In recent work, FedFin has focused on what we think the new rules should be.  Forthcoming work on leverage and operational risk-based capital will continue our quest for capital rules that actually make finance safer.  But, realists that we are, FedFin also spends a lot of time mostly behind closed doors in our proprietary practice, figuring out who wins and loses based on the rules as they are and will be, not on what we might prefer.  Earlier this week, we outed a bit of this thinking, issuing a report on what the new rules mean for U.S. housing finance.  As Karl Marx would have hoped, capital determines the GSE’ destiny.  All of the talk of privatization, utilization, or whateverization is for naught because the capital rules dictate the walk.  U.S. mortgage finance has two choices: live with GSEs providing up-front guarantees or learn to live without penalty-free pre-payable thirty-year fixed-rate mortgages.

I know, you thought the U.S. had learned the hard way that GSE doesn’t just spell government-sponsored enterprise – it also means giving shareholders everything until the day the G goes back to meaning government and taxpayers find out that S stands for sucker.  But, while we learned during the crisis how much GSEs can cost us, the U.S. hasn’t kicked the thirty-year, fixed-rate mortgage (FRM) habit.  We still want more even though this loan structure is insuperable without a federal guarantee, especially now that regulators demand that banks actually hold capital against all this now-recognized risk.  Banks can’t hold thirty-year FRMs, let alone guarantee them under the new rules.  Non-banks might be able to do this, but only because they aren’t under all these rules and the market is desperate for yield-chasing action.  With real capital, there’s no thirty-year FRM, at least not at a price that sustains anything like the mortgage volumes needed to underpin U.S. housing and all its related, dependent, and very, very large economic sectors.  Without real capital, you either take a whole lot of systemic risk or find yourself back in the taxpayer’s lap.

I don’t like these choices much either, but I tried to kick magical thinking after my early twenties.  The reason why we need something like a GSE to have anything like thirty-year FRMS is that these loans have significant amounts not only of credit risk, but also of interest-rate, duration, liquidity, and several other risks that cost a lot of money to mitigate.  GSEs or the U.S. Government’s direct mortgage guarantor Ginnie Mae make these risks disappear not because they are smarter than anyone else, but because their U.S. government seals create so large and liquid a market that maturity and liquidity transformation becomes almost effortless.  It still isn’t – there’s all too much risk left for the taxpayer because pricing and other disciplines don’t apply, but at least a government backstop calls the bet for what it is.

Could private-label securities (PLS) take at least some of the government out of U.S. mortgage finance?  Hope has sprung eternal for this since 2008, but the actual amount of PLS has continued to end up somewhere between dismal and minimal.  If yield-chasing and current financial dementia doesn’t stoke PLS demand now, I see no way to develop a sustainable market in normalized rate conditions.  The market of course would love lots of high-yield PLS now, but it’s not getting what it wants because bank capital rules in concert with several others stymie mortgage securitization.  Whisk the GSEs out of the picture and there would be some more PLS, but nowhere near the volume or in structures sufficient to soak up current mortgage volume or, still more challenging, that in a robust recovery.

Keeping mortgage rates low and yet attracting enough private-sector capital to sustain a multi-trillion residential sector will happen just as soon as I get to each of three chocolate sundaes and then lose ten pounds.  Capital is destiny for U.S. mortgage finance because loans have risk just as sundaes have calories.