Why can’t the US reform its housing-finance sector?

Published by the R Street Institute.

The attached policy brief appeared in the Summer 2016 edition of Housing Finance International, the quarterly journal of the International Union for Housing Finance (IUHF).

It is sobering to Americans that the U.S. housing finance collapsed twice in three decades: in the 1980s and again in the 2000s. This is certainly an embarrassing record.

The giant American housing finance sector, with $10 trillion in mortgage loans, is as important politically as it is financially. Many interest groups want to receive government subsidies through the housing finance system. This makes it very hard to reform.

From the 1980s to today, U.S. housing finance has been unique in the world for its overreliance on the so-called “government-sponsored enterprises,” Fannie Mae and Freddie Mac. Fannie and Freddie get government guarantees for free, which are said to be only “implicit,” but are utterly real. According to Fannie and Freddie in their former days of power and glory, this made American housing finance “the envy of the world.” In fact, the rest of the world did not feel such envy. But Fannie and Freddie did attract investment from the rest of the world, which correctly saw their issues as U.S. government credit with a higher yield. In the 2000s, this channeled the savings of thrifty Chinese and others into helping inflate American house prices into their amazing bubble. Fannie and Freddie became a key point of concentrated systemic vulnerability.

In 2008, Fannie and Freddie went broke. What schadenfreude my German housing finance colleagues enjoyed after years of being lectured on the superiority of the American system!  Official bodies in the rest of the world pressured the U.S. Treasury to protect their investments in the obligations of the insolvent Fannie and Freddie, which the Treasury did and continues to do. The Federal Reserve, in the meantime, has become the world’s biggest investor in Fannie and Freddie securities.

Almost eight years after the financial collapse, America is still unique in the world for centering its housing finance sector on Fannie and Freddie, even though they have equity capital that rounds to zero. They are primarily government-owned and entirely government-controlled housing finance operations, completely dependent on the taxpayers. Nobody likes this situation, but it has outlasted numerous reform efforts.

Is there a way out of this statist scheme—can we move American housing finance toward something more like a market?  Is there a way to reduce the distortions caused by Fannie and Freddie, to control the hyper-leverage that inflates house prices and the excessive credit that sets up both borrowers and lenders for failure?  Can we reduce the chance of repeating the mistakes of 1980 to 2007?  Here are some ideas.

Restructure Fannie and Freddie

The original government bailout of Fannie and Freddie created senior preferred stock with a 10 percent dividend, which the U.S. Treasury bought on behalf of the taxpayers. This was later amended to make the dividend be all their net profit. That meant there would never be any reduction of the principal, and the GSEs will be permanent wards of the state.

It is easy, however, to calculate the cash-on-cash internal rate of return (IRR) to the Treasury on its $189.5 billion investment in senior preferred stock, given the dividend payments so far of $245 billion. This represents a return of about 7 percent – positive, but short of the required 10 percent. As Fannie and Freddie keep sending cash to the Treasury, the IRR will rise, and will reach a point when total cash paid is equivalent to a 10 percent compound return, plus repayment of the entire principal. That is what I call the “10 Percent Moment.” It provides a uniquely logical point for reform, and it is not far off, perhaps in early 2018.

At the 10 Percent Moment, whenever it arrives, Congress should declare the senior preferred stock fully repaid and retired, as in financial substance, it will have been. Simultaneously, Congress should formally designate Fannie and Freddie as systemically important financial institutions (SIFIs). They are unquestionably SIFIs – indeed, they are Global SIFIs – able to put not only the entire financial system, but also the finances of the U.S. government at risk. This is beyond the slightest doubt.

As soon as Fannie and Freddie are designated officially – as well as in economic fact – as SIFIs, they will get the same minimum equity capital requirement as bank SIFIs: 5 percent of total assets. At their current size, this would require about $250 billion in equity. They must, of course, be regulated as undercapitalized until they aren’t. Among other things, this means no dividends on any class of stock until the capital requirement is met.

As SIFIs, Fannie and Freddie will and should get the Federal Reserve as their systemic risk regulator, in addition to their housing finance regulator.

It is impossible to take away Fannie and Freddie’s too-big-to-fail status, no matter what any government official says or does. Therefore they should pay the government for its ongoing credit guaranty, on the same basis as banks have to pay for deposit insurance. I recommend a fee of 0.15 percent of total liabilities per year.

Then Fannie and Freddie will be able to compete in mortgage finance on a level basis with other SIFIs, and swim or sink according to their competence.

Promote skin in the game for mortgage originators

A universally recognized lesson from the American housing bubble was the need for more “skin in the game” of credit risk by those involved in mortgage securitization. Lost in the discussion is the optimal point at which to apply credit risk skin in the game. This optimal point is originator of mortgage loans, which should have a junior credit risk position for the life of the loan. The entity making the original mortgage is in the best position to know the most about the borrower and the credit risk of the borrower. It is the most important point at which to align incentives for creating sound credits.

The Mortgage Partnership Finance (MPF) program of the Federal Home Loan Banks was and is based on this principle (I had the pleasure of leading the creation of this program). It finances interest-rate risk in the bond market but keeps the junior credit risk with the original lender. The result was excellent credit performance of the MPF mortgage loans, including through the 2000s crisis.

I believe this credit-risk principle is obvious to most of the world. Why not to the United States?

Create countercyclical LTVs

As the famous investor Benjamin Graham pointed out long ago, price and value are not the same:  “Price is what you pay, and value is what you get.” Likewise, in mortgage finance, the price of the house being financed is not the same as its value: in bubbles, prices greatly exceed the sustainable value of houses. Whenever house prices are in a boom, the ratio of the loan to the sound lendable value becomes something much bigger than the ratio of the loan to the inflated current price.

As the price of any asset (including houses) goes rapidly higher and further over its trend line, the riskiness of the future price behavior becomes greater—the probability that the price will fall continues to increase. Just when lenders and borrowers feel most confident because of high collateral “values” (which really are prices), their danger is in fact growing. Just when they are most tempted to lend and borrow more against the price of the asset, they should be lending and borrowing less.

A countercyclical LTV (loan-to-value ratio) regime reduces the maximum loan size relative to current prices, in order to keep the maximum ratio of loan size to underlying lendable value more stable. The boom would thus induce smaller LTVs, and greater down payments, in bubbly markets—thus providing a financial stabilizer and an automatic dampening of price inflation.

Countercyclical capital requirements for financial institutions reduce the leverage of those lending against riskier prices. The same logic applies to reducing the leverage of those who are borrowing against risky prices. We should do both.

Canada provides an interesting example of where countercyclical LTVs have actually been used; Germany uses sustainable lendable value as the same basic idea. The United States needs to import this approach.

Liquidate the Fed’s mortgage portfolio

What is the Federal Reserve doing holding $1.7 trillion of mortgage-backed securities (MBS)? The authors of the Federal Reserve Act and generations of Fed chairmen since would have found that impossible even to imagine. This massive MBS portfolio means the Fed allocates credit to housing through its own balance sheet. Its goal was to push up house prices, as part of its general scheme to create “wealth effects.” It succeeded— house prices have not only risen rapidly, but are back over their trend line on a national average basis. This means, by definition, that the Fed also has made houses less affordable for new buyers.

Why in 2016 is the Fed still holding all these mortgages?  For one thing, it doesn’t want to recognize losses when selling its vastly outsized position would drive the market against it. Some economists argue that losses of many times your capital do not matter if you are a fiat currency central bank. Perhaps or perhaps not, but they would be embarrassing and cut off the profits the Fed sends the Treasury to reduce the deficit.

Whatever justification there might have been in the wake of the collapsed housing bubble, the Fed should now get out of the business of manipulating the mortgage-securities market. If it is unwilling to sell, it can simply let its mortgage portfolio run off to zero over time through maturities and prepayments. It should do so, and cease acting as the world’s biggest savings and loan.

The collapses of the 1980s and 2000s should have taught the American government a lesson about the effects of subsidized, overleveraged mortgage markets. It didn’t. The reform of the Fannie and Freddie-centric U.S. housing finance sector has not arrived, nor is there any sign of its approach. But we need to keep working on it.