In the last few days, a furor of surprising – and not accidental – energy has arisen over non-bank mortgage servicers. The CFPB is threatening bloody murder, Rep. Waters wants servicing to stay stuck in banks, New York’s regulator has launched an all-points investigation of Ocwen, and banks are blaming the new-found market clout of these nonbanks on a raft of rules they think go far too easy on the competition. Some of these accusations are right, some not so much. All of them, though, point to the rapid transformation of mortgage finance into a shadow sector free from most of the rules regulators hoped safeguard borrowers and financial markets. Is this good for homeownership? Surely not. Did regulators anticipate this shift to the shadows in any of their rules? Again, no way. It will take a while for the balance between bank and non-bank mortgage players to settle, but the huge shift in servicing to non-banks in the past two or three years shows clearly which way the business is heading. There are several drivers here, the first of which was the sharply-increased capital treatment for mortgage servicing rights (MSRs) in the Basel rules. On its own, banks could have weathered this and stayed in servicing, but legal and reputational risks have made servicing so costly a business that most want out once the lawyers have finished with them. To be sure, non-banks can be sued too, but it’s nowhere near as much fun since they’ve nowhere near as much money. For starters, rules don’t require them to hold legal reserves and to test them for adequacy under stressed conditions.

Another big driver of mortgage finance to the shadows comes from the CFPB’s new servicing rules. These only became effective in January, so few servicers have yet really reckoned with them. But, they dramatically change the structure of mortgage credit risk. Servicers pretty much just do one thing – take money from borrowers and send it to all the places it’s supposed to go – taxes, insurance, and RMBS investors among them. The new rules say that, if a borrower misses a payment, the servicer is to work with him or her to modify the loan in some way or help the borrower come current. Each time a deal is broken, the help-them-out process starts anew, driving the time between first delinquency to default from months to years. All this time, though, the servicer is on the hook to pay investors – cash out the door that has to come from somewhere if it doesn’t come from the ne’er-do-well borrower.

For bank servicers, this money comes from capital and reserves, both of which now have to be a lot higher. For non-bank servicers, the money comes from a promise to investors that some of it will be found when needed. Over time, investors may well grow wary, but for now this means that non-bank servicing is a lot more profitable than bank servicing because it’s way, way more leveraged and, for that matter, exempt from all the tough new liquidity rules we assessed earlier this week for clients.

In the years before the financial crisis, non-banks were formidable mortgage players – think Countrywide before BofA thought it got lucky, Merrill Lynch – ditto BofA, New Century, Fremont, and the rest. Big S&Ls – WaMU, Indy Mac – also had fun with mortgages because capital and other rules largely didn’t apply to them. With banks stuck back in the regulatory box, non-banks are showing yet again how profitable mortgages can be if regulators aren’t involved.

Can this business model last? Not if recent history is any guide – as I think it is. Maybe this time the non-banks won’t grow so big because secondary markets will be smarter. But, so far, some secondary market players – FHA – aren’t showing much smarts because they’re still securitizing lots of high-risk loans dependent on vigilant servicing practices and a good-sized taxpayer backstop. Fannie and Freddie are keeping their skirts clean so far, but what of the replacement model for them or new profit pressures on the GSEs if left to languish in conservatorship for years to come?

This shift to non-bank mortgage finance is still way, way dangerous even if it doesn’t prove systemic all over again. Taking risks without holding capital is a recipe for disaster. I thought we learned that in 2008, but maybe not.