No rest for the weary...

...Fannie Mae investigator. Just when you thought it couldn't get any worse for the government-sponsored housing enterprise, we discover that regulators have stumbled across even more potential accounting irregularities.

Some of the emerging details on this story suggest that it could get really rocky (and the Street is taking the hint; Fannie's price is down around 5 percent just today). It appears that Fannie created several of a certain type of "subsidiary" business entity to issue and sell the mortgage-backed securities that form the core of the GSE's business model. It could then use a particular section of the accounting rules to define these "subsidiaries" in such a way that their assets and liabilities would be kept off Fannie's books. This raises the specter of a potentially enormous problem:

Fannie's core business is (supposed to be) buying mortgages from banks and then bundling those mortgages into securities (mortgage-backed securities, or MBSs) that can be traded on the market. This is a risky sort of asset, however -- it disappears as soon as the underlying mortgage is paid off. And a unique feature of the American mortgage market is that homeowners can pre-pay the principal with no penalty. (This is most likely to be a big problem in periods when interest rates are low, leading many homeowners to refinance.) So Fannie, or the Fannie "subsidiary" issuing the MBS, also sells a form of "insurance policy" to the MBS purchasers, guaranteeing a certain minimum return on the investment.

What that means is that Fannie, through the operation of these trusts, was prone to significant interest-rate risk. How much risk? We'll have to see how that was accounted for (or even if it was accounted for in one place). There are ways to create sophisticated portfolios of derivatives to hedge against these risks (in fact, the structure of other hedge portfolios led to an earlier round of embarrassing accounting revelations for Fannie).

Which leads to this list of some things to look for in the news reports in coming days:

First, how effectively were these units hedged? Strikes me as a reasonable question, since, if Fannie was so eager to keep these trusts off the books, one wonders how aggressive they were about going through the trouble to hedge.

Second, how were those hedge positions accounted for? A very interesting question, given Fannie's apparent predilection for sloppy hedge accounting.

All of which matters because it is precisely this interest rate risk that would tip Fannie (or Freddie, for that matter) over the edge into collapse.

My Cranky Prediction: We haven't seen nearly the end of this story yet.