What can we say? Our summer reading is, “The Role of the Securitization Process in the Expansion of Subprime Credit.” A new Fed research paper by two senior economists. One does so hate to get sun-block on this, but we soldiered on because the paper proves a critical point: regulatory capital matters and, in fact, it was a precipitating cause of the financial crisis. Given that bank regulators are about to rewrite the U.S. capital rules, the research is not only an important retrospective, but also valuable insight into what not to do this time around.

Point one: don’t trust the credit ratings agencies (CRAs). We’ve said this before, but it bears repeating especially because the Administration ratings-agency reform legislation – like the SEC – ducks the issue. Congress told the SEC in 2005 to look hard at the role of ratings in bank and securities regulation. The Commission in fact issued a very tough proposal but then completely backed away from it. The Administration just suggests yet another study, sending the question back to the beginning four years ago.

A look at the Fed paper points to the urgency in fact of getting around to doing something about the CRAs. It proves that the CRAs gave the most favorable AAAs to MBS pooled from loans in markets with the highest rates of house-price appreciation. A moment of reflection might have led to the obvious conclusion that what goes up not only comes down, but also that the faster prices go up, the harder they fall. We have yet to see sustained self-levitation in any financial market, but the CRAs “stress-tested” their assumptions over a very short time horizon so they never saw the downturn coming.

Maybe the CRAs are chastened now. And, even if they aren’t, the Administration bill does include some very useful methodology and disclosure reforms. But, unless or until CRAs try harder and stress cycles prove they do, regulatory decisions should use CRAs as nothing more than benchmarks and, even then, do so warily.

Point two: given an inch, astute firms will arbitrage the living daylights out of flaws in their risk-based capital. The Fed paper takes a look at the five bulge-bracket investment banks – remember them? – allowed in 2004 by the SEC to become “consolidated supervised entities” (CSEs). In our last look at Basel II, we noted that the CSEs were the only U.S. firms allowed under Basel II. When it did so, the SEC granted them an even bigger gift by failing at the same time to do anything to enhance prudential supervision or disclosure. This gave the CSEs open field, and the Fed paper shows the alacrity with which they took it and the disastrous results that then ensued.

Under Basel II, the internal ratings-based (IRB) approach permitted virtually no risk-based capital for AAA-rated mortgage obligations. With the CRAs doing the voo-doo they did over subprime and Alt-A MBS, the supply of AAA-rated MBS skyrocketed from 2004 on in parallel with house-price appreciation in the most speculative markets. Seduced by what was called “vertical integration,” the CSEs not only got big into MBS, but also into mortgage origination. We well remember when Bear Stearns opened up its mortgage bank without any capacity to handle foreclosures because, as it subsequently said, the firm didn’t think mortgages ever went into default. Oops.

Prudential regulation from the SEC might have spotted this glaring flaw, but in its absence all of the bulge-bracket firms followed Bear Stearns’ lead and went big into the full range of mortgage origination and securitization operations, betting bigger and bigger as the market boomed ever more loudly. The CSE risk-based capital rules did have one effect: it made it costly for the CSEs to hold the lower-rated tranches of their junky MBS. Since the market was replete with investors without even this minimal capital discipline, though, that proved no impediment. The CSEs get the AAA-rated tranches and sold the rest, contributing to the now all-too-familiar systemic risk.

Of course, the global financial system isn’t the only victim of this capital catastrophe. All the homeowners pushed into mortgages they couldn’t afford in markets with unsustainable house-price appreciation are also victims, as are the neighborhoods in which they live and, now, the taxpayers trying to pick up the foreclosure mess’ pieces. Do seemingly technical details buried in the fine print of complex rules matter? As the Fed’s new paper makes clear, you bet they do.