A correction published Nov. 20th, 2002, which is too extensive to fit here, pertaining to this article has been appended to the beginning of the text below.

*Published Correction* An article and graphic in the Nov. 11 Business section misstated and omitted certain facts that may have created a false impression about the role of local businessman Jonathan Ledecky in the affairs of four companies. In each case, Ledecky was no longer-if he had ever been-an officer or director of the companies when they experienced financial difficulties.

Ledecky sold a company, E2E Net, to U.S. Technologies, but Ledecky was never an officer, director, controlling shareholder or employee of U.S. Technologies and therefore did not "help" or participate in the management of the combined entity or its portfolio companies. There is one lawsuit, not several, alleging fraud against C. Gregory Earls in his role as U.S. Technologies' chief executive. There are no lawsuits alleging fraud against the company, which is a high-tech incubator, not a "roll-up," as incorrectly described in the article. The company had an accumulated deficit, not outstanding debt, of $68.9 million on March 31.

At U.S. Office Products, Ledecky had resigned as chief executive and president two months before the company announced a recapitalization that included borrowing $1 billion. When the recapitalization was announced in January 1998, Ledecky said he would step down as chairman on the day the transaction was completed. New management took over in June 1998. U.S. Office Products declared bankruptcy in March 2001, nearly three years after Ledecky left the company.

Ledecky provided "seed" capital to OneMain.com but was never an officer or director of the company. OneMain was acquired by EarthLink in September 2000 for $308 million in cash and stock.

Ledecky had already resigned as Building One Services Corp.'s chief executive and president when, in March 1999, the Apollo Group offered $22.50 a share for approximately 60 percent of the company in a recapitalization. At that time, the company had $200 million in cash and no debt. The recapitalization was complete in May 1999. Ledecky stepped down as chairman of the company in February 2000, when the company merged with Encompass Services Corp., and had no further management role. Encompass filed for bankruptcy protection this week. (Published 11/20/02) Lost in the recent uproar about the accounting problems of Washington-based U.S. Technologies -- and how much William H. Webster, the choice to head a new accounting-oversight board, knew about them -- was that the concept behind the now-insolvent Internet incubator was fatally flawed.

U.S. Technologies Inc., the troubles of which indirectly led to the resignation of Securities and Exchange Commission Harvey L. Pitt last week, tried to combine two of the most spectacular failures of recent years: the dot-com, and the structure known as a roll-up.

The roll-up was designed to amalgamate, with public financing, a number of small private companies within a fragmented industry, such as auto repair or funeral services. The Washington area was a hub of roll-up activity, led by entrepreneur Jonathan J. Ledecky, whose early but fleeting success inspired others. Eventually, more than 100 roll-ups were created in the mid-1990s.

Today, most of them are gone or struggling to survive. Nearly 90 percent of the remaining independent roll-ups have lost money for their investors, and half are in bankruptcy or trade for less than $1 a share, according to Advest Inc., a Hartford, Conn., brokerage.

"The concept behind roll-up companies was compelling, but the execution was extraordinarily difficult," said George Yeonas, who was chief executive of the Fortress Group Inc., a McLean-based roll-up home builder, when it closed in July. "I think there was a certain amount of naivete about how bloody difficult they really are to create."

Although roll-ups have been around since the 1970s, Ledecky got in at the start of the boom in 1994 when he combined four office-supply businesses to create U.S. Office Products Co. He took the new company public in 1995 and then acquired more than 260 competitors over the ensuing three years. In its first 16 months, the company's stock price more than quadrupled.

Ledecky left U.S. Office Products in 1996 to form U.S.A. Floral Products Inc., which united florists around the country. Shares of that company nearly doubled within four months of its initial public offering.

In 1997, Ledecky created a blind roll-up; investors did not know what industry or companies he was going to target. Nevertheless, Ledecky sold more than $500 million in stock and started rolling up firms that serviced and maintained commercial buildings.

With Ledecky's apparent success as a model, entrepreneurs and investment houses started roll-ups in almost every industry.

Using the proceeds from the IPO and debt offerings, the new company would go on an acquisition binge. The theory was that the resulting economies of scale would increase earnings, which would enhance the roll-up's stock price, enabling more acquisitions and growth.

Everyone seemed to win at first. The entrepreneur instantly controlled a large public company; Ledecky went from being a small-time executive to joining the Washington power elite. The owners who sold their companies into the roll-up received more value than they could realize in a private sale. Investors watched the value of their holdings expand exponentially. And the investment-banking firms that underwrote the IPOs collected substantial fees.

It was in this exuberant market that James J. Martell created the Fortress Group.

In April 1996, Martell rolled up four home-building companies: Buffington Homes in Austin; Genesee Co. in Denver; Sunstar Homes in Raleigh, N.C.; and Christopher Homes in Las Vegas. Fortress sold $27 million worth of stock and $100 million in bonds in its initial public financing. The company leased a posh suite of offices in Tysons Corner, next to the Ritz-Carlton. Martell talked about reaching $1 billion in sales within four years.

But things went bad almost from the start.

A Weak Foundation

For most of its history, Wall Street has viewed home-building stocks cautiously, considering them primarily interest rate plays. In recent years, several home-building companies, most notably Pulte Homes Inc. and Lennar Corp., have grown to a sufficient size and reputation to overcome investors' negative perceptions, but Martell ran right up against them.

Fortress initially proposed to price its stock at $11 per share, but it was able to fetch only $9 per share. The bond sale went even worse: Originally the bonds were supposed to be redeemable after 2001 and mature in 2006. But they ended up non-callable, maturing in 2003, and bearing interest at a crippling rate of 13.75 percent.

"Those bonds are the whole Fortress story," said Ed Horne, who sold Wilshire Homes Inc. to Fortress in 1997 and then bought it back this year. "The bonds were dead weight. We made money everywhere, but we couldn't get out from under the debt."

To stay ahead of the debt payments, Fortress went on a buying spree. Within two years, the company acquired eight more companies. In 1998, Fortress made a profit of almost $13 million on revenue of $693 million, and it sold more than 4,000 houses.

Despite that apparent success, however, Fortress Group was a house of cards.

"To do a roll-up successfully, you have to do three things well: acquire, integrate and operate a business plan," said Jeffrey R. Manning, managing director of the special-situations practice at the Baltimore investment firm Legg Mason Inc. "The first one, acquiring, is actually the easiest of the three. Integrating is by far the hardest. And if you can't integrate, operating the business plan becomes tremendously difficult."

Martell and his management team were not integrators. They were not interested in integrating; they wanted each of the companies to operate autonomously, with the main office in McLean supplying public money to build the top line and management expertise to boost the bottom line.

Each division had its own accounting system, its own e-mail system, even its own health-insurance provider.

"The way they structured the company was very decentralized," said Scott Campbell, a housing analyst at Raymond James & Associates Inc. in St. Petersburg, Fla. "Unfortunately, that didn't work."

Seeing that the strategy was failing, Fortress's board ousted Martell in July 1998 and replaced him with Yeonas, who had been recruited a year earlier to be chief operating officer.

Integrating was an "inordinate challenge," Yeonas said. "We didn't even have a single definition of what constituted a sale of a house."

With the bond debt hanging over their heads, Yeonas and the board started selling assets. Some divisions, such as Christopher Homes in Las Vegas, were losing money. But operations in Florida, North Carolina and South Carolina were profitable, and selling them cut deeply into the muscle of the overall organization. In many cases, Fortress sold the operations back to their original owners; in almost all cases, Fortress sold at a loss.

Selling the operations in the Carolinas to Lennar in late 2001 cut its senior debt to less than $33 million from $100 million , but Fortress had only its Texas and Colorado divisions left. Operating losses for the year totaled almost $9 million.

"By the time we sold the Carolina division, the handwriting was probably on the wall that we were no longer large enough to really be a true public company," said Robert Short, who was one of the original four builders and the only one to stay with the company until the end.

Earlier this year, Fortress sold the Texas operation back to Horne and its Denver division to Lennar. It used to the proceeds to pay off the remaining debt and give stockholders $3.68 per share.

Incompatible Systems

Despite its missteps, and ultimate failure, the Fortress Group fared better than many roll-ups. According to Legg Mason's Manning, most roll-ups failed within three years. U.S. Technologies lasted a little more than two years. Another Ledecky venture, OneMain.com, crashed after just 18 months, after burning through $200 million in investor funds.

Fortress Group's problems were the same ones that doomed so many other roll-ups: too much debt and inability to integrate the acquired companies.

In the case of U.S. Office Products, long-term debt ballooned from less than $19 million at the end of its first fiscal year to nearly $1.2 billion five years later, when the company filed for bankruptcy protection. The debt of Ledecky's second roll-up, U.S.A. Floral Products, climbed to almost $200 million in one year from $309,000. It filed for bankruptcy protection within a month of U.S. Office Products. Ledecky's third venture, Building One Services Corp., was near bankruptcy when it was acquired by a Houston firm in 2000.

Ledecky also was involved in U.S. Technologies. He created E2Enet Inc. in 1999 as an Internet incubator. After abandoning efforts to go public, E2Enet agreed to be acquired for $11 million by U.S. Technologies, which tried to adopt Ledecky's strategy. But its acquisitions didn't pan out, and by early 2002 U.S. Technologies had accumulated $68.9 million in debt and stated in a regulatory filing that there was "substantial doubt" about its survival.

The integration issue is equally problematic. Roll-ups don't start with a base operating system to which the acquired companies must conform. Often, officials at each acquired company think their systems are best. Also, the financial types who put together the roll-ups often do not have the managerial skills to integrate the businesses.

A key issue is that small private companies, especially start-ups, frequently have poor financial reporting systems. Not only do they struggle to accurately forecast quarterly performance, they often are hard-pressed to report their past performance on a timely basis.

U.S. Technologies was a prime example. According to its former auditor, BDO Seidman LLP, the firm had problems with business basics such as retaining and storing documents and promptly recording transactions. During Webster's tenure on the board and the company's audit committee, from April 2000 until July of this year, U.S. Technologies filed its quarterly reports late four times.

By 2000, according to a letter BDO Seidman sent to the audit committee last year, "the company's financial and accounting infrastructure became inadequate for purposes of properly addressing certain financial and accounting issues that the organization faced given its size, complexity and public stature."

The auditing firm contends it was fired after bringing those problems to the attention of Webster and his audit committee; Webster denies that, although BDO Seidman has released documents that it says prove its assertions. But Webster was sufficiently concerned that U.S. Technologies' history might pose problems for his nomination that he warned Pitt, whose failure to pass on the information to his fellow commissioners before they narrowly approved Webster's appointment proved to be his final misstep.

Predictable Results

Roll-up artists "may be really smart people, but what are needed are people who are into pennies," Manning of Legg Mason said. "They need the discipline and ruthlessness to force other people to adhere to their business model."

Without integration, many of the benefits that are supposed to accrue from a roll-up never materialize. The projected economies of scale are undermined by territorial in-fighting and redundancies. In the final days of U.S.A. Floral Products, there were more than 150 middle managers each earning more than $150,000 a year.

Greed also played a large role in the rise and fall of roll-ups. Lured by early promise of quick returns and big fees, investors and investment bankers constructed roll-ups in ill-suited industries, such as orthopedic practices, or rolled-up companies with poor fundamentals.

The results were predictable. According to a March 2002 study by Jeffrey Evans of Advest, of 113 recent roll-ups that had not been acquired, 17 were in bankruptcy protection and 83 were trading below their offering price; 39 of the companies had a share price of less than $1. Wall Street finally woke up to the dismal performance of roll-ups in late 1999. Since then, growth-oriented companies have adopted a consolidation strategy: A large public company acquires a small public or private company in its industry, integrates the new division into its business model and then goes out and buys another.

"Over the long term, consolidation has proven to be a successful strategy. Just look at General Electric," Manning said. Conversely, he said, "You could practically count the number of successful roll-ups on one hand."

J. Marshall Coleman, front, James J. Martell, top, and George Yeonas, right, ran the Fortress Group, a McLean roll-up.