Page 286 - untitled

LATE  TO EARLY : BILLIONS IN SUBPRIME LOSSES 

MERRILL LYNCH: “DAWNING AWARENESS
OVER THE COURSE OF THE SUMMER”
On October , Merrill Lynch stunned investors when it announced that third-
quarter earnings would include a . billion loss on CDOs and  billion on sub-
prime mortgages—. billion in total, the largest Wall Street write-down to that
point, and nearly twice the . billion loss that the company had warned investors to
expect just three weeks earlier. Six days later, the embattled CEO Stanley O’Neal, a
-year Merrill veteran, resigned.
Much of this write-down came from the firm’s holdings of the super-senior
tranches of mortgage-related CDOs that Merrill had previously thought to be ex-
tremely safe. As late as fall , its management had been “bullish on growth” and

“bullish on [the subprime] asset class.” But later that year, the signs of trouble were
becoming difficult even for Merrill to ignore. Two mortgage originators to which the
firm had extended credit lines failed: Ownit, in which Merrill also had a small equity
stake, and Mortgage Lenders Network. Merrill seized the collateral backing those
loans: . billion from Mortgage Lenders, . billion from Ownit.
Merrill, like many of its competitors, started to ramp up its sales efforts, packag-
ing its inventory of mortgage loans and securities into CDOs with new vigor. Its goal
was to reduce the firm’s risk by getting those loans and securities off its balance sheet.
Yet it found that it could not sell the super-senior tranches of those CDOs at accept-
able prices; it therefore had to “take down senior tranches into inventory in order to
execute deals” —leading to the accumulation of tens of billions of dollars of those

tranches on Merrill’s books. Dow Kim, then the co-president of Merrill’s investment
banking segment, told FCIC staff that the buildup of the retained super-senior
tranches in the CDO positions was actually part of a strategy begun in late  to
reduce the firm’s inventory of subprime and Alt-A mortgages. Sell the lower-rated
CDO tranches, retain the super-senior tranches: those had been his instructions to
his managers at the end of , Kim recalled. He believed that this strategy would
reduce overall credit risk. After all, the super-senior tranches were theoretically the

safest pieces of those investments. To some degree, however, the strategy was invol-
untary: his people were having trouble selling these investments, and some were even
sold at a loss. 
Initially, the strategy seemed to work. By May, the amount of mortgage loans and
securities to be packaged into CDOs had declined to . billion from . billion

in March. According to a September  internal Merrill presentation, the net
amount in retained super-senior CDO tranches had increased from . billion in
September  to . billion by March  and . billion by May. But as the

mortgage market came under increasing pressure and as the market value of even su-
per-senior tranches crumbled, the strategy would come back to haunt the firm.
Merrill’s first-quarter earnings for —net revenues of . billion—were its
second-highest quarterly results ever, including a record for the Fixed Income, Cur-
rencies and Commodities business, which housed the retained CDO positions. These