Lehman/VERT revisited (nice piece of journalism, the author is to be commended):
interactive.wsj.com
July 24, 2000
Raising the Stakes: As Wall Street Seeks
Pre-IPO Investments, Conflicts May Arise
By MARK MAREMONT
Staff Reporter of THE WALL STREET JOURNAL
In December 1999, the investment bank Chase H&Q published an analyst's
report praising Infospace Inc. as a "must-own holding." A few weeks later,
H&Q reiterated the "buy" recommendation.
But while the bank's analysts were telling customers to load up, H&Q itself
was bailing out. In late January, with the buy rating still in place, the Chase
Manhattan Corp. unit sold all of the Infospace stock owned by its
venture-capital arm, plus a stake owned by an H&Q employee fund, for
about $72 a share, or a total of roughly $42 million.
H&Q and some fortunate employees pocketed a 7,000% gain in 18 months
in a company H&Q had taken public in late 1998. Investors who bought
Infospace stock early this year based on H&Q's repeated recommendations
aren't doing as well. Some may even have lost money. The stock is currently
trading at about $48.
Questions of Credibility
Lured by huge returns in technology stocks, Wall Street banks have
dramatically stepped up their venture-style investments in the past year or
so. But as investing in privately held companies on the verge of going public
becomes a new profit center on Wall Street, it also is creating potential
conflicts of interest for the banks.
An investment bank's analyst, for example, may be tempted to boost the
stock of a company in which his employer -- and possibly the analyst himself
-- owns a stake. This possibility, while almost impossible to prove in
particular cases, nevertheless can erode the credibility of analysts'
recommendations, on which many investors rely.
Wall Street houses deny that their venture investing creates any conflicts. A
Chase Manhattan spokesman says that H&Q's sale of Infospace stock in
January was "completely separate" from the H&Q analyst
recommendations.
Not every investment firm is selling while its analysts tell the public to buy, of
course. But selling a stock while simultaneously recommending it is just the
final stage of what some bankers and venture capitalists say is an
increasingly common cycle on Wall Street: First, an investment bank secures
a stake in a promising start-up, sometimes by suggesting a willingness to
take it public. After a successful IPO, the bank's research department issues
positive opinions on the stock. When securities regulations allow it to sell its
venture stake, the bank does so, frequently reaping a big profit.
A New Model
In some cases, analysts and bank executives have a personal interest in the
outcome, either by means of individual pre-IPO investments, or, more
commonly, through employee venture pools that many banks recently have
set up.
Ted R. Dintersmith, a principal at Charles River Ventures, a Waltham,
Mass., firm that has backed many successful start-ups, believes the pattern
represents a "new investment-banking model," in which Wall Street firms
"push like crazy to get a large amount of [their own] money" into a promising
young technology firm, then "turn around and take the company public."
"Where it crosses the line," Mr. Dintersmith adds, is when "you see analysts
at investment-banking firms issuing aggressive buys at high prices and see
the same investment banks aggressively selling. Something about that seems
highly inappropriate." (The question of whether securities-fraud laws were
violated would arise only if there were manipulation -- for instance, if a
bank's venture-investment executives asked an analyst to effectively deceive
investors by postponing downgrading a stock to allow the bank to sell its
stake at a higher price.)
Growing Trend
Controversy has percolated for years over the widely held perception that
some Wall Street analysts issue glowing reports on companies to help attract
and keep those companies as investment-banking clients. Now the
phenomenon appears to have escalated. Banks are seeking opportunities to
make pre-IPO investments with potential payoffs dwarfing ordinary Wall
Street fees. And the banks' chances of landing those pre-IPO stakes are
enhanced because technology start-ups are eager to associate with Wall
Street institutions that can take them public and then publish positive
analysts' reports about them.
Bankers generally brush off such concerns. They point out that they have
navigated similar potential conflicts for years. Their money-management
arms often buy and sell stocks that are publicly assessed by their research
departments, the bankers say. Analysts, they add, operate independently
from traders and investment bankers.
It's beyond debate that venture investing has become more important to
Wall Street. Surveys by VentureOne, a San Francisco research firm, show
investment banks put money into 305 venture-backed companies last year,
up from 104 in 1998. Some of the deals have been huge hits, making
venture investing a significant source of profits at many banks. Goldman
Sachs Group Inc.'s $36 million pre-IPO investment in StorageNetworks
Inc., a provider of computer storage services near Boston, is now valued at
about $1.6 billion.
Between 13% and 18% of Goldman's net income last year came from
"private equity" gains, primarily venture-style technology investments,
compared with 4% to 5% in 1998, according to Salomon Smith Barney
analyst Guy Moszkowski. At Lehman Brothers Holdings Inc., he says, such
gains jumped to between 17% and 22% in this year's first quarter, from
about 4% of net income for all of last year; much of the increase came from
a huge windfall in VerticalNet Inc., a provider of online business-to-business
trading sites that Lehman invested in and then took public.
Although some banks' venture investments drooped in the second quarter of
this year, overall, "banks have figured out that the investing business is more
attractive than their fundamental business" of providing advice and
underwriting, says Greg Taxin, a former Goldman technology banker who
left earlier this year for Epoch Partners, a smaller investment bank in San
Francisco. "At a lot of places," including Goldman, he adds, "this is driving
the firm's economics." Mr. Taxin says many of the most able people he used
to work with at Goldman have requested transfers to the bank's
venture-investing area, "in part, because they know that's where the bread is
getting buttered."
Gains from venture investing now sometimes eclipse Wall Street's
more-traditional underwriting fees, as was the case with Vignette Corp., an
Austin, Texas, firm that makes Internet customer-tracking software. Vignette
went public in February 1999, paying underwriting fees of less than $6
million to a syndicate of 13 investment banks. But three Wall Street firms,
including Vignette's two main underwriters, also had venture stakes in the
company that have since yielded profits totaling nearly $135 million.
Wall Street's honeymoon with Vignette started with Goldman and H&Q,
which put money into the software start-up 10 months before its IPO. Both
banks wanted to take Vignette public, says the company's chief executive,
Greg Peters. But Vignette chose Morgan Stanley Dean Witter & Co. to
lead the offering, partly because the company wanted Morgan Stanley's star
Internet analyst, Mary Meeker, to publish research on the stock, Mr. Peters
says.
Goldman bowed out of the offering, while H&Q stayed on as a secondary
underwriter.
When Morgan Stanley put $1.5 million into Vignette in late 1998, Ms.
Meeker was closely involved in the decision. She "was obviously very
positive" about the investment, says Mr. Peters. "That's why they did it."
Morgan Stanley spokesman Raymond O'Rourke confirms Ms. Meeker was
involved in the bank's investment but says "she didn't make the decision
unilaterally." Ms. Meeker declines to comment.
Boosting the Stock
Vignette's stock quickly soared from its split-adjusted IPO price of $3.17.
The swift rise caused Ms. Meeker to be cautious at first about the company
in her published research. She became more bullish as 1999 wore on and
Vignette's revenues beat projections, although the company was still losing
money. In January of this year, she and a colleague heralded Vignette's
"knockout" fourth quarter of 1999 in a report headlined, "They crushed the
nums."
Morgan Stanley was required by securities regulations to hold its Vignette
stake for a "lockup" period of a year after the offering. In March, a month
after the lockup ended, and just two weeks after the stock briefly touched
its peak of $100, the bank sold half of its Vignette shares for about $76 a
split-adjusted share, or $41.2 million. That was a 5,400% gain in just over a
year.
While Morgan Stanley was selling, Ms. Meeker was telling investors to buy.
She rated the stock "outperform," the second-highest ranking on her firm's
four-point scale and a rating she had maintained since the prior July. The
bank's research reports included a standard footnote, saying Morgan
Stanley or its employees "have or may have a long or short position or
holding" in the securities mentioned.
Vignette shares closed Friday at $48.0625, 37% below the price at which
Morgan Stanley sold in March. The bank's remaining stake is now valued at
about $26 million.
Morgan Stanley's Mr. O'Rourke denies any connection between the firm's
positive rating on Vignette and its decision to sell half of its own stake. "We
have been a long-term investor in Vignette, so no one should be surprised
that we would eventually sell part of our position," the spokesman says. He
says that the sale came 15 months after the venture investment and eight
months after the bank's research department first urged investors to acquire
the stock.
The pattern was the same with Chase H&Q's venture investment in
Vignette. H&Q's research department maintained a buy rating on the stock
beginning in March 1999. But between August 1999 and January 2000, the
bank and a fund made up of H&Q employees and others filed papers with
the Securities and Exchange Commission declaring their intention to sell
more than $25 million of Vignette stock. Such filings usually result in
short-term sales, but H&Q won't confirm the total actually sold. (H&Q was
called Hambrecht & Quist Group Inc. until it was acquired by Chase last
year.)
Goldman sold about $75 million of Vignette shares in the second half of last
year for about 15 times its original per-share price. Goldman analysts never
covered the stock.
Wall Street's Appeal
For those hot start-ups that, despite the overall decline in Internet stocks,
still have their pick of pre-IPO financiers, Wall Street has an edge over
other sources of venture capital: the capacity to bring fledgling companies
public.
Some banks are telling tech start-ups that they are more likely to take them
public if the banks get an opportunity to make pre-IPO investments,
according to Mr. Dintersmith and others familiar with the venture-capital
world. "Not only is the venture-capital tail wagging the investment-banking
dog, it's crushing it," Mr. Dintersmith adds. He declines to identify on the
record any banks that employ this quid pro quo approach, and most
bankers deny they use it.
But Mr. Taxin, the former Goldman banker now with Epoch Partners, is an
exception. Banks faced with stiff competition from other financiers
increasingly are "trying to strong-arm companies" into letting them invest, he
says.
Mr. Taxin says a pitch he twice made successfully while at Goldman went
something like this: "We've got the best analyst in [your area], and we'd
really like to work with you. But we have a lot of opportunities. If you really
want to work with us, you have to give us a slice of your private round" of
pre-IPO financing. Mr. Taxin says he endorses this approach because if a
bank is ready to take a company public, it should be willing to invest in the
company.
Banks Less Selective?
Some in the investment community muse that a hunger for venture-investing
profits may even contribute to some banks becoming less selective about
which companies they take public.
But at Goldman Sachs, Joseph H. Gleberman, co-head of the firm's own
technology investments, strongly rejects the notion that his bank would
underwrite a weak IPO just because it had a venture stake. He also says
that Goldman wouldn't put any sort of quid pro quo proposal to a start-up
client. He acknowledges that for many start-up companies, "the fact that we
have this terrific banking franchise is a very attractive part of having us as an
investor." But he says that Goldman's venture and investment-banking arms
"make their own decisions."
So why do these two arms often encircle the same tech start-ups? The firm's
investment bankers are scouting out "leading companies" as potential clients,
while the venture unit is "out looking for exciting, dynamic growth
companies," Mr. Gleberman says. "There are some that overlap."
A look at the most recent nine months of cases of such overlap -- where
Goldman had venture stakes in its tech-IPOs -- reveals less-than-stellar
results:
In the last quarter of 1999 and the first two quarters of 2000, Goldman led
IPOs for nine technology firms in which it had prior venture investments,
including Webvan Group Inc., PlanetRx.com Inc., FreeMarkets Inc. and
Nuance Communications Inc. Of the nine, three are trading below their IPO
price; three others are trading below their first-day closing prices.
StorageNetworks, by contrast, the company in which Goldman has run up a
huge paper profit, has gone up fivefold since its IPO in late June; Nuance
has similarly soared. And to be sure, Goldman and other banks take tech
companies public in which they haven't invested, and there is often fierce
competition among banks for IPO assignments.
A Venture Pioneer
Chase H&Q, in San Francisco, regards itself as a pioneer in closely
integrating venture dealings with investment banking, according to former
employees. "It's the goal of any banker, especially one financing companies
in the New Economy, to get to know great companies at very early stages
of development," says Robert Keller, a former head of H&Q's Internet
investment-banking practice who in April joined Rosewood Venture Group
in San Francisco. "What better way to get to know a company than
investing in that company?"
Like an increasing number of investment banks that have lost top people to
tech start-ups, H&Q uses employee venture-investment funds, and in some
cases individual investment opportunities, as incentives to stay and excel.
One who has benefited is Daniel H. Rimer, the
senior H&Q Internet analyst who issued some of the
firm's positive research reports on Vignette. Mr.
Rimer says he was individually given a small
percentage of H&Q's total investment in the
software company as a "sourcing fee" for helping
find the deal. He says he also made a separate
pre-IPO investment in Vignette through an H&Q
fund, although he declines to specify the amount of
these investments.
Considered among H&Q's best deal spotters, Mr.
Rimer, a 29-year-old Swiss native educated at
Harvard University, says he helped the bank make more than 25 venture
investments, including those in Vignette, Inktomi Corp., Critical Path Inc.
and Infospace. H&Q helped take all four companies public, and Mr. Rimer
wrote bullish research reports on all four. He says he received equity
sourcing fees on some of the deals in which he participated, but they weren't
enough to keep him at the bank. In late 1999, he left to join a venture capital
firm run by Netscape Communications Corp.'s former chief executive,
James Barksdale.
Mr. Rimer's and H&Q's involvement with Infospace offers an illuminating
case study of the new style of combining venture financing with traditional
investment banking.
In mid-1998, Mr. Rimer was part of an H&Q team that made a pitch to
Infospace, a provider of Internet information services that was hoping to go
public. Naveen Jain, a former Microsoft Corp. executive who founded
Infospace, says he ultimately chose H&Q over rival bankers largely because
he was impressed with Mr. Rimer's grasp of the company. As the analyst,
"he's your salesman to investors," Mr. Jain explains.
Based on Mr. Rimer's recommendation, H&Q also was eager to invest in
the company, Mr. Jain recalls. He welcomed that eagerness. H&Q's
obtaining a stake would help motivate the bank to "make it a successful deal,
both at the IPO and afterward, when they would support the stock," says
the Infospace chairman.
In early August 1998, H&Q and a limited-partnership fund for employees
and others paid a total of $2 million for just over 1% of Infospace stock, at
a price of $1 per share, adjusted for splits. Mr. Rimer says he participated in
the deal through the fund but can't recall his precise investment. Three weeks
later, Infospace filed with the SEC for its initial public offering.
The company went public in December 1998 at a split-adjusted $1.875,
and the stock quickly shot up. It stayed in the teens for most of 1999, but
then, with Mr. Rimer still praising the stock in his research reports, the price
started to rise sharply. In late January 2000, a month after the lockup period
expired, came the sale in which H&Q and its employees made their 7,000%
gain. Participants in the limited partnership also have sold some Infospace
shares separately.
The stock briefly soared to a high of about $130 in March before falling to
close Friday at $48.0625 on the Nasdaq Stock Market.
After Mr. Rimer left H&Q, his successor as Internet analyst maintained the
buy rating for Infospace and kept it on the firm's "Focus" list of especially
promising stocks. Both during Mr. Rimer's tenure and after, all of H&Q's
reports contained footnotes noting that the firm and the responsible analyst
had "an investment position" in Infospace.
Perception vs. Reality
Chase spokesman John Meyers says the H&Q executives responsible for
the bank's venture portfolio "don't care what the analyst's recommendation
is." That's because of what Mr. Meyers says is the bank's policy of selling its
venture shares "as quickly as possible" after the lockup ends, so that the
money can be reinvested in privately held companies.
"This is one of those cases," Mr. Meyers says of the Infospace engagement,
"where there is an appearance of a conflict of interest, but there is none" in
fact.
Chase H&Q's "quick-as-possible" sale policy doesn't appear to be hard and
fast, however. SEC filings indicate that in the case of Vignette, the bank sold
some of its stock over a five-week period; H&Q then filed papers five
weeks later, indicating a proposed sale of another chunk of stock.
Mr. Rimer says the bank's venture-investing executives generally would ask
his opinion about whether to sell a particular stock in their portfolio but
"would make their own decisions" and wouldn't tell him when they were
selling.
He denies that he had any conflict of interest when recommending Infospace
or any other stock in which he and his firm held an interest. Instead, he says
his and the bank's venture investing amount to "putting your money where
your mouth is." The lengthy lockup provisions in such deals, Mr. Rimer
adds, "in a lot of ways makes you more objective." |
