What do oil and Google have in common? Well, beyond the obvious – they both contain the letter “o” – the price of both skyrocketed late last week.

On Friday, U.S. crude oil hit a record at $49.40 on the New York Mercantile Exchange, warming the hearts of oilmen and scaring the heck out of everyone else. By Friday afternoon, U.S. crude for September delivery had failed to touch $50, settling just below $48 per barrel.

That took selling pressure off of stocks, and the market rallied.

Google priced its IPO at $85 per share on Wednesday night and it shot up more than $15 on Thursday, its first day of trading. By the close on Thursday, August 19, Google’s stock was up 18 percent, at $100.33 per share. On Friday, August 20, Google leaped again, shooting up almost 8 percent to close at $108.31.

By the time the weary denizens of Wall Street headed out the door after Friday’s closing bell, the price of one share of Google was up $23.31 – or 27.4 percent – from its IPO price. Meanwhile, the lesser IPOs fell by the wayside.

The Nasdaq trade winds

We may never know Google’s secret, but it was clear that the stock market began to blow warm trade winds on the back of the newly minted public company.

The Nasdaq Composite Index closed the week at 1,838.02, up 4.6 percent from one week ago – the sharpest weekly gain since the week ending April 2, 2004. Nevertheless, for the year, the Nasdaq was still down 8.25 percent from its close at 2,003.37 on December 31, 2003.

At the beginning of the week, the IPO calendar showed three deals, plus Google, which was listed as a “possible.” By the end of the week, none of the three scheduled three IPOs made it to market. Google, of course, did.

Two IPOs were postponed “due to market conditions.” The one remaining IPO was rescheduled and listed as “day-to-day.”

Getting the hook

Pulled off the IPO calendar werePRN, a San Francisco, California-based broadcasting company that operates the fifth-largest broadcast network in the United States, and WebSideStory, a San Diego-based provider of on-demand Web analytics services.

The bankers of the WebSideStory IPO even cut the size of the offering to 5 million shares at $8 to $9 each, down from an initial filing range of 5 million shares at $10 to $14 each. Apparently the deal didn’t draw enough interest to get it into the market.

Only the lonely

eCOST.com was the only deal left on the calendar. Based in Torrance, California, the company is a multi-category online discount retailer of high quality new, closeout, and refurbished brand-name merchandise.

The bankers of the eCOST.com IPO even cut the size of the offering to 3.15 million shares at $7 each, down from an initial filing range of 3.15 million shares at $9 to $10 each. Apparently bankers believe the deal will draw enough interest to get it into the market sometime during the week of August 23.

Spread formations

Now back to last week and the Google IPO. One of the original selling points of the Dutch auction bidding system was the savings it offered for the company going public. The underwriting discount for Google was just that.

The underwriting fee, called a spread in Wall Street jargon, was $46.7 million, as reported in the final prospectus. That was 2.8 percent of the total amount raised. Google priced 19.5 million shares at $85 each to raise $1.667 billion.

Wall Street’s underwriting discounts, or spreads, generally run as high as 10 percent or as low as 2.8 percent. The spread for a small-cap IPO, under $10 million, is 10 percent. For a normal IPO, the spread is 6.5 to 7 percent. The spread for blockbuster IPOs, more than $1 billion, varies between 3 percent and 3.5 percent.  

In the largest U.S. IPO in 2000, when AT&T Wireless priced 360 million shares at $29.50 each to raise $10.6 billion, the underwriting discount was $309 million, or 2.9 percent.

The fee structure depends on the risks involved for bankers, who have to buy all unsold shares for their own accounts. If the bankers misjudge the market, the cost can add up. They could end up owning millions of shares at the offering price (less the underwriting discount) and see the IPO trade at a sharply lower price. This was not an uncommon event in the 70s and 80s.

In contrast, the Dutch auction system is much like a “best efforts” approach. In a “best effort” offering, the bankers will sell what they can and assume no liability for any unsold shares.

The Dutch auction is not an underwriting – it is an auction. And the offering price is set at the highest price it takes to sell all the shares being offered by the issuer or issuers. This sharply reduces the risks an underwriter faces, and the underwriting discounts can be lowered. But this isn’t always the case.

Exclusive of the Google Dutch auction (the underwriting discount was 2.8 percent, just below the 2.9 percent of the blockbuster AT&T Wireless deal), the average underwriting discount for all 10 Dutch auction IPOs ever done was 6.04 percent.

Dutch history

The first Dutch auction IPO was Ravenswood Winery, a Sonoma, California-based wine producer. Ravenswood Winery priced its IPO of 1 million shares at $10.50 each to raise $10.5 million in April of 1999. The underwriting discount was $420,000, or 4 percent. In a traditional IPO, an underwriting of this size would have cost the company $1.05 million in banking fees.

Salon.com, a San Francisco-based media company, used the Dutch auction system for its IPO in June of 1999. Salon.com priced its IPO of 2.5 million shares at $10.50 each to raise $26.25 million. The underwriting discount was about $1.31 million, or 5 percent.

The last Dutch auction IPO was New River Pharmaceuticals, a Radford, Virginia-based pharmaceutical company that develops new versions of existing drugs. New River Pharmaceuticals priced its IPO of 4.2 million shares at $8 each to raise $33.6 million on August 8. The underwriting discount was $2.35 million, or 7 percent.

Overall, excluding the blockbuster Google deal, the Dutch auction system has raised $329.3 million, with total underwriting discounts of $19.9 million, or 6.04 percent.

The underwriting discount, or spread, is a three-part charge that the issuer pays the bankers. The first part is the management fee, divvied up among the managing underwriters. The next part is the underwriting fee, shared by all the underwriters. The third part is the selling concession, essentially a commission, which goes to the brokers selling the deal.

Besides the excessive hype and hooplah, there was one interesting thing about the Google IPO: there was no selling concession.

So it’s clear why no investors received a late evening phone call from their stockbrokers, offering an investment opportunity they couldn’t afford to pass up. Maybe the investment banking firm got a fee, but the brokers handling the Google IPO got no commission.

No crocodile tears, please

So the underwriters did not make out like bandits on the Google IPO. Instead, they earned a spread of 2.8 percent, or $46.7 million.

In Wall Street terms, those aren’t big bucks – but it’s not chicken feed, either. Google went public and raised $1.67 billion, and it paid its underwriters at a discount – $46.7 million. Even on Wall Street, that’s enough to cover some nice bonuses and stock up on cases of Dom Perignon.

And what about the IPO investors? Well, they got a chance to buy Google stock at $85 per share in the Dutch auction, and then wade in again to buy more shares at a premium in the aftermarket.

Bankers and stock buyers may not have gotten everything they hoped for with the Google IPO, but they didn’t wind up empty handed, either.