Last week I posted the first of a three-part series on my new article with Mike Stegemoller on SPACs. In this post I'll discuss our second finding, which relates to underwriter discounts, or the gross spread. In a normal IPO, the nominal way an underwriting bank makes money is by buying stock from the company at a discount, and then turning around and selling it to the public at full price. The difference, or "spread" between the purchase price and the sale price, is the investment bank's official compensation for bringing the fledgling firm to market. Some scholars would point to underpricing--and the banks' corresponding ability to curry favor with client--as the true compensation for underwriting, but we covered underpricing last week.
What do underwriters do to earn their discount? Well, they expend a lot of effort trying to value the firm correctly (I'm looking at you, Morgan Stanley), because in a firm-commitment offering they take on the "risk" that they won't be able to unload the shares they buy from the company and offer to the public. "Risk" is in quotation marks because the underwriters build a "book" of pre-sale offers of interest, and generally won't agree to take the firm public without being sure all of its shares will sell. The underwriter does face a real risk of liability for false statements in the offering documents, however, and the spread may also compensate the bankers for this risk.
A curious fact about traditional U.S. IPOs is that the gross spread is sticky--underwriters almost always buy company shares at a 7% discount, as documented in this fabulously named Chen Ritter paper, shouting out to Sherlock Holmes, no less. As Chen & Ritter, describe, the clustering of spreads around 7% is 1) a relatively recent phenomenon, 2) specific to the U.S., and 3) roughly twice as high as in other countries.1
Suspicious. After all, shouldn't some firms be riskier to underwrite than others, and thus command a higher spread? And others by the same token be surer bets, with a correspondingly smaller spread? Chen and Ritter conclude that most spreads for firms over $30 million are above competitive levels and discuss explanations ranging from implicit or explicit collusion to reducing underpricing.
SPACs, essentially piles of cash, should be easier to value than the typical company. The information asymmetries that characterize the average company just aren't there. Moreover, bookbuilding should be an easier process because there isn't much of a "story" needed to persuade potential customers to buy. By the same token, SPACs should be less risky to underwrite, since the trust fund creates a floor price below which the stock should not drop. As a shell company, disclosures are boilerplate and close to risk-free for the underwriter. So we hypothesize that the spread should be lower than 7%.
Yet the mean and median underwriting discount for our sample is 7%. We argue this offers further evidence that spreads are above competitive levels. Moreover, as the SPAC form evolved, investment banks began to accept a portion of the spread as deferred compensation. Meaning that the nominal discount remained 7%, but some of that cash was tied up in escrow, and only released to the bank if and when a subsequent acquisition occurred. The fact that the SPAC underwriting discount initially clustered around 7%, but all of the banks quickly proved willing to sacrifice nearly half of their compensation, without demanding any increase in overall spread as compensation for the delay in payment, is further evidence that a 7% spread is not the product of a competitive market. In other words, if competition is fierce enough, underwriters can apparently operate at spread levels much lower than 7%.
That sticky 7% is looking curiouser and curiouser...
1 The Chen Ritter study is from 2000, but Abrahamson, Jenkinson, and Jones (2011) confirm their findings.
TrackBack URL for this entry:
Links to weblogs that reference SPACs: The Finance Article (Part 2, The Underwriter Discount):