Three reasons why Im not jumping at the investment banks at present:

• There are still significant areas of concern that have
  not unwound yet residential housing exposure will increase as housing
  prices fall further, including lawsuits which will eventually prove not
  meritorious, and CDO exposure.
• It is my firm belief that their
  hedges hold in minor moves, but not major moves. VAR modeling is fine
  for when the winds are calm, but not when they are gale force. At gale
  force the Extreme Value Theory models kick in, and they are untested at
  present.

  Berkshire Hathaways experience in unwinding GenRes swap
  book was telling; few things were marked conservatively. That is
  probably true industrywide, partly because auditors are incapable of
  audit the swap books in all of their complexity, or theyd be working
  for the investment banks themselves.

• New accounting regulations
  make earnings quality more opaque, and less comparable across time
  periods and companies. This should result in lower multiples, akin to
  big commercial insurers.

Thats all. Personally I think the investment banks will be a
buy sometime in 2008, but I am waiting to see how the current leverage
unwind affects them.

Originaly from Source