The old "implicit-guarantee-of-GSEs-didn't-work-so-make-the-government-guarantee-private-mortgage-securitizations" routine
Speaking on a panel on GSE reform at the Annual Distressed Investing Conference, Michael Frantantoni, chief economist for the Mortgage Bankers Association, invoked the, "implicit-GSE-guarantee-didn't-work-so-we-must-have-an-explicit-government-guarantee-of-residential-mortgage-securitizations" meme. As he put it:
What were some of the issues? A lot of it had to do with the ambiguity of these [government sponsored] enterprises. What was talked about was an implicit guarantee. You remember that during the end of June 2008 there were concerns about Freddie's ability to roll over their short term debt. And if you read Hank Paulson's biography, he was allegedly being threatened that China and Russia were going to pull back on their holdings of GSE securities, right?
So this ambiguity about what the government was standing behind and what they weren't was a real problem for the market. And again for our members having a stable secondary market was incredibly important. I think fixing that implicit versus explicit backing was incredibly important.The second aspect of that is if you want to support the market it probably makes more sense to support the MBS."
Ponder that for about 30 seconds. If the GSEs ever faced difficulty rolling over their short-term debt, do you think you might remember the headlines? In 2008 the GSEs were the only major source of liquidity in the $11 trillion U.S. mortgage market. So any suggestion of a disruption would have destabilized the broader system. Which is why any investor, including Chinese officials, would have wanted assurances that the unthinkable would never come to pass.
What debt funding problems? And nothing remotely like that ever came to pass. There was never a moment, ever, when the GSEs lacked access the unsecured debt markets. It is misleading to suggest otherwise. In 2008, the companies' cost of debt, measured as a spread over Treasuries, did rise, but not nearly as much as it did for other financial institutions, including other beneficiaries of an implicit guarantee, such as Citigroup, Bank of America and JPMorgan.
"Fannie Mae 10-year yield spreads narrowed by 1 basis point to 76.6 basis points, [over Treasuries]," reported Reuters on July 1, 2008. On September 3, 2008 the Reuters headline was, "Fannie Mae, Freddie Mac debt funding smooth," because both companies, "drew solid demand for $5 billion of new securities on Wednesday." On Friday evening the GSEs were informed of the impending government takeovers.
Anyone who followed the GSEs would know why the market never signaled any concerns about debt rollovers. Both companies held sufficient cash and cash equivalents to repay any debt that matured. Because the GSEs never had access to the discount window at the Fed, they always maintained 90-days liquidity. (Though banks could always pledge GSE securities and securitizations with the Fed.) Unlike companies in the shadow banking system, the GSEs did not fund long-term assets (loans) primarily with short-term debt. Look at their 2008 balance sheets.
Unlike the Wall Street banks, the GSEs never consumed a lot of liquidity in trading securities and derivatives, or in funding their corporate commitments when the commercial paper market shut down. Most of the credit extended by the GSEs was unfunded, in the form of guarantees. Unlike the banks, the GSEs always generated robust positive cash flows because mortgages on the balance sheet always show a significant prepayment rate.
GSE stock prices did get hammered, largely because of White House leaks detailing the long-planned takeover of the GSEs. (And because of Hank Paulson's illegal tipoff to hedge fund managers who could turn a quick buck by shorting the stock.) True, plummeting stock prices may be an early warning sign of future liquidity problems, but not when the government signals that its equity investment to shore up the companies may be highly dilutive.
Moreover, Fannie's public reports show precisely how it funded itself each month. And you can see how nothing much changed during the months leading up to the government takeover imposed on September 6, 2008.
The imaginary problem of an implicit GSE guarantee is used as the pretext for GSE "reform" that extends government guarantees on private label deals.
The imaginary problem of an implicit GSE guarantee is used as the pretext for GSE "reform" that extends government guarantees on private label deals.Until very recently, conventional wisdom in Washington dictated a singular path toward housing finance reform. That path leads to the abolition of Fannie Mae and Freddie Mac, which would be replaced by some new, untested system.
The basic idea is that that the government shall insure the catastrophic risk in privately issued mortgage securitizations. That is, private players shall assume the credit and interest rate risk on the first 10% loss, while the federal government shall insure the remaining 90%. This setup of private-public risk sharing would replace the GSEs. The template was a centerpiece in a 2011 Treasury Report, "Path Forward for Reforming America’s Housing Finance Market.” Later the template would be embodied in the Corker-Warner bill, the Johnson-Crapo bill, and in "A More Promising Road to GSE Reform," and in papers by like-minded individuals.
Bernanke decreed that the future must not include the GSE business model, which financed mortgages as a private corporation. He conveyed the impression that the replacement for the GSEs must operate with an explicit U.S. government guarantee, making it the functional equivalent to FHA or Ginnie Mae, could work. He said:
[T]he recent legislation [HERA] does not fully resolve the fundamental conflict between private shareholders and public purpose that is the source of many concerns about the GSEs. Considering some alternative forms for the GSEs (or for mortgage securitization generally) during this "time out" thus seems worthwhile. Needless to say, however, even if alternative organizational structures are considered for the future, the U.S. government's strong and effective guarantee of the obligations issued under the current GSE structure must be maintained. [Emphasis added.]
Bernanke explained the GSEs’with a made-up story:
[T]he existing GSE model involves an inherent conflict between the objectives of the companies' private shareholders and the objectives of public policy. For example, the GSEs were reluctant earlier this year to raise capital and to expand their operations, even though this would have helped financial and macroeconomic stability at a time of much-reduced mortgage availability. The GSEs' disinclination to support the mortgage market was motivated by the fact that raising additional capital would have diluted the values of the holdings of the existing private shareholders.
It's nonsense, because the GSEs were in the process of raising preferred shares, which are not dilutive. In May 2008 Fannie had raised $7 billion in capital this way.
It’s important to remember that when Bernanke spoke in October 2008, the GSEs had not yet released their third quarter 2008 financials. According to their second quarter financials, they were more than adequately capitalized. So Bernanke’s condemnation of the GSE business model was never based on any empirical data or analysis. It reflected his ideological agenda.
"It would seem advisable to retain some means of providing government support to the mortgage securitization process during times of turmoil," said Bernanke. "One possible approach, suggested by Federal Reserve Board economists Diana Hancockand Wayne Passmore, is to create a government bond insurer, analogous to the Federal Deposit Insurance Corporation." (Note that Bernanke referred to the “mortgage securitization process,” which was a much broader category than GSE securitization.)
A few hours later at the same event, Hancock and Passmore presented their idea to create a government bond insurer for all securitizations. The final version of their 2008 presentation was published in The B.E. Journal of Economic Analysis & Policy in 2009.
Hancock and Passmore continued to develop their idea for a GSE alternative, and in August 2010 the Fed published their paper, "An Analysis of Government Guarantees and the Functioning of Asset-Backed Securities Markets," which was then repackaged as another paper presented at a Brookings Institution conference on GSE reform on February 11, 2011. It's title: “Catastrophic Mortgage Insurance And The Reform Of Fannie Mae And Freddie Mac
Though the text of their papers is somewhat obfuscatory, Hancock and Passmore's oral and slide presentations are very clear. They leave no doubt as to why they believe the GSE implicit guarantee didn't work.
"The GSE debt holders ran, given the probability of substantial losses even with the credit guarantee,” said Passmore. “And that was because capital was inadequate, values of the portfolio were opaque and they were unsecured."
Passmore and Hancock did not misspeak, or exaggerate or embellish. They just made it up this wild totally false story out of thin air. Why? because they needed to show that the implicit guarantee of Fannie and Freddie did not work. So they assigned the failures of the private label securitization market on to the GSEs. That way, they could justify their agenda, which was to abolish the GSEs and get a government guarantee covering private label deals. If you're going to lie, LIE BIG.
None of the other speakers at the Brookings conference, including Frantantoni, seemed to notice.
5 comments - The old "implicit-guarantee-of-GSEs-didn't-work-so-make-the-government-guarantee-private-mortgage-securitizations" routine
Swap spreads (which were considered 'AA' in 2008) were only about 60 bp over Treasury -- 70 bp over means the GSEs were starting to trade below AA.
Their fixed income trading portfolio (the major source of revenue back in 2008) required yields very close to Treasury rates. Spreads of 70 bp over Treasury meant the fixed income book was toast.
It is hard for me to tell, but looking at the files on the Fannie Mae website, they are able to issue, in October 2016, non-callable five-year debt at 5 to 10 bp over the Treasury rate.
So 70 bp spread would have been a reason to send shockwaves.
Thanks for raising an excellent point about something I just mentioned in passing. I modified the post to illustrate and make the point more clearly.
Its been more than eight years, so many may not remember how all financial and corporates saw dramatic widening of spreads over Treasuries during 2008. The GSEs' spreads did not widen to the same extent as they did for others.
Exhibit 5 from this Fannie Mae Report illustrates the phenomenon:
This chapter by James Barth and other also illustrates how spreads had spiked during the runup to the September meltdown that occurred subsequent to the GSE takeovers.
Delete | Approve [ Hide ]
The key difference between banks and the GSEs is that banks can get funding through FDIC deposits (sub-LIBOR). Banks also were not leveraged 100X
The GSEs were able to get funding sub-Treasury!!!
GSEs need to borrow in the debt markets. And their original business model (at least Fannie Mae dating back to 1968), was the ability to borrow at or below Treasury (let alone LIBOR). And your charts show they were able to borrow below Treasury rates until the summer of 2008.
The lifeblood of the GSEs from 1981 to 2008 was making money from leverage. If they have to borrow above Treasury rates, they go broke almost immediately. Again, borrowing 70 bp over Treasury when the historic rate was at or below Treasury rates was horrific.
Each 10 bp increase in borrowing costs would cost the GSEs over $1.5 billion.
That is why the implicit guarantee model failed.
I have a different view. I think we both agree that a bank's or a GSE's effective cost of funds is derived from many different sources, or moving parts, and that credit spreads can widen or compress depending on market conditions. Before 2007, the difference between Libor, the fed funds rate, and financial companies commercial paper was negligible.
I do not think that anything that happened in 2008 supports the popular meme that the GSEs cannot operate without an express government guarantee because the GSEs always had robust liquidity.
If the GSEs' cost of funds spiked, then they had time to adjust. 30-year FRMs on their balance sheets are traditionally priced at about 185 bps above 10-year treasuries. And let's not forget that most of the credit extended by the GSEs, guarantees, is unfunded.
i don't recall seeing anything to indicate that their net interest margins got seriously impaired back then.
Since 1968, Fannie's existence (and Freddie after 1981) was based on their ability to borrow at or below Treasuries to fund this portfolio. As your chart shows their cost of borrowing went from below Treasuries to 70 bp over -- they were now just another AA company (albeit one with $1.5 trillion in debt).
And I suspect the spread they paid on swaps to hedge the interest rate risk on the retained portfolio also went up by 70 bp or more. They would have to issue debt to meet collateral calls on the swaps.
There was liquidity -- but the liquidity premium was the highest in nearly 40 years. The markets, in effect, had repudiated the 'implicit guarantee' and wanted Fannie to pay market rates.
The takeover dramatically lowered their funding costs -- which is why Treasury took them over.
Leave a reply
Is the Common Securitization Platform Legal?
10 Jan, 2017
GSEs' Liquidity At the Time They Were Taken Over
8 Jan, 2017
GSE Loan Performance
8 Jan, 2017
Shooting Down Dubious Arguments Against GSE Recapitalization
7 Jan, 2017
FHFA's Case For Zero Equity GSEs
30 Dec, 2016
Wharton Prof. Reminds Us Of The Power Of Social Stereotypes
27 Dec, 2016
GSE Reform And A Conspiracy of Silence
26 Dec, 2016