Jerome Powell has some curious ideas about housing finance

American housing finance has changed a lot since the bubble burst — for the better.

Yet some officials now wax nostalgic for the pre-crisis model. In January, William Dudley of the Federal Reserve Bank of New York seemed disappointed households weren’t using their homes as ATMs as much as in the go-go years. Federal Reserve governor Jerome Powell has gone even further in a recent speech to the right-leaning American Enterprise Institute.

The title of Powell’s speech was “The Case for Housing Finance Reform”. It could have been about the tax deductibility of interest payments, the troubles with mortgage servicers, or derivatives to hedge against changes in house prices.

Instead, Powell focused on the status of Fannie Mae and Freddie Mac, the “government-sponsored enterprises” that have been wards of the state since 2008. According to him, the current situation “may feel comfortable, but it is also unsustainable”. The status quo is “unacceptable”. He darkly warned that “the next few years may present our last best chance” to “address the ultimate status of Fannie Mae and Freddie Mac” and avoid “repeating the mistakes of the past”.

Why is he so hot and bothered?

After all, America’s housing finance system is currently a stable closed system mostly involving transactions within the government. A brief summary of how things work today:

Banks and others originate mortgages that conform to strict standards so that they’re eligible to be purchased by the GSEs. The GSEs buy the loans, bundle them into agency mortgage-backed securities, and insure them against the risk of default in exchange for “guarantee fees” paid ultimately by borrowers.

Combining the magic of securitisation with the depth of the To Be Announced market for agency MBS transforms millions of individual loans into a few dozen liquid contracts. Traders, freed from having to worry about defaults, focus their energies on pricing the odds of prepayment risk. (That’s what gives agency MBS their negative convexity.)

When the GSEs collect more in insurance premiums than they pay out to bond investors — one of the biggest being the Federal Reserve — they send the difference to the Treasury. The Fed, which owns more than a fourth of the total supply of these agency mortgage-backed securities, also sends its profits to the Treasury on a regular basis.

The result is that taxpayers, who are always on the hook when things go really wrong, get to capture a fair amount of the upside. You might dislike the arrangement for ideological reasons, but it seems like a reasonably fair deal for the government as long as g-fees are high enough. (Some hedge funds argue taxpayers are being treated too well, but that’s outside the scope of this post.)

So what’s not to like?

It’s reasonable to wonder if the government’s focus on “affordable” mortgages encourages excessively high prices and an inefficient distribution of home ownership. Without its heavy state subsidies for debt, America might look more like Germany or Switzerland, where the majority of the population rents housing.

It’s even possible to imagine large swathes of America’s housing stock owned and professionally managed by listed companies or real estate investment trusts, allowing regular people to get exposure to residential real estate as a diversified asset class while maintaining the flexibility that comes from renting.

None of this seems to be what Powell has in mind, however. Instead, he has two main concerns:

Taxpayers are on the hook for mortgage defaults
Lending standards are “too rigid”, with riskier borrowers harmed by “a lack of innovation and product choice”

Powell thinks both can be addressed by adding “ample amounts of private capital to support housing finance activities”. Mortgage credit risk isn’t inherently different from other kinds of credit risk. If investors can be trusted to lend responsibly to businesses and municipalities — and bear the pain when they lend poorly — why can’t they do the same for loans backed by residential real estate? Privatisation could also encourage competition, potentially lowering costs for borrowers and promoting “innovation and product choice”.

This sounds reasonable, but it’s not as if private creditors are currently prevented from making mortgages right now.

There are no laws or regulations telling banks not to hold mortgages on their books directly. Even if you think risk-weighted capital requirements and liquidity rules discourage banks from holding mortgages, Powell himself doesn’t seem to think those are a problem. Instead, he holds those regulations up as models for what ought to apply to privatised mortgage securitiser-insurers. (There’s also the lingering question of whether those rules have actually been sufficient to eliminate wealth transfers from taxpayers to bankers and bank shareholders…)

“Private-label” securitisations aren’t really limited by any of the post-crisis reforms, either. Issuers can stuff whatever they want into their bonds as long they’re willing to retain 5 per cent of the default risk. And if they’re unwilling to eat their cooking, they can avoid retaining any risk as long as the mortgages in the pool meet certain minimum standards. Things have changed, but not nearly enough to explain the dearth of new issuance. It’s far from obvious how ending the conservatorship of the GSEs would change any of this.

Part of the confusion might come from Powell’s misleading narrative of the origins of the crisis.

The story he told in his speech focused almost exclusively on the growth and transformation of Fannie and Freddie, with barely a mention of other causes:

Beginning in the early 1980s, Fannie and Freddie helped to facilitate the development of the securitization market for home mortgages. They purchased and bundled mortgage loans, and sold the resulting mortgage-backed securities (MBS) to investors. Fannie and Freddie also guaranteed payment of principal and interest on the MBS.
With this guarantee in place, MBS investors took the risk of changing interest rates, and the GSEs took the risk of default on the underlying mortgages…The system ultimately failed due to fundamental flaws in its structure. In the early days of securitization, the chance that either GSE would ever fail to honor its guarantee seemed remote.
But the question always loomed in the background: Who would bear the credit risk if a GSE became insolvent and could not perform?…Investors understandably came to believe that the two GSEs were too-big-to-fail, and priced in an implicit federal government guarantee behind GSE obligations.
In the end the investors were right, of course. The implicit government guarantee also meant that investors–including banks, the GSEs themselves, and investors around the world–did not do careful due diligence on the underlying mortgage pools. Thus, securitization also enabled declining lending standards. This was not only a problem of the GSEs–private label securitizations also helped to enable lower underwriting standards.
Over time, the system’s bad incentives caused the two GSEs to change their behavior and take on ever greater risks. The GSEs became powerful advocates for their own bottom lines, providing substantial financial support for political candidates who supported the GSE agenda.
Legislative reforms in the 1990s and the public/private structure led managements to expand the GSEs’ balance sheets to enormous size, underpinned by wafer-thin slivers of capital, driving high shareholder returns and very high compensation for management.
These factors and others eventually led to extremely lax lending conditions. The early 2000s became the era of Alt-A, low doc, and no doc loans. These practices contributed to the catastrophic failure of the housing finance system.

That’s factually correct, but profoundly misleading.

Yes, Fannie and Freddie were mismanaged in the period before the crisis. Thanks to the perception of taxpayer guarantees — in turn facilitated by favourable regulation bought with generous donations to politicians — they operated with excessive leverage, extracting wealth from regular people to reward shareholders and executives. Also, in 2007, the GSEs lowered their standards to regain market share.

But it was their 50-1 gearing that compelled the government to pump in so much money to cover their losses. Default rates on Fannie’s and Freddie’s mortgages were quite low, even on their subprime MBS, especially when compared to everyone else who invested in American mortgages. Blaming Fannie and Freddie for the excesses of the 2000s is nonsensical. They may have exacerbated things on the margins at the end, but they weren’t the main event.

The real explantion for the bubble and bust is the growth and collapse of the private-label market.

From the start of 1991 through the end of 2003, a little less than 14 per cent of the growth in residential mortgage debt was funded by the expansion of the private-label MBS market. In the go-go years of 2004-2006, however, half of the growth in mortgage debt came from this nontraditional source:

Less than a tenth of the stock of household mortgage debt was funded by private-label MBS as late as 2003. By 2006 the share had ballooned to more than 22 per cent:

Most of the additional credit funded dubious loans. From 1990 through 2003, only about 10 per cent of mortgage originations (including refinancings) were classified as subprime or “Alt-A”, according to data from Inside Mortgage Finance. In 2005 and 2006, however, more than a third of originations were in these toxic categories:

That didn’t work out so well. Investors lacked the tools to evaluate and protect against credit risk, which was bad enough, but they were also systematically misled by originators and securitisers.

Since the peak in mid-2007, the private-label MBS market has shrunk by $2 trillion. It is now as big (small?) as it was at the start of 2002 in absolute terms, and back to where it was in 1992 in relative terms:

You could therefore be forgiven for thinking the entire experience was a mistake that shouldn’t be repeated. The ongoing shrinkage of the private-label market — down another $20 billion in the first quarter of this year — suggests there is little appetite for its revival among creditors.

The bubble in private-label issuance came at the expense of the GSEs.

Until 2004, about 55 per cent of all mortgage credit risk was borne by Fannie and Freddie. By 2006 this had dropped to 41 per cent. The growth in their market share since then simply reflects the ongoing demise of the private-label market:

In principle, some of the contraction in the private-label market could have been offset by traditional banks holding more mortgages on their balance sheets. But this hasn’t happened. While half of all US mortgage debt was held directly by US-chartered depository institutions as recently as 1991, American banks now have credit exposure to just 30 per cent:

The government has therefore been forced to pick up the slack. The growth in market share isn’t a sinister plot or a failure to learn from “the mistakes of the past”, it’s a natural consequence of the desire to sustain mortgage borrowing at a time when the private sector is unwilling to take the risk.

It could be worthwhile to change the current status of the GSEs as part of a broader effort to discourage household indebtedness, but it’s bizarre to suggest, as Powell seemed to, that you could make the housing market both more stable and less “rigid” by privatisating a functional utility.