Levine on Wall Street: Internet IPOs Past and Future
Twitter IPO speculation: fees
Twitter is in the process of adding banks and negotiating fees for its upcoming initial public offering. Twitter seems to want to avoid the fate of Facebook, which overpriced its IPO and ticked off the market, as well as of LinkedIn, which underpriced its IPO and left money on the table. Facebook paid banks a notably skinny 1.1 percent fee for executing its IPO (which, to be fair, was significantly larger than Twitter's will likely be), and split that fee among three bookrunners and 30 other underwriters. I'd sort of expect Goldman Sachs, picked to lead this deal, to tell Twitter that in IPOs, as in much else, you get what you pay for, and that paying 1 percent for a botched IPO costs you far more than paying 3 percent for a successful one.
Twitter IPO speculation: price
I'm late on this, but Aswath Damodaran had a spectacularly good breakdown this week of how Twitter, its bankers, and its buyers will likely think about pricing Twitter's IPO. He gets a range of $10 to $15 billion, and he does a pretty detailed and compelling analysis without any financial performance data except a public revenue number. The point here, really, is not to help you figure out how much you should pay for a share of Twitter; it's to explain the difference between valuing a business and pricing an IPO. If you can do the latter with just one performance number, it may not have much to do with the former.
Also in hyped internet IPOs: the eToys settlement
Goldman Sachs knows the difference between value and price, though the case of eToys may not help you sort out which is which. EToys, a Pets.com-for-toys, IPOed in 1999 and was bankrupt within two years. Goldman led the deal and seems to have underpriced it -- relative to demand -- in order to please its investor clients. There are some bad facts, eToys's estate sued, and yesterday the bankruptcy court approved a $7.5 million settlement. "Both sides sought to keep the terms undisclosed, claiming they contained 'confidential and sensitive' material," but the judge said no dice; you can see how Goldman wouldn't want sordid details of its past IPO underwriting mistakes to come out while it prepares for the Twitter IPO. Because the eToys IPO managed to be both wildly underpriced (it priced at $20 per share and traded as high as $78 the first day) and wildly overpriced (it got to zero pretty quick).
Things that JPMorgan can and can't admit to
The SEC's newfound desire to make firms admit wrongdoing in settlements is largely symbolic on the SEC's side, but has real-world implications for the settling firms. Certain admissions of wrongdoing make it much easier for plaintiffs' lawyers to win lawsuits and shuffle vast quantities of money from one set of innocent public shareholders to another set of innocent public shareholders, taking out enormous chunks of that money for themselves. Everyone -- the settling firms and the SEC alike -- realizes that that's sort of a bad outcome. Anyway here is a good piece from Peter Henning explaining why JPMorgan admitted to the things it did admit to in its SEC settlement (basically, they're unlikely to create private liability), and why it will fight harder against CFTC market manipulation charges (basically, those charges could open JPMorgan up to lawsuits from other participants in the credit derivative index markets that it maybe manipulated).
Humans replaced by computers say "I told you so"
This is a story about a former stock exchange specialist, rendered obsolete by computerized trading, "allow[ing] himself a bitter laugh" at the recent high-profile glitches in those computerized markets. A plausible model here is that computerized markets are cheaper, faster, more accurate, and more efficient than humans with handwritten trade tickets 99.9 percent of the time. But the computers, not being programmed to gloat, don't. The 0.1 percent of the time that the computers mess up, the humans fire their gloating circuits.
To contact the author on this story:
Matthew S Levine at email@example.com