegations carefully and fully, when they were first raised. Footnoted’s claims have always been vague, unsubstantiated, and baseless.”
Maddox continued: “The last letter we received from Footnoted’s lawyers was in January 2018 and we have not heard from Footnoted, Ms. Leder, or their lawyers since then, so we understood that the matter had been resolved.”
One obvious question: Who cares about a hedge fund’s dubious, data-driven machinations? After all, Footnoted, with less than $1 million in annual sales, was in most respects just a bit player in the burgeoning, increasingly prosperous industry.
The answer: Leder’s predicament isn’t unique — and when revealed stands as a stark warning and cautionary tale in what could become a Wild West-style alternative-data free-for-all.
“This has been a pervasive issue for the longest time,” says John White, founder of Investment Data Licensing Advisors. “There are people who say, ‘Hey, I know how to get this data and I’m going to grab it.’”
Data vendors aren’t keen on discussing data misappropriation. One vendor, who asked not to be identified, recalls a recent instance when he enthusiastically offered up a trial subscription only to find the sales prospect accessing off-limits parts of his database.
“They were crawling all over it,” he gripes, adding that he ended the trial but opted not to pursue legal action. Who needs the negative publicity and expenses?
“Some engineers think, ‘It’s just data. We just want to work with it,’” says Elizabeth Pritchard, founder of White Rock Data Solutions, an advisory firm. “There is a mindset that when there are data scientists coming in from other industries, that the data is free.”
For quants, meanwhile, the challenge is to speedily hoover up new data sets and trade profitably before rivals do. Squarepoint is a choice example of such a data-hungry enterprise, describing itself in an SEC filing as leveraging “large historical data sets, either compiled in-house over time, or licensed or purchased from third parties.”
But beyond that, who exactly is Squarepoint?
More than many quants in this secretive business, Squarepoint operates under the radar.
The firm’s genesis lies in the quantitative trading acumen of quants like Gregoire Schneider and Olivier Durantel, chairman and a founding partner of Squarepoint, respectively.
Both attended École Polytechnique — the storied Paris university and incubator of fabled mathematicians — as did fellow Squarepoint majority owners Maxime Fortin and Antoine Fillet, according to their LinkedIn profiles. Founded in 1794, the school’s alumni also include former Credit Suisse CEO Tidjane Thiam, former French president Valery Giscard d’Estaing, and engineer Gustave Eiffel of Eiffel Tower fame. But for math geeks, its true rock stars were the polymath Henri Poincaré, the father of algebraic topology, and Benoit Mandelbrot, the developer of fractal geometry, which plays a key role in the quantitative canon.
Schneider and Durantel joined Lehman Brothers in 2000, according to media reports, building a trading desk known as nQuants. Eight years later, after Lehman’s collapse, the team had a new owner in Barclays, which bought up most of the bankrupt U.S. firm. As banks de-risked under increased regulatory pressure, Schneider and Durantel departed, leading a group of their colleagues in founding Squarepoint, which started trading in 2015.
(Elbaze, who is based in Switzerland, did not attend École Polytechnique, having picked up master’s degrees from École Nationale des Ponts et Chaussées and Université Pierre et Marie Curie, according to his LinkedIn profile. He responded to a LinkedIn email request for comment, saying he would like to try to help if possible. But after he was sent a list of detailed questions, he referred communications to Maddox and FTI Consulting. In September, Maddox wrote, “Client has declined to engage further.” Leder says that in late August, Elbaze accessed the Footnoted website.)
The fund has far outgrown boutique status. The filing of the New York affiliate, Squarepoint Ops, shows the firm has about 440 employees globally. Eighty percent of its investors are non-U.S., according to the document. (In December, the San Francisco Employees’ Retirement System approved a $200 million investment in Squarepoint funds, $25 million of which closed on May 1.)
Regulatory assets under management, which includes borrowed money, stood at $34.6 billion. Assuming the fund is leveraged at a ratio of 5 to 1, that would give it total assets of nearly $7 billion.
Today, the fund’s barebones website — which does not disclose the names of its principals — identifies only two strategies: long-short equities and low-latency, or high-frequency, trading. The SEC filing adds little. “The portfolio combines low-frequency trading with high-frequency trading,” it reads. “Many different data sets are used to design, improve, and run strategies that are often cross-validated using different but overlapping data sets. The majority of investments are made on a systematic basis.”
Squarepoint isn’t afraid to call out lawyers when it comes to what it claims as its own intellectual property. The fund in April 2018 brought arbitration proceedings against a former researcher, Vojislav Sesum, who it claimed had developed a trading strategy while at Squarepoint and subsequently offered it to Millennium Management, according to a filing of the case posted on Leagle.com.
The arbitrator initially directed Sesum to pay Squarepoint $188,137 in damages plus $919,053 in disgorged profits he received from Millennium. Sesum sought to have the award vacated, and though a court largely upheld the arbitration, the partially redacted final order, issued earlier this year, was sealed.
Sesum did not respond to a LinkedIn message seeking comment.
Brash and stocky at 5 feet 4 inches, Leder often dresses, well, like a hippy. She wears white shorts, lavender Birkenstocks, an orange blouse, and turquoise earrings. An Apple Watch completes her outfit — along with a blue cotton face mask, which she tosses on a table, settling into her chair at a proper social distance as her rescue dog, Max, lies down beside her.
Leder is a self-described wiseass who loves to strip away anodyne corporate facades and reveal the driving forces of avarice that underpin much of executive decision- making. That is what Footnoted strove to do every day — although she admits that she started the site to help boost sales of a 2003 book she wrote on financial fraud in the wake of accounting scandals at Enron Corp., WorldCom, and Tyco International.
An only child, Leder grew up in the middle-class neighborhood of Sheepshead Bay in Brooklyn. The journalism bug bit early: In high school, she worked at the school’s weekly as well as a local paper; at Brandeis University, she copy-edited the school’s newspaper. After graduating in 1988 with a B.A. in economics, the 22-year-old bagged a choice job as a business reporter at The Tampa Tribune. When an editor groused to Leder about how her ascent at the paper had been so much quicker than his own, she wisecracked, “Well, John, I’m smarter than you.” Leder was promptly reassigned to an obscure county bureau.
“I learned my editor was an idiot and his editor was an idiot and the editor-in-chief was an idiot,” Leder laughs. “I was the perfect employee.”
That kiss-off attitude got Leder pushed out or fired repeatedly over the next decade, despite commendable work on everything from the 1989 Eastern Air Lines strike to an investigative piece on an IBM toxic chemical leak. Her resume showed stints at Florida’s Bradenton Herald, Connecticut’s New Haven Register, and New York’s Poughkeepsie Journal.
Leder chucked corporate newspapers for freelancing in the late 1990s, writing for The New York Times, BusinessWeek, and Money. The economy was percolating, magazines paid well, and Leder was even making money in the stock market, particularly in a white-hot telecom company called Qwest Communications International.
The collapse of the dot-com bubble began in March 2000. By the time she sold Qwest in 2002, Leder calculated her holdings had fallen to less than $500 from around $4,000. She trawled through Qwest’s financial filings to divine what warnings she might have overlooked. “What did I miss?” she asked herself.
A lot, it turns out. The impressive earnings Qwest announced were an illusion. The 2000 pro forma net income of $1 billion the company had published in a January 2001 press release was revealed as an $81 million loss when Qwest reported GAAP results in its SEC 10-K filing less than two months later. Few noticed.
This was through-the-looking-glass accounting — but it was everywhere.
The results inspired a book from Leder: Financial Fine Print: Uncovering a Company’s True Value. It got endorsements, but to goose sales further, Leder launched a website on which she would blog about corporate accounting legerdemain and other matters.
Leder started slow. She didn’t charge for the first five years, only doing so in 2008 for an annual subscription of $1,000. She also hired her first full-time employee.
In 2010, Leder sold Footnoted to Morningstar, the Chicago financial publisher, for an undisclosed price, with the aim of expanding both the volume of the site’s stories and its subscriber base. Leder hired a dedicated sales rep and Footnoted grew to half a dozen staffers and a like number of freelancers. Bolstered, Leder was eventually able to raise the standard subscription fee to $10,000 annually.
“Michelle was really a player-coach,” says Theo Francis, who joined Footnoted as a reporter from Bloomberg News and is now a special writer at The Wall Street Journal. “Nobody was doing what Footnoted did. We were flagging the things that were important in filings; we were using search algorithms and typical journalist sleuthing.”
The service’s success in part derived from researchers’ ability to use search algorithms to quickly scan thousands of filings for specific language that flagged important corporate developments — words like “subpoena” and phrases like “government request for information” or “administrative inquiry.” In short, all the euphemisms a corporate lawyer could dream up to legally obscure the truth.
Leder became a master of all of them and entered them into Footnoted’s algorithms. Not knowing whether the algorithms were compromised was key to her decision to shut down, she says.
After selling her firm to Morningstar, Leder soon realized that clients were intent on tapping her personally for insight — flattering, but not good for a company with an expansion plan. “It was ‘The Michelle Leder Show,’” she says. “I wasn’t able to scale the business properly.” Leder bought Footnoted back from Morningstar in 2012 for under $500,000.
The Footnoted fare has always been varied, covering companies large and small, and digging into compensation, related-party transactions, and risk factors. “She does the digging up of the stuff you don’t want to see,” says Hendrik Leber, founder of Acatis, the German fund.
Executive compensation abuses and other perks were among the most common staples, as in 2004, when Gannett Co., the media giant at the time, revealed in its proxy that it would be paying its recently retired executive vice president for operations $600,000 annually — for life. That was in addition to picking up the ongoing tab on a company car, a club membership, and financial counseling.
Dubious accounting was always rife, as Leder’s Qwest exploration showed.
In a Walmart 2011 10-Q, Footnoted found that the retail giant had disclosed it was investigating whether it was complying with the U.S. Foreign Corrupt Practices Act. The reference was in the section entitled “Environmental and Other Matters.”
Four months later, The New York Times ran the first installment of a two-part, Pulitzer Prize-winning investigative series on bribery at the chain’s Walmart de Mexico affiliate. The Bentonville, Arkansas, company last year agreed to pay more than $282 million to settle parallel charges brought by the SEC and the Justice Department under the FCPA.
There was also rampant self-dealing. Healthcare technology company Cerner Corp. in 2013 disclosed in its 10-K that it was investing $88 million to acquire office space for employees at a local sports complex whose developer and Kansas City soccer team tenant were both controlled by Cerner’s chairman and vice chairman.
“They had wonderful numbers; it looked good,” recalls Acatis’s Leber. “Michelle alerted us to the intercompany dealings, company expenses, related-party transactions.”
Acatis decided not to invest in Cerner. “You never win when you deal with people without integrity,” Leber says.
There were ambitious projects too, trend stories that generated buzz — and made money for subscribers. In early 2011, Footnoted published a list of the top ten takeover candidates based on language in the companies’ various financial filings. Three of them — Smurfit-Stone Container Corp., Pride International, and Lawson Software — announced sales within three months of the report’s publication. Two others — Copano Energy and Leap Wireless International — eventually followed suit.
“That’s a pretty good batting average, if I say so myself,” says Leder.
Fridays were the busiest days, when legions of companies would file 8-Ks containing their most embarrassing disclosures in the hopes that investors would overlook them as they prepared for the weekend. On occasion, the SEC’s Edgar website would break down under the volume of the Friday filings. Leder quarterbacked coverage of the weekly surge with “Friday Data Dumps” that detailed the disclosures by the dozen.
Leder had a motto, a version of which she would repeat endlessly to her staff: “There are no accidents in SEC filings. Everything is there for a reason, even when we’re not exactly sure what the reason is at the time.”
There are obvious steps Footnoted could have taken. LMG’s digital forensic report for Footnoted gave some examples: In addition to updating some unpatched and out-of-date software, it recommended moving the website to a dedicated server from its shared WordPress cloud platform. At the very least, that would have allowed Leder to occasionally probe her site for security gaps.
LMG also suggested that Footnoted set up a log system that would have allowed it to differentiate between the pages that have merely been visited and those whose content had been accessed. “There was nothing that would tell us exactly what files were taken,” says LMG’s Davidoff. “It’s really important to have granular information.”
There should also have been automatic blocks on users whose activities breached specific thresholds. “There is often free software for this,” Davidoff says. “There are inexpensive ways of setting up alerts.”
A dose of paranoia isn’t necessarily bad. Investment Data Licensing Advisors’ White suggests that prospective subscribers be offered just a small subset of the data. “You don’t give them the family jewels,” he says. “As [they] are asking for more information, the level of security should increase.”
That means the specificity of terms in trial contracts should rise as well as the level of detail as to who will have access. “You need to show me where the data is going and who is going to see it,” White says.
Leder did not offer a trial contract, which she believes could have provided her an airtight legal case against Squarepoint. She is adamant, however, that Elbaze’s actions clearly violated the website’s terms of use, which read in part: “You may not copy, reproduce, republish, upload, post, transmit, or distribute any component of Footnoted.”
“I would still love to resurrect” the website, says Leder, who has been freelancing for The New York Times’s DealBook site.
She wonders whether she made the right decision in shuttering Footnoted. “It was a really difficult time for me emotionally,” she says. “I had been really burned by this. I felt violated.”
And, naturally, Leder wonders whether she should have sold her firm when hedge funds expressed some interest in buying it all those years ago.
After a quick lunch of lobster rolls, fries, and Diet Coke, Leder dons her blue cotton face mask, musters her dog Max, and heads out to her gray Honda CR-V under the clear blue skies of Downeast Maine.
Her immediate plan in July: a cross-country drive back to California. Perhaps more freelance work. Maybe even eventually relaunching Footnoted.
But two months later, in Covid-plagued Los Angeles, as summer drew to a close under smoky skies and pounding, 112-degree heat, Leder was bedeviled by one further development.
On September 9, Simmons & Simmons, Squarepoint’s U.K. law firm, emailed Leder, citing her conversations with Institutional Investor. “You have deliberately mischaracterized the events and the email exchanges between the parties,” it read. “You have falsely alleged that our client acted illegally, breached the terms of the trial, and accessed vast amounts of material on the Footed [sic] site without permission.”
The email went on to say Leder had published “defamatory statements” that could hurt Squarepoint’s reputation. It listed four demands: that Leder “withdraw” statements made to Institutional Investor, agree not to encourage third parties to publish them, provide Simmons & Simmons with details of her conversations with Institutional Investor, and divulge to the law firm the identities of any other parties she had talked to about the matter, including contact details.
Simmons & Simmons gave her until noon London time on September 11 to respond. “In the event that you are not willing to comply with the above requests, our client reserves the right to take all steps necessary to protect its reputation,” it stated.
At press time, Leder still hadn’t responded to the letter.
The author of this article, who worked at The New York Times between 1998 and 2000, edited articles written by Leder when she was a freelance financial reporter and then again when he was an editor at a Time Inc. personal finance magazine in 2001 and 2002.
