So here’s a business strategy you could follow if you wanted to make a lot of money in the mutual fund business: Get a computer, load it with the past 10 years of stock market data, and then ask the computer to come up with a strategy that only would have bought the stocks that went up. (It could be that companies that start with the letter “T” are good buys on Tuesdays. It doesn’t really matter as long as it works. And no one will ask.) Tell the world you know how to beat the market.

You don’t. You don’t know what’s going to happen over the next 10 years, or even next year. You may not even know if you can buy and sell stocks based on your strategy. You haven’t done it. All you have is a computer that is very good at picking stocks that have already gone up. But you don’t have to emphasize that.

That’s what F-Squared, a hot fund asset management company, did, very, very successfully.

I am exaggerating and simplifying here, but, basically, back in 2008, F-Squared licensed an investment strategy from a company called Newfound Research, which said its computers knew when to buy and sell different sectors of the S&P 500. F-Squared said that, based on the technology it licensed, its investment strategy returned 200% from mid-2001 to September 2008 (a period that ended right before the messiness of the financial crisis), or roughly double what the market did in the same time. It then pitched a bunch of mutual funds touting the past outsize returns of this strategy.

It was a winning marketing pitch. Money poured in. At the end of June 2014, F-Squared managed $27.7 billion. By contrast, the firm was managing just $100 million in late 2008.

It’s also a tad deceitful. So, perhaps the news this week is not that surprising: the Securities and Exchange Commission has told F-Squared that it is considering suing the company for civil fraud.

But here is the surprising thing: the SEC is not suing the company over any of the things I described above. You can do all that. You just have to attach a little asterisk and say that your past performance may have been helped by the fact that when you put together the strategy for your portfolio, you already knew what stocks had gone up in the past.

F-Squared is in trouble for not including that asterisk. It said that the company that it contracted with, Newfound Research, had actually invested money based on its computer models from 2001 to 2008. It hadn’t. The returns were only based on what Newfound thought would have happened if it had invested money during that period, which it didn’t.

But that wasn’t it. When F-Squared’s mutual funds launched, the company didn’t even end up using the strategy that had produced the fake returns that it had bragged about in its marketing materials. It did something slightly different that sounded the same.

You can probably guess what happened next. The mutual fund purportedly based on F-Squared’s strategy, the Virtus Premium AlphaSector fund, didn’t end up beating the market. The fund has turned out to be a bit of a laggard compared to the overall market. Over the past three years, the fund has gained about 13% a year, or about three percentage points less than the S&P 500. That’s not horrible, but it is far less than the double returns of the S&P 500 that F-Squared marketed the fund on.

F-Squared officials declined to comment for this story. Last October, F-Squared sent a letter to clients saying they should ignore the numbers it bragged about when it first marketed the fund. F-Squared CEO Howard Present has told clients that he and his company had been misled by Newfound. He also said a computer programming error explains why the fund’s actual strategy differed from the strategy that the fake portfolio would have engaged in; that is, if it had been real in the first place.

To be sure, this isn’t Bernie Madoff. No one has lost money with F-Squared. And I guess that’s true, as long as you don’t consider the money investors would have made if they had just put their money in a low-cost index fund.

But the bigger issue is that more and more Wall Street firms are launching mutual funds and ETFs based on similar computer models. This approach allows them to publish theoretical track records that seem to be real, and impressive. And as these funds grow more complicated, the marketing is allowed to move farther from reality.

Take the Doubleline Shiller Enhance CAPE fund, which is marketed as a fund that will produce returns slightly better than the S&P 500 by picking the most undervalued sectors of the stock market. The metric for what’s cheap was devised by Noble Prize-winning economist Robert Shiller.

Where does the fund actually invest all of its money? In bonds. And it uses those bonds to buy swaps that will mimic the returns of the stock market. But what will happen if the stock market rises but bonds go down? The fund won’t do very well. The fund is invested in bonds.

Perhaps the most amazing thing about the F-Squared story is that the revelations about its past marketing abuses have not seemed to turn away investors. Since October, when the problems surfaced, F-Squared has managed to attract an additional $6.6 billion, brought in from brokers at the banks that sell their funds. That might say something about Wall Street’s marketing machine, the blindness of investors, or the combination of both factors.

F-Squared’s marketing materials play up the fact that its strategy will protect investors from losing money if the market falls. Of course, most of F-Squared’s mutual funds launched in 2009 or later. Since then, the market has mostly gone up. So, how does it know that its funds will not lose money when the market drops? It doesn’t. But neither that fact nor the SEC can stop F-Squared from saying so.

Editor’s note: A previous version of this story incorrectly stated that F-Squared had $22.7 billion under management. In fact, the firm had $27.7 billion under management as of late June 2014. This story initially stated that this amount represented “about double” the assets the firm had been managing in late 2008. In fact, the firm had less than $100 million in assets under management as of 2008. A previous version of this story incorrectly stated that F-Squared had purchased Newfound Research in 2008. In fact, F-Squared began to license an investment strategy from Newfound Research at that time.