Stocks were tumbling last fall as the new school year began, but at Harvard University, it was as if the boom had never ended.

Workers were digging across the river from Harvard's Cambridge, Mass., home, the start of a grand expansion that was to eventually almost double the size of the university. Budgets were plump, and students from middle class families were getting big tuition breaks under an ambitious new financial aid program.

The lavish spending was made possible by the earnings from Harvard's $36.9 billion endowment, the world's largest. That pot was supposed to be good for $1.4 billion in annual earnings.

Behind the scenes, though, a different story was unfolding.

In a glassed-walled conference room overlooking downtown Boston, traders at Harvard Management Co., the subsidiary that invests the school's money, were fielding questions from their new boss, Jane Mendillo, about exotic financial instruments that were suddenly backfiring.

Harvard had derivatives that gave it exposure to $7.2 billion in commodities and foreign stocks. With prices of both crashing, the university was getting margin calls--demands from counterparties (among them, JPMorgan Chase and Goldman Sachs) for more collateral. Another bunch of derivatives burdened Harvard with a multibillion-dollar bet on interest rates that went against it.

It would have been nice to have cash on hand to meet margin calls, but Harvard had next to none. That was because these supremely self-confident money managers were more than fully invested. As of June 30, they had, thanks to the fancy derivatives, a 105 per cent long position in risky assets. The effect is akin to putting every last dollar of your portfolio to work and then borrowing another 5 per cent to buy more stocks.

Desperate for cash, Harvard Management went to outside money managers begging for a return of money it had expected to keep parked away for a long time. It tried to sell off illiquid stakes in private equity partnerships but couldn't get a decent price. It unloaded two-thirds of a $2.9 billion stock portfolio into a falling market.

Now, in the last phase of the cash-raising panic, the university is borrowing money, much like a homeowner who takes out a second mortgage in order to pay off credit card bills. Since December, Harvard has raised $2.5 billion by selling IOUs in the bond market. Roughly a third of these Harvard bonds are tax exempt and carry interest rates of 3.2 per cent to 5.8 per cent. The rest are taxable, with rates of 5 per cent to 6.5 per cent.

It doesn't feel good to be borrowing at 6% while holding assets with negative returns. Harvard has oversize positions in emerging-market stocks and private equity partnerships, both disaster areas in the past eight months.

The one category that has done well since last June is conventional Treasury bonds, and Harvard appears to have owned little of these. As of its last public disclosure on this score, it had a modest 16 per cent allocation to fixed income, consisting of 7 per cent in inflation-indexed bonds, 4 per cent in corporates and the rest in high-yield and foreign debt.

For a long while, Harvard's daring investment style was the envy of the endowment world. It made light bets in plain old stocks and bonds and went hell-for-leather into exotic and illiquid holdings: commodities, timberland, hedge funds, emerging-market equities and private equity partnerships.

The risky strategy paid off with market-beating results as long as the market was going up. But risk brings pain in a market crash. Although the full extent of the damage won't be known until Harvard releases the endowment numbers for June 30, 2009, the university is already working on the assumption that the portfolio will be down 30 per cent, or $11 billion.

The strain of market turmoil is visible in staff turnover at the management company, which axed 25 per cent of its staff recently and is on its fifth chief in four years. Mendillo, 50, came to Harvard last July after running Wellesley's small endowment.

She declines to comment. But how much blame she should get is unclear; the big bets on derivatives and exotic holdings were in place before she got there. The bad bet on interest rates--a swap in which Harvard was paying a high fixed-interest rate and collecting a low short-term rate--goes back to a mandate from former Harvard President Lawrence Summers.

Jack R. Meyer, 64, a revered money manager who headed Harvard's endowment until 2005, offers a few guarded comments. "The liquidity thing most concerns me--that should not have happened," he says. Though he wasn't there at the time, Meyer says Harvard Management bought the commodity and foreign stock derivatives as a way to get exposure to those asset classes while freeing up cash to put to work elsewhere. The strategy, he says, "drained liquidity" from the endowment in recent months. "Many endowments stretched too far, and I think Harvard did as well," he says.

The endowment will remain stretched. Harvard has been counting on it to fund more than a third of its $3.5 billion operating budget. Assuming the fiscal year ends with around a $24 billion endowment value, the university will be drawing down half again as high a percentage of its assets as it did in 2004, the last time the endowment was around that size.

That can't go on forever. The strain on liquidity will continue, as the private equity partnerships compel Harvard to meet billions in capital calls in future years. Why not just unload those partnerships along with the liabilities that stick to them? Because no one wants to buy them. Private equity stakes like Harvard's are selling at 40 per cent to 60 per cent discounts in various markets. "Endowments will be shocked at the valuations of their (private equity) portfolios," says Stewart Massey, an endowment consultant at Massey Quick. "It's going to be an absolute bloodbath."

Harvard's woes are in some ways no different from those at other universities or in the market generally (the S&P 500 is down 37 per cent since last July 1). "A loss in these kinds of markets is inevitable," says Michael Eisenson, a former HMC staffer who now runs private equity firm Charlesbank. The average endowment is down 23 per cent in the five months through November, according to a university trade group.

But Harvard was supposed to be different. In the 15 years through last June, it returned an annual 15.7 per cent versus 9.2 per cent for the S&P. Meyer landed at Harvard in 1990 after scoring big investment returns at the Rockefeller Foundation.

In an unorthodox move for an endowment chief, Meyer built a Wall Street-like trading operation and managed most of HMC's money in-house. It looked like a giant hedge fund, and it had paychecks to match. A high-level HMC manager would make as much as $35 million in good years.

Those sums triggered what became an annual Harvard tradition: first, the disclosure (compelled by tax laws applying to nonprofits) of the HMC bonuses, followed by an outcry led by the late William Strauss and a group of Harvard alumni from his class of 1969.

HMC not only became a place to make big bonuses, it was also where you could make a name for yourself and become a "crimson puppy," meaning launching your own private equity firm or hedge fund with Harvard's backing. One of the puppies, Jeffrey Larson, left in 2004 to start Sowood Capital. That pile of smart money cratered in 2007, losing $350 million for Harvard.

By September 2005, Meyer himself decided it was time to go. Some people say it was because of the persistent criticism about bonuses, which were reduced near the end of his tenure; others say he had run-ins with former US Treasury Secretaries Lawrence Summers and Robert Rubin, who assumed Harvard leadership positions at the start of the decade. Meyer denies both reasons and says 16 years at Harvard was simply enough.

Meyer formed his own hedge fund, Convexity Capital, which seems to have held up well in the current market. He took with him the Harvard heads of domestic and international fixed income and both their staffs, as well as the chief risk officer, chief technology officer and chief operating officer. The survivors were demoralized. "You walked onto the trading floor, and it was just 10 per cent full," says someone who was there at the time. "There was a sense that if you were good, you left."

Five months later, Mohamed El-Erian, now 50, took over. The son of an Egyptian diplomat, he had risen to deputy director of the International Monetary Fund before joining giant bond manager Pimco. He seemed perfect for smoothing relations between HMC and the university. Filling the hole that Meyer left was another matter.

One solution: Don't even try, just hand over all of the endowment to outside money managers. But El-Erian insisted on keeping things intact. He talked of the "structural advantages" of investing a big endowment backed by an AAA-rated university, such as allowing you to borrow at low rates when making leveraged bets. The former Pimco emerging-market superstar also believed that the developing countries offered big profits to smart investors like HMC because they had become less risky thanks to ample dollar reserves and a growing middle class.

So El-Erian upped HMC's exposure to emerging-market stocks, which rose from 6 per cent of assets when Meyer left, to 11 per cent two years later. He also used total return swaps to bet on developed world stocks and commodities on the cheap, freeing up money for other investments. Tapping former Stanford endowment staffer Mark Taborsky (an "important hire," El-Erian would later write in a book), El-Erian also took money from hedge funds he didn't like and redirected it to ones he thought were winners, putting hundreds of millions into funds in Latin America, Asia and the Middle East.

The moves looked brilliant. For the year ended June 2007, Harvard returned 23 per cent versus 17.7 per cent for 151 other big institutional investors (and 20.6 per cent for the S&P 500). Fearing all markets could soon fall, El-Erian injected what he referred to as "Armageddon insurance" into HMC's portfolio for the first time by buying interest rate floors, or a wager that rates would fall, and betting, via credit default swaps, that companies could soon struggle to pay their debts.

For the following year, through June 2008, Harvard gained 8.6 per cent, versus a 13 per cent fall in the S&P. El-Erian's insurance accounted for much of HMC's outperformance. Hedge funds, however, were sucking up cash--HMC had increased investments in those areas to 19 per cent from 12 per cent a year earlier. The returns were flat. It's unclear how much of the results--good or otherwise--were El-Erian's doing. He left at the end of 2007, six months before the results came in, citing a desire to move back near his wife's family in California and return to Pimco as heir apparent to founder Bill Gross.

Since July, emerging-market shares have been a disaster, falling 50 per cent, as measured by the MSCI Emerging Markets Index, worse than US stocks. Another problem: El-Erian's insurance has been partly taken off since he left, leaving HMC vulnerable when markets plunged this fall.

The total return swaps, which easily could have been terminated, were left alone. The EFG-Hermes Middle East North Africa Opportunities Fund, a hedge fund launched in September 2007 with some $200 million of HMC cash, was down 35 per cent in 2008. El-Erian's big hire, Taborsky, left HMC in September. He's since joined El-Erian at Pimco. El-Erian and Taborsky decline to comment.

By the time Jane Mendillo walked into HMC's offices in July 2008, she figured some changes needed to be made. A former consultant who worked for years at HMC under Meyer, Mendillo got the HMC gig partly as a result of Meyer's recommendation.

She had spent the last six years running the $1.6 billion Wellesley College endowment, which was completely outsourced to external managers. Her detractors say that she was ill prepared for Harvard's liquidity crisis and slow to take cognizance of the swap exposure. But they concede that the crisis came fast on the heels of her arrival.

Mendillo did move quickly to deal with the private equity portfolio. One of her first moves at HMC, which she initiated before the markets started to fall in earnest, was to sell between $1 billion to $1.5 billion of Harvard's private equity assets in one of the biggest such sales ever attempted.

The high bids on such assets have recently been 60 cents on the dollar, says Cogent Capital, an investment bank that advised Harvard on the sale. Cogent says the big discounts are due to "unrealistic pricing levels at which funds continued to hold their investments" and "fantasy valuations."

Defenders of Harvard's portfolio argue the secondary market is discounting private equity stakes too much. The market is made up of a dozen secondary funds with at most $15 billion available, says Bryon Sheets, a partner at San Francisco secondary firm Paul Capital.

That makes it a buyer's market, given the slew of desperate banks, pension funds and endowments looking to unload assets to meet obligations. So what are Harvard's private equity stakes worth? Most private equity investors like Harvard have been waiting for their money managers to finish marking down their assets following a brutal 2008. It is a slow process that lags the public markets by as much as 180 days, says William Frieske, a performance consultant at Northern Trust, which administers endowment accounts.

But one clue to what may be coming can be found in Harvard's own portfolio. It owns units of Conversus Capital, a publicly traded vehicle that holds slices of 210 private equity funds. Conversus has cut its net asset value by 21 per cent since last summer to make a "best estimate."

Yet stock investors think things are a lot worse. Conversus shares have fallen 67 per cent since June 30 and are trading at a 62 per cent discount to the net asset value. The Conversus stock drop translates into a potential $168 million loss for Harvard, which, as of Jan. 31, was still listed as a "strategic investor."

Conversus is run by Robert Long, a former Bank of America exec who went to Boston and got $250 million from El-Erian to help him set up the firm and buy $1.9 billion of Bank of America's private equity assets. Harvard also owns a piece of Garnett & Helfrich Capital, a $350 million fund opened in 2004.

Garnett has purchased six companies but, five years later, is yet to realize any returns. The value of one of those investments, software maker Ingres, has been reduced by its minority owner to nothing "as a result of reported losses." Then there is Tallwood Venture II, a $180 million fund raised in 2002 to invest in semiconductors. It has hardly exited any of its portfolio companies, according to Thomson Reuters and SEC filings.

The fact that a fifth of HMC's portfolio is in private-equity-like investments makes it vulnerable to the kind of problems HMC faced this fall. HMC has made $11 billion of capital commitments to investment partnerships through 2018, says Moody's. HMC used to make good on those commitments with income generated by the existing private equity portfolio. "Endowments are afraid capital calls will come quickly and far ahead of any liquidity from private equity funds," says Colin McGrady, managing director at Cogent Partners.

Watching all of this, the group of 10 Harvard alumni from the class of 1969 feel vindicated. "The events of the last year show that the whole procedure of rewarding people so handsomely based on increases on paper value of the endowment was deeply flawed," says a spokesman for the group, which recently sent a letter to the Harvard president suggesting HMC staffers return $21 million of their latest bonuses. "Even now, we don't really know how well it has done in the last 10 years."