Initial public offerings (IPOs) earn fortunes for investment banks and their investing clients. At the heart of the IPO business is ‘bookbuilding’, the name for the way most IPOs are run.
Bookbuilding allows banks to serve clients, but also to exploit them. A recent survey of investors in IPOs suggests that even after the IPO-related scandals of the bubble period, IPO practice is not always in the interests of those who are paying the fees—the owners selling old shares and the companies issuing new ones.
MALAY KARMAKAR / MINT
In the first half of 2007, according to Thomson Financial, worldwide IPO volumes were $135 billion (about Rs5.54 trillion), up about 30% from the same period of 2006, with Asia (excluding Japan) the second largest contributor after Europe, West Asia and Africa. IPOs are the most lucrative part of the ‘primary equity’ business, which earned banks fees estimated at $10.4 billion in that six-month period.
Bookbuilding is like a Dutch auction in which the price of the shares is lowered until there is enough demand, and everyone bidding at, or above, that level gets shares at the highest clearing price.
The sellers in the auction are the existing shareholders or the company itself; the bidders are investors, mainly institutions; and the auctioneer is the investment bank. But bookbuilding differs from an auction in two key ways.
First, the price is almost always set below the highest clearing price.
Second, the bidders disclose their identity as well as their bid price, so the bank can favour some bidders and blackball others when allotting the shares. Once issued, IPOs tend to jump around 10-15%, reflecting the fact that the shares are priced below investors’ bids.
Most agree that this IPO discount is not a free gift and must be “buying” something—but what?
The “academic” view is that the discount is buying from institutions their views on the value of the shares, which the bank and issuer need to set the right price.
The “investment banking” view is that underpricing is buying for the issuer a choice of investors: If the price were set at the highest clearing price, the shares would be allotted to everyone who bid at least that level, but a lower price allows the issuer, through its bank, to pick and choose.
Finally, the “quid pro quo” view is that the bank is buying for itself the secondary trading business of the investors to whom it allots cheap IPO shares. This is not the explicit arrangement outlawed by the US Securities Exchange Commission and the UK’s Financial Services Authority after the Wall Street Settlement of 2003. Rather, institutions, which can direct their secondary trading business wherever they like, place it with the banks that feed them IPOs. This is an implicit quid pro quo, whereby part of the IPO discount is effectively rebated to the bank—over and above their fees on the deal.
My Oxford colleague Tim Jenkinson and I have conducted a survey of London-based institutions on just these questions. We found that little information is produced by institutions or revealed to banks during bookbuilding, which casts doubt on the “academic” explanation. We also heard that investment banks tend to allot shares to investors who give those banks their secondary trading business. Now, since these investors are nowadays largely hedge funds, they are unlikely to be the investors the issuer would choose to have as its shareholders. So, investment banks are probably not giving issuers the shareholders they want. Instead, our research squarely supports the “quid pro quo” interpretation of bookbuilding.
We are not alone in this interpretation. Issuers have tried in the last few years to reverse the bias against them by adjusting IPO design. For its flotation in 2004, Google promised to run a real Dutch auction, with shares going to the highest bidders, and investment banks excluded from allotment decisions. In the event, the Google IPO was launched in a tough market and it’s not certain that the IPO was, in fact, run this way.
Also in that year, the French company Pages Jaunes was sold by France Télécom in a “competitive IPO”. Rather than appoint the banks months in advance to run the IPO, and see its own influence over the process wane, France Télécom waited until well after the IPO was launched before formally appointing firms. This worked, but ‘competitive IPOs’ since then have had mixed success. FSA warns that they encourage banks to write inflated research opinions in the hope of being mandated. Market critics say competition is destructive so late in the process, because banks hide bad news from the issuer, e.g. on market conditions, to avoid looking half-hearted and being passed over for the top role.
The solution for issuers may be in an innocuous-looking rule change by FSA, which says banks bringing IPOs must be able to demonstrate that IPO allotments are unconnected to other business done with the same clients. Our survey, carried out in 2005-2006, suggests the connection then was close and that there was a quid pro quo between bank and investor. If FSA enforced its own rules, the rebate would disappear and banks would have to use cheap IPO shares for other purposes, like getting the price right or giving issuers the shareholders they want.
Send your comments to firstname.lastname@example.org
Howard Jones is director of the MSc in finance economics programme at the Saïd Business School, University of Oxford