I’ve spent much of the last two days dissecting the second amendment to the Groupon S-1. I was hoping to post my complete thoughts tonight, but I’ve seen so much shoddy reporting on this S-1 that I want to make sure that I’ve polished the most important pieces. What I’ve read in the S-1 makes me so angry that I’ve expanded the scope of my analysis to essentially re-calculate all of the key numbers from Groupon’s highly misleading document.

Although Groupon has largely eliminated the controversial adjusted CSOI metric — for which it was rightfully beaten up by just about everybody — there is still a lot of stink on this S-1. Groupon and its underwriters have done a masterful job of torturing the English language to make things seem very different from what they are. At every turn, Groupon tries to get investors to focus on misleading and inflated metrics while ignoring the ones that are economically important.

There are the major issues, such as gift card liability, gross vs. net accounting and inflated customer numbers that I talked about before this amendment was released. (See my blog post about Groupon’s accounting practices and my conversation with Emily Chang on Bloomberg TV.) Based on my revised estimates, Groupon has between $500 million and $750 million in liabilities it is not showing on its books.

But there are many others.

In the revised S-1, there are some new metrics. Although they should elucidate, they again only serve to mislead. One new metric, the average Groupons purchased per consumer, would lead the average investor to believe the business is going well. It has been reported in the press that the average consumer purchased four Groupons. While that is technically true, it is highly misleading.

The number that Groupon provides is the arithmetic mean; the most relevant number is the median number of Groupons purchased. That number is 1 — more than half of Groupon customers have only purchased 1 Groupon.

Here is a quick explanation of mean vs. median. Assume that there are 100 people in a room. Each is worth $100,000. The mean and median in this case is $100,000. Now, Bill Gates walks in. Add in his $56 billion net worth and the new mean is $555 million, but the median is still $100,000. (See Khan Academy’s video on averages for a more in-depth explanation.) Groupon is similarly distorting its financial reporting. I’ll dive more into this later.

If Groupon truly believes that the arithmetic mean is the appropriate measure, its management is grossly incompetent. If they are putting out a number to deliberately mislead, then the SEC should freeze this offering.

Groupon’s median sales to Groupon list subscribers is… Zero. This company literally can’t get 80% of its mailing list to buy money at half off.

Groupon also shifted its financial reporting to emphasize year-over-year comparisons vs. quarter-over-quarter comparisons. Given that the company is still at an early stage, the only numbers that matter are the quarter-over-quarter comparisons. Groupon blended its first and second quarter results in the front section of the document. This has the effect of hiding the substantial declines in its business in the second quarter.

Likewise, Groupon is hiding the numbers for its poor performance of its Asian operations in its more lucrative European operations. The characteristics of the markets are so different that they should be reported separately.

Even the basic terminology Groupon uses has the effect of misleading potential investors. The term “gross profit” has no resemblance whatsoever to a profit. It is what should be considered “revenue”. What Groupon calls “gross margin” is really its weighted average (mean) revenue share percentage. (This is an important metric and I expect it to collapse within the next year.)

It isn’t unusual for IPOs to have shareholder lawsuits. Groupon co-founder Eric Lefkosky has had more than his share with previous companies, according to Fortune. Investors and employees have been left holding the bag while Lefkosky has profited immensely. Lefkosky and his family have already pocketed $382 million from Groupon, according to Fortune. The depth of deception in this S-1, if gone unchanged, will give plaintiff’s lawyers a lot of ammunition. (As will the many news stories generated every time Groupon fudges metrics.)

I know I’ve disappointed many interviewers when I’ve refused to describe Groupon as a Ponzi scheme or a pyramid scheme. I researched those terms today. Those terms are not technically accurate. But we may soon start using the phrase Groupon scheme.

For some of the best analysis on Groupon’s second amendment to the S-1, see:

• Hagey on Groupon’s Amended S-1 Filing, Accounting
• Groupon In Major Trouble as Q2 Results Show Plunging Revenue Growth
• New Filing Reveals Groupon’s Oldest Markets Got Even Worse