SAN DIEGO ( TheStreet) -- Arena Pharmaceuticals' (ARNA - Get Report) most ardent shareholders -- a group of retail investors known affectionately known as the Areniacs -- are unhappy with the early prescription numbers for the company's weight-loss pill Belviq, so they're making up their own numbers.
Doctors wrote 1,829 prescriptions for Belviq for the week ended June 21, a 68 percent increase over the 1,087 scripts written the week before, according to IMS Health.
Arena and its marketing partner Eisai began selling Belviq on June 11. It's hard to say much about two weeks of prescription data but that matters little to Arena's retail investors because they've brainwashed themselves into believing the marginally effective weight loss pill would be an instant blockbuster.When Belviq weekly scripts didn't double, triple or even quadruple this week, as they expected (unrealistically), the Areniacs blamed IMS, short sellers and anyone or anything other than the drug or themselves. "Because IMS data is only 73% accurate, I adjusted it to estimate actual number of scripts and based my analysis on the adjusted number," writes "Reasonable Risk," in a Seeking Alpha column Monday titled Improper Analysis of Belviq's IMS Prescription Data Causes A 9% Sell-Off Of Arena On Good News. "Reasonable Risk" doesn't explain how he reached the conclusion that IMS reporting of weekly Belviq prescription is only 73 percent accurate. It's true that IMS weekly prescription data are rarely 100 percent accurate so require adjustments, but two weeks of data is not enough to calculate an error rate. The anonymous author of the Seeking Alpha article starts with the conclusion that Belviq's early launch must be going great, so if actually reported IMS data aren't supportive, "adjust" the data to make them look better. Unfortunately, even his adjusted analysis of Belviq scripts isn't looking too hot: "A linear growth rate based on the second week's 501 scripts per workday suggests pharmacies will fill about 9,000 scripts the first month. Although that is about 30% below my predicted initial estimate, two data points are inadequate for making predictions," he writes.