It was the right pitch at the right time. In the fall of 2008, just as the stock market was tanking in the financial crisis, a tiny Boston-based firm called F-Squared Investments began marketing a new investment strategy that, it said, had trampled the returns of the S&P 500—beating the market by an average of three percentage points annually for years while taking on less risk. Avoiding losses, in fact, was a guiding principle of the computer model behind the strategy’s success. “Although the performance numbers seem too good to be true,” a press release quoted a former Merrill Lynch strategist as saying, “they are based on a live track record and can’t be ignored.” The validity of F-Squared’s approach was seemingly confirmed when the firm later reported that its main portfolio had fallen just 1.9% in 2008, vs. a drop of 37% for the S&P 500. Money soon began flooding in.

Today F-Squared’s investment products are sold by financial advisers around the country and used to ­manage the accounts of tens of thousands of individual investors. The company’s fee-generating assets have leaped to more than $28 billion. The F-Squared model—in which a fund trades in and out of exchange-traded funds to ride what’s hot and avoid underperforming market sectors—has been hailed by some on Wall Street as a new paradigm.

But now it appears that it was too good to be true after all.

The Securities and Exchange Commission has been investigating F-Squared for over a year. The SEC looks close to reaching a settlement with the company, and it could come as early as this week, according to sources close to the negotiations. (Update: On Monday, F-Squared settled with the SEC paying a $35 million fine and admitting wrongdoing.) Privately the SEC has told F-Squared that regulators believe the firm and its founder, Howard Present, committed “reckless violations” of securities law, according to a person close to the investigation. In November, Present abruptly stepped down as CEO of F-Squared and left the company, though he remains a minority owner.

The consensus seems to be now that much of F-Squared’s amazing early track record was faked, either a product of a hypothetical backtest or just an outright fraud. A year ago F-Squared removed more than half of its performance history—a period covering April 2001 to October 2008—from both its website and its marketing materials. In May, according to someone close to the investigation, the SEC told F-Squared that its advertised returns for its main strategy prior to October 2008 were “systematically incorrect” and “clearly overstated.”

Source: Morningstar.


What’s less clear is who’s to blame. According to a person familiar with the SEC investigation, F-Squared and its founder, Present, have claimed that they were duped by the firm that developed and licensed Present the portfolio model. (Present did not respond to repeated attempts by Fortune seeking comment for this article.)

One thing that’s certain is that the quick rise of F-Squared proves once again that when Wall Street comes calling with its latest “can’t lose” idea, investors should be wary.

When Present launched F-Squared in 2006, he was looking for a way to profit from the boom in ETFs. Indeed, the amount of assets invested in the hybrid funds—which can be traded like stocks—has exploded from $300 billion in 2005 to nearly $2 trillion today. Present was a petroleum engineer by training who had gotten into the mutual fund business in the mid-1990s. In 2002 he landed a top job at Evergreen Investments (now part of Wells Fargo) as the head of the product division. Former co-workers describe Present as a natural salesman, but he had setbacks. At Evergreen he launched five closed-end funds, a precursor to ETFs, that had mixed results. In 2004, Present left the fund giant, and two years later he founded F-Squared. His first product launch, a line of ETFs that mimicked popular mutual funds, fizzled. He began casting around for something new to sell.

Present’s big break came when he discovered a startup called Newfound Research in 2008. Newfound, like F-Squared, is based in Boston. It was formed by investment adviser David “Jay” Morton, a technology lawyer named Tom Rosedale, and a wunderkind computer science nerd named Corey Hoffstein. Morton’s career stretches back to 1992, and he has run into trouble in the past. In 2003 a client sued Morton for $750,070, accusing the broker of unauthorized and excessive trading in his account. The client was paid $49,000 in a settlement.

In the summer of 2007, after his sophomore year at college, Hoffstein had interned for Morton, his father’s broker. Hoffstein was struck by the fact that fund managers were often locked into their strategies no matter what was happening around them. Back at college, Hoffstein, who had been programming since he was 14, decided that he would design his own trading algorithm. Hoffstein’s model for timing the market, the one that later became the basis for F-Squared’s investment product, is something he calls momentum plus volatility, or “dynamic momentum.” Generally the idea is to buy stocks—or, through ETFs, market sectors—that have been going up and doing so steadily, and stay away from those that are more volatile. Based on historical data, it appeared to work quite well. When Hoffstein returned as an intern for Morton in 2008, he showed Morton the program, and the financial adviser was blown away. Morton introduced Hoffstein to Rosedale, a lawyer he knew, and eventually to Present. By the end of the summer Hoffstein, Morton, and Rosedale had formed a new company and made a deal with Present. Hoffstein headed back to college at Cornell and eventually to grad school at Carnegie Mellon, where he earned a master’s degree in computational finance.

According to a source with knowledge of the investigation, F-Squared claims that Present initially approached Morton about licensing mutual fund indexes that Present had created. But Morton turned around and began selling Present on a “sector rotation model” that Morton said he’d used for years. Morton and Present traded emails about how long money had been managed against the portfolio and how the methodology had evolved over time, including changes to reflect Hoffstein’s approach. Present asked Morton to send him a file documenting the returns.

On Sept. 3, 2008, Present emailed Hoffstein directly asking for details about how and when the model had changed, according to two sources who have knowledge of the correspondence. He also asked again for the exact month in which client assets had first been managed against the strategy. Instead of replying to Present, Hoffstein forwarded the email to Morton with a message: “Since I have never ‘managed money’ using the technique—just researched its viability, I think you should handle it.”

A couple of weeks later, Present was ready to go to market. On Sept. 18, 2008, Present issued a press release announcing the launch of F-Squared’s AlphaSector Rotation Strategy, which could be used by money managers and mutual funds to run portfolios. It said AlphaSector had been used to manage client accounts since 2001 and had beaten the S&P 500 by three percentage points a year. The model’s supposed ability to time major market moves was even more impressive. In July 2007, according to F-Squared’s press release, AlphaSector had completely sold off its financial stocks. F-Squared also said its AlphaSector had moved 50% of its portfolio to cash shortly before Leh­man Brothers ­collapsed. F-Squared said that $100 million in client assets were invested in the strategy. But according to the disclosures at the bottom of the press release, AlphaSector’s returns were “hypothetical” and did not “represent returns that any investor actually attained.” Nine months later there was just $5 million tracking the strategy, according to Morningstar.

F-Squared’s disclosures changed over time. In January 2009 it issued a press release that said the AlphaSector Rotation had outperformed 99% of all large-cap mutual funds. At the bottom of that release, F-Squared said that some of its return data were based on “hypothetical performance,” and that “hypo­thetical backtested performance has many inherent limitations.” Later F-Squared called some of its performance “theoretical” and “hypothetical.” Elsewhere F-Squared said the indexes that tracked the firm’s performance were based on an “active strategy” since April 1, 2001, and were “not backtested.”

Present’s boasts about the firm’s investing prowess also grew as time went on. In early 2010, F-Squared said its Alpha­Sector Premium Index had returned 198% since 2001, vs. 13.5% for the S&P 500. In interviews, the line between F-Squared and the strategies it licensed grew blurry. In October 2010, Present told trade publication Bank Investment Consultant that F-Squared had “dropped the technology sector in 2001,” thereby avoiding losses. F-Squared now says it didn’t start working with client assets until February 2009.

At Virtus Investment Partners, a $60 billion mutual fund firm that hired F-Squared to sub-advise some of its funds and separately managed accounts, Present got the nickname “Howie Backtest,” say two former employees. “He would come with us on sales calls and tell people to go to a website that showed he had the best-performing fund in the U.S.,” says one of the former Virtus salesmen. By late 2012 the Virtus Premium AlphaSector Fund, which was based on F-Squared’s index and for a long time listed Present as its co-manager, was one of the company’s largest and fastest-growing funds. In early 2013 a prospectus for the Virtus Premium AlphaSector Fund said its goal was to outperform the S&P 500 while also seeking to manage “downside risk.” According to the prospectus, the AlphaSector index had risen 12.7% annually since April 2001. A little over a year ago Virtus removed those numbers from its marketing materials. Virtus declined to comment for this story.

Inside the industry there were growing questions about the veracity of F-Squared’s returns. Eventually the SEC grew suspicious. In October 2013, Present sent a letter to clients saying that the firm was under investigation. According to the letter, the SEC had issues with some of the claims the firm has made in its marketing materials, particularly its “use of historic data in advertising materials.” That month F-Squared’s pre-October 2008 results disappeared from its website and marketing materials.

Over the past year, despite the SEC investigation, F-Squared’s funds have attracted an additional $7 billion in assets, though that flow has slowed lately. Wells Fargo has recently put F-Squared on a watch list and Raymond James has told its advisers to stop selling F-Squared’s products.

There are also signs that individual investors and brokers are growing more cautious in general about funds of ETFs—known in the industry as “tactical ETF funds”—like those offered by F-Squared. In the third quarter of 2014, assets dipped in those funds for the first time since Morningstar began tracking the category. Part of the problem may be the investigation into F-Squared, the largest manager in the group. But after a long bull market, selling funds that promise protection against market dips—and charge higher fees to do so—is more challenging. Especially when the funds haven’t done particularly well.

Over the past three years F-Squared’s AlphaSector Index has produced an annualized return of just over 20%, according to Morningstar. That trails the S&P 500’s 23% average return over the same period. F-Squared’s supporters say that’s not a fair comparison—that the company’s portfolios are safer than the S&P 500 and in a market downturn like 2008 will lose less. They say a few percentage points a year is not a lot to give up for more downside protection. Since we haven’t had a big market downturn since 2008, that’s hard to prove. But F-Squared hasn’t done well navigating the market’s recent volatility. The Virtus Premium AlphaSector fund had dropped nearly 4% since July, vs. a 4% gain for the S&P 500 through mid-December.

Source: Bloomberg.


An F-Squared spokesman says that the firm is committed to its objectives of delivering strong risk-adjusted returns: “Since F-Squared first had assets invested pursuant to the AlphaSector strategy in October 2008, the performance has been consistent with these expectations, and our clients have seen the results in their portfolios.”

F-Squared stopped using Newfound’s models in June of 2013—­cutting off the licensing fees it paid to Newfound. CEO Rosedale and Hoffstein, who returned after grad school as chief investment officer, have learned a lot about the mutual fund business since 2008. They’re no longer licensing Hoffstein’s portfolio model to others. Instead, they have decided to grow Newfound as an independent fund manager. And they’ve had some early success. Newfound now manages about $160 million, up from close to zero a year ago. Its largest fund, the Newfound Risk Managed Global Sectors, has attracted $63 million. But its performance has been so-so. Over the past six months the fund is down 2.4%, vs. a 2.3% drop in MSCI’s world index, the firm’s preferred benchmark.

Neither Hoffstein, now 27, nor Rosedale, 42, will comment on F-Squared or the SEC investigation. (According to its latest filing, Newfound is not the subject of an SEC investigation.) Morton is no longer an owner of ­Newfound, though Rosedale says the split was amicable. Morton did not respond to numerous requests for comment by Fortune.

Newfound’s small downtown office near Boston Commons has the feel of a startup. It was remodeled in the fall to accommodate nine employees, up from just five a year ago. Recently a visitor complimented Rosedale on adding a new plant to the premises, only to realize, upon closer inspection, that it wasn’t real. “Oh, yeah, that’s plastic,” said Rosedale. “Unlike other people, we’ll tell you when our stuff is fake.”

This story is from the January 2015 issue of Fortune.

Editor’s note: An earlier version of this story gave Hoffstein’s title as co-chief investment officer. He is the CIO of Newfound.