In his recent Slate column, Daniel Gross provides a few other less obvious explanations. He cites some findings from a study conducted by Oxera Consulting, and commissioned by the London Stock Exchange and the City of London:
* China (and other countries with rapidly expanding economies) are eager to support local capital markets.
* U.S. investment banking firms charge fees that are significantly higher than those charged by firms in other countries. As an example, the average fee charged for U.S. IPOs was about twice that charged for firms going public on London exchanges.'
* U.S. IPOs have greater first-day underpricing than those in other countries. While big first-day return is good for purchasers of IPO stocks, it amounts to money left on the table for the issuing company.
* While not specifically measured, the perceptiuon is that non-IB fees (for lawyers, accountants, printers, etc...) are much higher for US IPOs.
* A decreased value placed on the prestige of an American exchange listing.
I'm not sure I buy these arguments. In the grand scheme of things, I doubt a couple of percentage points of IB fees would be the reason firms choose to list offshore rather than in the U.S. My own reading of the academic literature on IPOs (and granted, it's not my primary area of expertise) is that firms going public focus far less on fee-related issues than on other matters (like the immense risks and payoffs associated with cashing out).
I think it's far more likely that firms see the increased costs associated with SOX as just the tip of the iceberg. While I don't have hard data to back up my arguments, my sense is that the present value of many, many years of SOX-related costs would swamp the few million in additional costs associated with a U.S. IPO.
Update: Equity Private has a quick comparison of SOX costs vs. additional IB fee over at Going Private.