When PE funds do these kinds of deals, they "agree to pay" $8 billion to buy the company (as the headline says). What actually happens is that PE funds come up $2-3bn themselves to buy the equity, and also come up with a group of people willing to put up another $5-6bn to buy bonds against the cash flows generated by the company. Usually, something like 50-75% of the takeover is funded with debt. The PE fund does not take on the debt, the company they are taking over takes on the debt (in this case, Harman). If Harman, separately, wanted to sell a $200mil bond in normal times for the credit market, it would not have a great deal of trouble (other than it's a small deal) and the spread (percent that they would pay over a benchmark - say government debt) would not be terribly wide. However, when they put themselves in debt to the extent that the PE funds like to do (let's say 10 times the amount of earnings generated before interest, tax, and depreciation are subtracted), the company has become less credit-worthy and therefore the people lending the money want to get a slightly higher interest rate for the increased risk. These are "junk bonds." Usually, PE funds get guaranteed financing and then make the deal. The 'guaranteed financing' includes a bridge loan (and sometimes more complicated things) to pay for the thing in cash, then a refinancing package into bonds, whereby the investment bank guarantees that the borrower will pay a certain interest rate. Investment banks (like Goldman Sachs) are eager to act as arrangers for the loans/debt because the fees are quite good on leveraged loans and junk bonds, and they tend to think it gets them an inside track to act as investment bank when the PE fund tries to sell the asset on (either in an IPO or to another buyer), which will of course generate more fees.
The recent credit crises sweeping financial markets has meant that all kinds of credit have seen spreads widen vs Treasuries - even to the extent that banks lending to each other in the money market, and highly credit-worthy companies borrowing extremely short-term money in the commercial paper market have found it difficult to borrow at anywhere near normal rates (or at all). This has meant that junk bond investors are no longer willing to buy 5yr bonds with a coupon of 1.5% over LIBOR and instead want to receive 3.5% over LIBOR. The investment bank who guaranteed to the PE fund that they would pay 1.5% over LIBOR signed a contract to that effect and the PE fund will generally want to hold them to it. If the PE fund does hold the investment bank to the deal, the investment bank will only be able to find buyers for its bonds at the 3.5% over LIBOR level, which in this case would be something like 90 cents on the dollar.
In this case, the math is that if GS is the sole arranger for the bonds (which could be $5-6bn) and they lose 10% on those bonds (they actually would only lose abt 8% if they sold at 90 because they keep the 2% fee), they would lose $400-500 million at this point. Davidny's suggestion is that it is possible that GS' private equity arm might not want to stick it to GS' investment bankers and so have claimed MAC (if GS and KKR share a $225mil break up fee, GS' share would be substantially less than the $400-500mil loss on the bonds).
The deal contract between Harman and the buyers precludes missed forecasts, a slowdown in the audio industry, or the economy at large as being a material adverse change (MAC). The financing has been stuck to the investment bank already so that is not (usually) a worry for the buyers. If the change really has been a MAC, theoretically, the buyers could walk away without paying the break-up fee though as Davidny suggests, Harman's initial reaction to say that they dispute the buyers' claim that there has been a MAC means that Harman is going to fight the buyers who suddenly have cold feet (the fall in the stock price is a weird one - it causes damage but the "damage" has to be claimed through tort (not living up to the contract), not through simple loss, which means it is really up to Harman's board to act on behalf of shareholders - though they will because they are afraid of shareholders' claims). All things considered though, it is likely that the vast majority of shares at this point were held by merger arbitrage desks trying to earn a spread too - though given what happened in August, I would have avoided plain-Jane PE-fund-buyer balance sheet merger arb trades in the US like the plague.
I also agree with Davidny's last comment - it is really difficult to imagine a deal like this (a leveraged "recapitalization" which usually have lots of covenants attached to the debt so that cash flow must be maintained at high levels) where product innovation and service "costs" are going to be well-supported. Though, one thing to say for KKR (disclaimer: no affiliation blah blah blah), on average they hold their companies for 7+yrs (vs average PE fund holding period of roughly 3-4yrs) so they do take a long-ish view (though it helps that they have a huge portfolio of companies, cross-sell products wherever possible (see the Sun deal) and all their portfolio companies pay consulting fees to their captive consulting company).
