If I understand correctly, a Securitized Debt Instrument (the evil engine of the economic meltdown) is similar to a stock mutual fund.  Mutual fund buys stock from companies A, B, and C.  I buy a portion of the fund.  The idea is that, while Company A may be down, B & C probably aren't.  That means I lose less money than if I had purchased stock from A, directly.  Buying into the fund, rather than buying stock directly, dilutes my losses.

An SDI then, is something like a mutual fund that buys I.O.U.s (mortgages, for example) rather than stock.  But it's the same idea, otherwise.  If Homeowner A defaults on a mortgage, B & C probably haven't.  I lose less money than if I had sold the mortgage to A, directly.  Buying into the SDI, rather than selling a mortgage directly to A, dilutes my losses.

That makes sense.  Clearly, I reduced my losses in both examples.  So what-the-hell-happened in the real world?  I have a couple of ideas:

• It only works if the risk of default is correctly assessed.  The idea is to mix loans with different risks.  Assume the guy who set-up the SDI chose A for high risk of default, B for moderate, and C for low risk of default.  If B is actually a high risk for default, then buying into the SDI doesn't dilute my risk.
  
• Diluting the risk makes more people eligible for loans; specifically, people who wouldn't have been eligible without the SDI.  This is (probably) good -- it allows people just outside the envelope of home-ownership to purchase a house, with the social benefits that entails.
• There are always more high-risk-loan opportunities than moderate & low-risk ones.  That creates pressure to reclassify moderate as safe, high as moderate, and "bad idea" as low-risk.  Picking-up new sales at the bad-idea/low-risk boundary increases the pressure still more.
  
• SDI-guy didn't sell the mortgages.  Banks did.  And the banks did it fully expecting to sell those loans to SDI-guy very quickly.  Banks, then, had reason to sell as many loans as possible, largely without caring about default.

So what happened?  Well, I Am Not A Finance-Guy but here's my take: SDIs (and other financial techniques) expanded the loan market by diluting the risk associated with making a loan.  That maybe created a bubble, or contributed to it.  In a bubble, everything's great -- right up until it isn't.  That makes the SDIs look safer than they are, and the "risk reclassification" seem legitimate ("That expired bologna-loaf hasn't made us sick so far, so it must still be good!").  The mix that SDI-guy called high/med/low was actually more like oh-God-NO/bad-idea/low.    Enough of those defaulted, that everybody called in his markers.  At which point even more loans defaulted, and banks started refusing to issue new loans (aka "credit") in order to cover their own markers (as they are legally required to do), and here we are now.

The thing is, in the beginning, SDIs worked really well.  It's just, as usual with us humans, we had to push it too far.