“The Fed Can Exit Housing, or Keep Mortgages Cheap. Not Both”

Newly installed as Fed chair, Kevin Warsh wants the Federal Reserve out of the mortgage business. The organization still holds almost two trillion dollars in mortgage-backed securities, the residue of two rounds of crisis buying and something the law allows but never required it to hold. A central bank that keeps housing as its favored sector, whatever the crisis-era logic for getting there, is running something closer to industrial policy than monetary policy.

The case for getting out is clean. The case for getting out cheaply is not.

A reform everyone claims to want

The Fed itself says it intends to hold "primarily Treasury securities" over time, precisely to minimize its effect on the allocation of credit across sectors. The free-market critics want the same thing, only faster. Cato's Norbert Michel has argued for years that the Fed should trade only short-term Treasuries and shed its mortgage book, and he is right that the holdings tilt the field toward housing. Warsh looks ready to go further than his predecessor and shrink the book faster than waiting for the loans to mature.

On the destination, most of the market-minded agree: among free-market economists and many former officials, the Fed does not belong in housing for good. Some do defend the holdings -- as a prop for housing, or as a spare lever for when interest rates hit zero -- and that is a fair argument about the destination, not the cost. But few of them would call the current size a permanent fixture. For those who want the Fed out, the question is how to leave, not whether.

The exit nobody wants to price

What too few of them say out loud is what leaving costs. For much of the past fifteen years the Fed has been a buyer far less sensitive to price than anyone else in the market, willing to soak up prepayment risk at yields below what private investors require. Remove that buyer and the risk does not disappear. It returns to investors who price it, and the gap between mortgage rates and Treasury yields widens.

How much? No one can say with precision, because the Fed has never sold a mortgage book anywhere near this size. That spread answers to many things, from hedging costs to bank balance sheets to the wiring of the mortgage market, and the Fed is only one of them. But it is the largest price-insensitive buyer in that market. Its years of buying helped pull the spread in; take that support away, and the pressure runs the other way.

Where the headline rate finally lands depends on where Treasury yields go too, so no one should promise a number. What no one should pretend is that borrowers come through it untouched. With the thirty-year fixed already around six and a half percent, and the spread already near its widest in years, even a modest widening is no rounding error to a family at the closing table.

This is not the Fed setting mortgage rates. The market sets those, as this page often argues. It is the Fed lifting a thumb it had been pressing on the scale, and letting the market reprice what it had been leaning on. That repricing is honest. It is also not free.

Slow is not the same as neutral

So the Fed has chosen the slow door. Since last December it has stopped shrinking its balance sheet altogether; the principal that trickles back from its mortgages now goes into Treasury bills, and the rest sits. Because almost none of those old, low-rate mortgages are being refinanced at today's rates, the mortgage book barely moves. On the Fed's own staff projections, the book will still hold some seven hundred billion dollars in mortgages a decade from now, and will not clear for years after that.

There is a respectable reason to go slow. A sudden sale could jolt the mortgage market, and sparing borrowers that kind of shock is a fair goal. But slow is still a choice, and it is a choice to stay. Holdings that large keep tilting credit toward housing as they sit there, the effect fading only gradually, so the distortion the Fed says it wants gone lives on for another generation.

A generation is not an exit. It is a way of looking like you left.

The slow runoff lets the Fed claim a neutrality it has not paid for. It gets to say it is leaving housing while still holding almost two trillion dollars of it, and while putting off the day borrowers feel the full cost. That is not patience. It is a decision to stay.

Who pays, and why that is the honest part

A real exit would be felt, and it would be felt by borrowers -- wider mortgage spreads, at least for a time. Mortgage analysts say so plainly. The reform case rarely does. The same writers who carefully document that the Fed's buying pushed mortgage rates down rarely dwell on the other side of it: that its leaving pushes them back up.

The answer is to do it anyway. A mortgage rate that reflects the real cost of lending for thirty years is a true price, not an injury. A rate held down by a central bank parked in the market is a subsidy, one the Fed was never asked to grant. Ending the subsidy is the point, not an embarrassment to be buried.

And this is not what truly ails housing. Its deepest troubles, from too few homes to prices out of reach to owners frozen in place by mortgages they will not give up, are largely structural and national, problems the Fed's mortgage book did not cause and is not the tool to fix. That cost of leaving is real to the family who pays it, but set against problems that large it is not what decides whether the country can house itself. A central bank is not housing policy, and as long as the Fed sits in the market the true price of a mortgage stays muddied.

Say the price out loud

Warsh has a cleaner choice than the one he is being handed. He should commit to a dated path out of mortgages and tell the country it will cost a little at the margin. Going slow to spare the market a shock is defensible; going slow while pretending the Fed has already left is not. That pretense is the one option that should be off the table.

You cannot leave a market you refuse to price.

The Fed can leave the mortgage business, or it can keep mortgages cheap.

Not both.