The basic idea is that different people have different risk tolerances. Therefore, if some financial assets (say, bank loans) can be bundled and bits of it sold to buyers with varying appetite for risk, it will free the primary lender (the bank) from the obligation of carrying all the risk. So risk will be borne more efficiently and the bank’s ability to lend will increase.

Don’t sneer at this just because jokes about investment bankers outnumber jokes about lawyers today. Distributing risks is at the heart of such plain vanilla financial products like insurance or company shares. Issuing shares allows a company’s business risk to be spread widely between shareholders.

But this distributed risk idea works really well for markets if all market participants have equal access to all relevant information. Remove that assumption and there’s a problem.

When risk is distributed widely, as it was when assets of various kinds were bundled and bits sold to many financial firms, ideally everyone involved should have known if, say, particularly risky loans were bundled with relatively safe assets. But one of the things that happened in the financial crisis was that sellers of risky mortgages had better information than buyers of bundled assets. There was information asymmetry.

So when risky assets turned bad, holders of bundled assets, who didn’t have all the information, acted on the basis that the whole thing was risky. Since there were many buyers of these bundles — risk was widely distributed — the market for efficiently managing risk ended up increasing market participants’ perception of risk. Basically, the model failed.

... contd.