The Continuing Fight Over Fannie and Freddie, and the Real Problem of US Mortgages

I’m a bit late to comment on an important series by the New York Times’ Gretchen Morgenson on the continuing, until now largely behind the scenes, fight over the future of the giant mortgage guarnators, Fannie and Freddie.

In her first article, A Revolving Door Helps Big Banks’ Quiet Campaign to Muscle Out Fannie and Freddie, Morgenson showed, in gory detail, how Wall Street had labored mightily to take over the activities of the mortgage behemoths for their fun and profit. Mind you, it’s not as if they already take a big proportion of the savings from the mortgage guarantee for themselves now; you can see that in refis, where much of the benefit of the interest rate reduction is chewed up in fees and other charges. Her second article, Fannie and Freddie’s Government Rescue Has Come With Claws, discusses in depth how the Administration flagrantly violated the terms of its own bailout deal to hoover up the earnings from Fannie and Freddie for Treasury, rather than give shareholders the proportion they were due, and took other punitive measures that were contrary to the 2008 legislation that set forth how a conservatorship of Fannie and Freddie would work.

These pieces provide insight into the state of play with government sponsored enterprise “reform” and also as a case study of how banks influence government policy, and how eager the Obama Administration has been to take up their cause (not that a Republican or Clinton Administration would behave any differently).

Cynics may regard these two stories as a bit moot, since the banks appear to have lost the fight to take over the operations of the government sponsored enterprises, and the case over the alleged mistreatment of Fannie and Freddie shareholders is being adjudicated. But don’t mistake the fact that the story (save for Morgeson’s recap) is out of the headlines for the notion that the banks aren’t continuing to move forward. As Morgenson points out, they are working to see how much they can get the regulators to cede to them through administrative processes, meaning without getting legislation passed. And as a Congressional staffer stressed, the big financial firms are moving their game pieces into place so they can take up the fight again after the 2016 elections.

But Morgenson’s focus on the machinations of the banks and their allies in government resulted in her not incorporating the policy issues. And without an understanding the policy problems, it’s easy to draw the wrong conclusions about how we got where we are and what might be the best approaches going forward.

At the core is an issue that no one wants to ‘fess up to: that the wildly popular 30 year fixed rate mortgage, which in most cases a borrower can refinance at any time, is a product that would not exist absent government support. It does not exist in any other advanced economy mortgage market (at least in any size). The reason is that it is extremely unattractive to lenders. The lender is potentially exposed to a very long term credit risk*, and he is also exposed to the worst sort of interest rate risk. A long term fixed rate bond is already risky, and is particularly unattractive in a low-yield environment). But even worse, the refi option in US mortgages means that the mortgage disappears (as in gets refinanced) when it turns out to be a winner for the borrower, that is, when interest rates fall.**

A Short History of the Evolution of the Mortgage Market and the GSEs

We’ve gotten here through an odd set of events. Prior to the Depression, mortgages were typically five-year bullets (building and loan societies offered longer maturities of 10 to 12 years). The bank could reassess the borrower every five years and set new terms based on market conditions and any change in the borrower’s fortunes. Down payments were generally higher than now and most borrowers sought to pay down their mortgages over time. The bank failures of the 1930s left many destitute (my maternal grandparents had their money in three different banks, all of which collapsed. They eventually got three cents back on the dollar from one of them). And even people who managed to have funds in banks that survived still found it almost impossible to refinance when their mortgages came due to the general dearth of credit.

Although many had already lost their homes by then, the Federal government set up the Home Owners’ Loan Corporation in 1933. It bought mortgages from banks on a full cashout basis, and was able to give borrowers much lower payments by having lower interest rates than banks offered (5% for HOLC versus the prevailing rate of 6% to 8%) and longer maturities (15 years) with the monthly payments designed to amortize principal over the of the loan. The program was derided by critics as wildly reckless at the time, and 20% of the HOLC loans did default. Even so, when the program was finally wound down, it posted a modest profits.

That started the US down the path of longer-term, fixed rate fully amortizing mortgages.

Government guaranteed mortgages and Fannie Mae were also creations of the Depression. The FHA provided FHA insurance to FHA-aproved lenders who underwrote loans to FHA standards.

But because this type of mortgage is so pervasively used, eliminating it would have an immediate impact on housing prices, and even assuming someone were willing to try, gradually restricting its use would have to be done in a very attenuated manner. The FHA also supported the issuance of longer-term, fixed rate mortgages (which needless to say, by making housing more “affordable” and increasing the pool of eligible borrowers, helped support home prices).

Fannie Mae was established in 1938 as a Federal agency and put the government even more directly in the business of financing new mortgages, albeit in a backdoor way, by having Fannie Mae provide “liquidity”. Banks could sell mortgages to Fannie Mae, which would free up their balance sheets to make new mortgages. Private lenders eventually started complaining that Fannie was competing unfairly with them in buying mortgages in the secondary market, and demanded that Fannie be wound down or made private. Instead, the compromise was that it was made a “partly private” entity, with the government holding preferred stock and private investors owning common.

In mind that in 1968, Fannie’s was effectively split into two entities, due mainly to the need of President Johnson to no longer consolidate Fannie’s debt as part of Federal indebtedness.*** “government sponsored enterprise” (in part because Johnson did not want to consolidate Fannie’s debt on the government balance sheet), some of Fannie’s functions were returned to full government control and given to the agency Ginnie Mae, which was part of the newly-created HUD. Freddie Mac was created as a private company with an initial cash contribution from the government. It too was to provide “liquidity” to mortgage lenders, in particular savings & loans, that were getting whacked by interest rate increases on their deposits. Freddie’s strategy at the outset was different than Fannie’s, with Freddie packaging the mortgages it bought into mortgage-backed securities, while Fannie at the time was retaining. them. Over time, the two institutions’ strategies became indistinguishable.

Why Does This History Matter?

Fannie, Freddie, Ginnie, and the FHA were all created to deliver Federal subsidies to the housing market. They are also integral to the fact that the overwhelmingly dominant mortgage product, the 30 year fixed rate mortgage with a borrower refi option would not exist absent this extensive government support. Removing that support would mean that other types of mortgages that lenders and investors would be willing to hold absent government support would replace the staple of the 30 year fixed rate mortgage. The result would be much higher interest rates, shorter mortgage terms, and more borrower assumption of interest rate risk. All of those would mean more costly mortgages, which translates into much lower housing prices. That’s not acceptable from a policy standpoint, so the political fight will be over how these subsidies will be administered in the future.

Let us also not forget that despite overt denials (meaning on the face of the securities offerings) of government support of the GSEs, the market correctly assumed pre-crisis that they did enjoy an informal guarantee. Jim Hamilton discussed that fact at some length at the 2007 Jackson Hole conference, that the extreme undercapitalization and super low funding cost of the GSEs meant that investors regarded them as government guaranteed. We have also been told that US officials gave Chinese officials assurance that the US would always support the GSEs (recall that the Chinese government was and is a major holder of US bonds).

So while it might be preferable not to have the government so deeply involved in housing finance, particularly since housing finance is an inefficient way to support the housing market (it’s hard to analyze the impact of the subsidies and they tend to leak to all sorts of parties that are not necessarily the targets of policy), there’s no easy way out of this position. Any reduction of government support for the mortgage market, which would have to include a change in terms to conventional mortgages, would need to take place on an attenuated basis.

Why Have the Efforts to Dismantle the GSEs Failed So Far?

This is the elephant in the room that Morgenson’s otherwise fine articles fail to address. The Republican party hates the GSE, in part because Fannie Mae in particular helped cement a housing coalition that was staunchly Democratic. But as Morgenson points out, the Administration was on board with the program for winding down the GSEs, just not as quickly as the Republicans were. Was this a clever headfake (as in the delay was really a stealth plan to stymie the Republicans) or was something more at work?

In the interest of brevity, I’ll make some observations and let readers chime in.

First, I am actually a bit surprised (as well as gratified) that the efforts to dismember the GSEs have failed so far. As indicated, I’m not a fan of these subsidies, but we are where we are, and as questionable as these subsidies are, it’s better to have them under government control than in a public/private hybrid (which as we saw did not work out very well) or even worse, with purely private actors getting more subsidies and having even lower accountability than they do now by virtue of being private. The privatization plans are basically another scheme for looting, unless the new “private” entities are set up and run as utilities, with regulated profits and oversight of executive pay and perks.

Second, my guess as to why the plans put forward a few years ago never got anywhere is they were too greedy and lame. The only bennie of “private” ownership was that the new mortgage guarantors would be a bit better capitalized than Fannie and Freddie. Um, there’s no reason to dismantle Fannie and Freddie to to that. Another optical excuse was that there would be more “sons of Fannie and Freddie” created, so they would be more robust. As readers of Andrew Ross Sorkin’s Too Big to Fail may recall, Freddie was clearly in trouble, yet Paulson insisted on putting both Fannie and Freddie in conservatorship over the vociferous objections of Fannie’s board. Why? Paulson argued, and I think correctly, that the market did not see Fannie and Freddie as all that different, and putting Freddie in conservatorship would result in an immediate spike in borrowing rates and the availability of credit to Fannie, which had the potential to put it in a crisis. The same logic would apply to “sons of Fannie and Freddie” which would all be pursuing the same general strategy.

Third, and I suspect this is a far bigger consideration than anyone has acknowledged publicly, is that the GSEs have huge and very complex systems, both computer systems and the human processes and the documentation around them. Reconstituting them in new entities is an extremely daunting task. I saw nothing in the “son of Fannie and Freddie” proposals that gave me any comfort that the designers were taking these extremely large operational risks seriously.

As indicated at the outset of this post, the Morgenson articles are an important reminder that the battle over the future of Fannie and Freddie, and who controls government subsidies to housing, is ongoing even though it has retreated to the background for now. And given how the Presidential election appears to be giving social and broader economic issues higher priority, it’s not clear that any of the candidates will have much to say on this issue. But rest assured it will come back to the fore not all that long into a new Administration.

___
* A 30 year mortgage is shorter in duration than, say a 30 year Treasury or corporate bond, since the bond amortizes over its life. Due to the prepayment options, MBS are priced off of 5 to 7 year bonds, since the odds at a time when yields are “normal” that a lot of borrowers will experience enough of an interest rate decline in the first ten years of their mortgage for them to have a refi opportunity that most take. And mortgages are also paid off due to selling the house as a result of death, divorce, job loss, moving, or trading up. However, a top mortgage expert has been pointing out for years that the mortgage originators, as well as the Fed, the GSEs, and policy-makers, have greatly underestimated what will happen if/when interest rates go up. Everyone will hang onto the current crop of low-interest rate mortgages. He contends those mortgages that were modeled as being a 5 to 7 year risk will turn out to be a 15 year risk. That in turn will create all sorts of havoc for holders, who might have made some allowance for making a bad bet that was outstanding for the balance of the 5 to 7 year average life, but not for 15 years.

** Other countries, including ones with much higher ownership rates than the US, have borrowers assume at least some interest rate risk, via adjustable interest rate mortgages, or ones that reset periodically. The US has had some more borrower-friendly adjustable-rate mortgages, such as ones that float only in a certain range (as in they have both a rate ceiling and a floor). That means that borrowers get the benefit of declines in interest rates without having to go through the cost and hassle of refis.

*** One of the biggest macroeconomic mistakes of all time was Johnson’s refusal to raise taxes to finance the unpopular Vietnam War (or in MMT terms, to drain demand so as to combat inflation). Running government deficits at a time of full employment when the Federal government was sending a man to the moon, fighting a war on poverty, and leading a ground war in Asia set off the inflationary spiral of the early 1970s, which was made worse by the oil shock.

Print Friendly
Tweet about this on TwitterDigg thisShare on Reddit0Share on StumbleUpon0Share on Facebook0Share on LinkedIn3Share on Google+0Buffer this pageEmail this to someone

29 comments

  1. Michael Hudson

    Another way of looking at Morgenson’s article is, why didn’t the government treat Citigroup, Bank of America etc the SAME way they treated Freddie: Get their stock, charge them 10% interest, etc.
    Granted that there was asymmetry. But why not take the opportunity to take the offending Citigroup etc into the public domain as a public option, undercutting the credit c are rates and writing down loans as an alternative to the biggest bank consortia?

    1. Brian

      re; “why didn’t the government treat Citigroup, Bank of America etc the SAME way they treated Freddie: Get their stock, charge them 10% interest, etc.” Each of the above mentioned sold everything they purchased to private or the GSE buyer. If it was all new money backed by the full faith and credit, the GSE was now the finance arm for the treasury. Ownership was obfuscated due to irregularities, which bothered no one as the government skimmed money to pay interest on the national debt. If all the “lenders” were pretending they had lent money, or had their own money in the game when it all came from the purchasers of MBS, the legal niceties are shown to be ignored. They created MERS to hide what they were hiding.
      The end result was the separation of the payment stream for the trust deed and the ownership of the note. There is no legal obligation for the grantor to pay if the obligation was paid for them. Many grantors find that their note is alleged to be in one of these MBS trusts. Yet it can’t be unless it is being paid every month. How then can the borrower be in default if the payment is being made? Why is the payment being made? A Ponzi requires new money, and new MBS buyers are a must to keep the interest payment on the debt being made so there are new MBS buyers.
      They just didn’t tell anyone that they destroyed the land title system, UCC 3/9 ownership and transfer requirements, and paid off the loan for each buyer to make the system appear solvent and benefit the skim. Then they foreclose, illegally, working with others to subvert law, backed by the government who doesn’t care who they get the money from. Odd, because there are statutes that discuss this kind of act and its consequences. For now.

      1. alex morfesis

        Dont blame poor charles ponzi…he actually paid back his investors 95 cents on the dollar(nyt aug sept oct 1920 archives)…9.5 million in…9 million back out…to the victor go the history books…me thinks he was money laundering…most “investors” were boston police officers…

        But the capital cycle you are mentioning has a missing part in your argument…americans pay 4 to 8 percent interest rates on home loans…but the actual aaa rated global oecd rate for mortgages is around 2 percent globally…the skim is not from new investors…the skim is in illegal collusion to keep charging americans higher interest rates and then using the excess cash flow to create new capital merry-go-rounds with the left over money…trading on the lower at risk portions of home loan stream payments are in fact safe like treasuries…but money market rates from aaa ratings were never passed on to consumers as originally planned in the mid 90’s…

        This myth that americans Must pay a rate for home loans from the days before color television is confusing to me and not based on any empirical data…just rentier legacy nonsense…

        The world has changed and is always changing…looking back at data for investment purposes is amusing but useless…american capital markets have replaced gold as the store of value for now and probably for the next 50 years…there is only one statue of liberty and symbols and myths are much more powerful than most “intelligent” folks want to accept…if the usa said anyone could move here for a 20 grand payment per person payable over five years plus a 20% surcharge excise tax on income plus a minimum tax of 5 grand per year….there would be 20 million people lined up every year to step into this “mess” we complain about…

        We will never see behind the curtain as to the derivatives that i believe were winning hands at both fannie and freddie…but a decision was made that better to kill the gse group than kill off wall street…besides bnp paribas who pulled the carpet out from under the derivatives market on aug 7 2007 just did not much appreciate the subpoena it was being handed asking about its moving of funds in the un iraqi oil for food issues…thankfully for them and their iraqi main shareholder…no one has ever turned back to connect the dots…especially not the american firms that got a “national security” burp from that nice three letter agency with all those former abwehr girls and boys…

        The truth is hardly ever what it appears to be…

        Oh…and the problem is article 8 of ucc…not 3 and 9…besides the ucc is the original ttip…look at the fine print…it has a private “tribunal” of corporate lawyers who get to “interpret” it …thankfully for corp america two us supreme court justices “died off” within 40 days or so of each other in 1949 and truman replaced two basically liberal judges with conservatives…

        To the victor go the history books…especially when helped along by a few “convenient” demises…

  2. David NYC

    Loan performance on 30-year and 15-year fixed rate mortgages is vastly superior to ARMs, even in a low interest rate environment.

    When people talk about the government “subsidy” they’re really talking about the pre-conservatorship GSE business model. The GSEs can finance vast numbers of FRMs for two reasons: First, their mortgage backed securities transfer interest rate risk, but not credit risk. All GSE securitizations benefited from a corporate guarantee, which obviates the biggest risk, which is market timing. Second, the GSEs could actively manage the interest rate risk embedded in their balance sheet loans through a variety of interest rate derivatives.

    Also, it’s worth mentioning that the GSEs track record in credit underwriting is unmatched.
    You can quickly eyeball their 40-year history of credit losses on pp 89 and 106 in the regulator’s annual report to Congress.
    http://www.fhfa.gov/AboutUs/Reports/ReportDocuments/FHFA_2014_Report_to_Congress.pdf

    There also seems to be a reluctance to admit another inconvenient truth. Private label RMBS proved to be an unmitigated failure, which is why they haven’t returned except in nominal amounts. Aside from the moral hazard problem, which neither the banks nor the government will address, the math simply doesn’t work. The arranger of an RMBS deal must calculate an NPV for a static portfolio of loans in liquidation. There are no do overs post closing. Tranched RMBS were developed in the 1980s, when nobody refinanced to get a lower rate, so at that time, it was possible to estimate the cash flows based on historical data.

    1. washunate

      When people talk about the government “subsidy” they’re really talking about the pre-conservatorship GSE business model.

      David, maybe you have a separate article planned, but I’d be curious to hear more of your thoughts along those lines.

      Obviously there are mechanistic and legal differences pre-and post-conservatorship, but those operational nuances are not generally what people mean when railing against subsidies. The frustration with government subsidies is that they transfer money to more affluent households while driving up the price of housing for lower income households. Malinvestment is perhaps nowhere more clear in its social and environmental destruction than the mess that is using public policy to artificially inflate real estate prices.

      1. David in NYC

        It’s a long story, but basically it overlooks the fact that whatever subsidy one can estimate, the benefit to financial markets and the country is that credit losses on 30-year fixed rate mortgages are very, very low, and only the GSEs can provide them. As Lewis Ranieri has pointed out many times, there is no long term history that shows ARMs can work.

        I’ll mention two pivotal sources of that mythology:

        1. After the fact free Fannie Mae “accounting scandal” the company’s stock didn’t decline that much. Why? Because sophisticated investors can follow debits and credits, and see that the alleged accounting violations were BS. Once Fannie restated its financials the phony $11 billion “loss” was transformed into a $4 billion profit.

        But GSE critics saw the resilience if Fannie stock as “proof” that investors didn’t care about the GSEs’ financials, they interred that investors presumed the implicit guarantee was a firm guarantee, so the GSEs were de facto subsidized by the government.

        2. There were a series of fatally flawed studies done by Wayne Passmore at the Fed, which argued that the government was subsidizing the GSEs by the implicit guarantee, and this subsidy gave the GSEs a much lower cost of funds, so the private sector couldn’t compete. Passmore overlooked the fact that a bank’s cost of funds is a blended rate from debt of varying maturities. Plus, he ignored the cost of interest rate derivatives to hedge interest rate risk. Plus he ignored the fact that the cheapest source of funds, anywhere, is from consumer bank deposits. Plus he ignored the fact that credit spreads widen or tighten, depending on the overall interest rate environment.

        I think those two items served to instill the belief that the GSEs were subsidized, which caused the usual suspects from right wing think tanks to say that 30-year fixed rates are bad. They have zero empirical data to support their case.

        1. Yves Smith Post author

          You are ignoring a basic fact: no other marker in the world has as 30 year fixed rate mortgage as a widespread, much the less standard product. Everywhere else in the world has borrowers taking more interest rate risk than the US.

          Ranieri is invoking American exceptionalism, our refusal to learn from the experience of other countries. to argue for more government subsidies for his market.

          1. washunate

            Well said. I was hoping originally there might be more to the thought process, but alas this is what it appears to be.

            Housing policy has created some fascinating cognitive dissonance.

    2. Yves Smith Post author

      That’s because ARMs in US pre-crisis were targeting subprime borrowers and on top of that had low teaser rates and very large rate resets. They are not comparable to the sort of ARMs that are widely used in other markets.

      In other mortgage markets in the world, there are either no government guaranteed mortgages or the mortgages that are subsidized represent a much smaller % of the market, generally because the government sets much more stringent requirements.

      Finally, the reason there have been non-gov’t guaranteed mortgages post crisis. There are tons of jumbos being done, but they are with higher downpayments in the conforming mortgage market, and are being kept on bank balance sheets.

      And the reasons there is no subprime securitization is sell-side greed. The FDIC surveyed investors and told the industry the four reforms they’d need to implement to have a private label subprime market:

      1. One year seasoning

      2. No CDOs

      3. Buyers have access to loan tapes

      4. Risk retention by seller (I forget the level but it was a stipulated %).

      So it is not that we can’t have a subprime market. It is that instead they like their fat margins and aren’t prepared to give them up.

      1. Ryan

        Exactly Yves – David, the data suggests otherwise on ARMS, specifically those guaranteed by fannie and freddie. The average credit score on agency guaranteed ARMs in 2006, 2007 and (what little was originated) in 2008 was >100 points lower that the average credit score originated today. Even the non option ARMS that came without teaser rates had sub 700 FICO scores and frequently came with 80% LTVs and another 20% second lien behind it…oh and sometimes with interest-only payments for…I’m not kidding…10 years. These pools went 40+% delinquent by 2010. But if you had the cash and didn’t default, you pretty much got the most advantageous mortgage loan possible, as they reset down to libor+225bp (during ZIRP) after 3-5 years with no principal payments for another 5.

        Today it’s a different world, and the borrower base within the ARM sector is dramatically different, both due to (stating the obvious) stricter underwriting standards and market aversion to anything called an “ARM”. In fact all of the available pooling and prepayment data suggests that credit quality, borrower sophistication, and delinquencies in Fannie/Freddie ARMs today are more favorable than they are in fixed-rates. ARMs used to be a dumping ground for subprime borrowers who needed them as an affordibility product. Today they are really used by the savvy as a cash-flow management tool, plus, for shorter tenor Arms you still have to qualify to make payments at the current note rate + 2%. Would be happy to discuss more offline, but there’s a lot of structures in the ARM space that would make sense as an alternative to the 30yr sector.

      2. Leonard

        It has become undeniable that the wide swings in interest rates is very damaging to our market economy. The very high and very low interest rates are caused by flaws in our monetary and fiscal policies.
        I have developed a policy that would help stabilize our economy and interest rates.
        I would consider it an honor if you could read and review the article I have written about the policy. The article is located at wp.me/p42WQA-7c or at http://www.taxpolicyusa.wordpress.com
        Thank you

    3. Jim Haygood

      ‘The GSEs could actively manage the interest rate risk embedded in their balance sheet loans through a variety of interest rate derivatives.’

      That was the theory. In practice, both Freddie Mac and Fannie Mae ended up restating earnings by multi-billions, in the years before they blow up In Freddie Mac’s case, “the company understated net income by a total of $7.6 billion. The largest single component, nearly $5 billion, was a result of improper accounting for investments in derivatives.”

      http://www.nytimes.com/2003/11/22/business/freddie-mac-understated-its-earnings-by-5-billion.html

      As for “transferring interest rate risk but not credit risk,” Fannie Mae still retains a $367 billion mortgage portfolio as of 9/30/2015. According to Fannie Mae’s calculations, it would lurch downward in value by $61 billion in the event of a 50-basis point interest rate shock (page 39):

      http://www.freddiemac.com/investors/pdffiles/investor-presentation.pdf

      Fannie is still using derivatives to hedge this interest rate risk. Meanwhile, as of 9/30/2015, its assets of $3.23 trillion are 813 times its shareholder equity of under $4 billion. That’s “prudence,” Kongressional style!

      1. David in NYC

        A few basics.
        You may not be aware, but Treasury has drained the GSEs off about $250 billion in cash that could have built up the companies’ capital. That’s not Congress and that’s not the GSE business mode, that’s Treasury’s agenda to put the companies out of business.

        Every balance sheet mortgage lender is required to use interest rate derivatives, whether the rate is fixed for 3 years or 30 years. This became a fact of life when interest rates were deregulated in the early 1980s. The same applies to all corporate fixed rate bank loans, but for retail lending it’s done on a portfolio basis.

        Also, this is a hard fact about accounting for hedges and derivatives. There’s always a disconnect if one side of the hedge is marked to market and the other side isn’t. Interest rate derivatives are always marked to market, whereas bank loans are recorded at historical cost. The lender will report mark-to-market gains and losses from derivatives, which tell you nothing about the cash flow, or economic, hedges.

        A mark-to-market gain or loss may reflect a change in the yield curve, which may have nothing to do with the hedges locking in interest rate exposure.

        That’s why the GSEs publish their duration gap each month to show how the tenor of their assets match up to the tenor of their interest rate obligations, (the GSEs actively swap the 90-day rate into the 10-year rate.)

        Again, the securitizations transfer interest rate risk but not credit risk, which is why they are profoundly different from private securitizations, which assume both risks. GSE balance sheet loans assume both risks, and again, that is actively managed, as it is with all balance sheet lenders.

        As the footnote says on page 39 of the Freddie presentation, that $61 billion potential exposure assumma, “an instantaneous 50 basis point shock to interest rates, assuming no rebalancing actions are undertaken and assuming the mortgage-to-LIBOR basis does not change.” The GSEs take rebalancing actions every business day. If the 10-year rate changed by 50 bps, I can assure you mortgage-to-LIBOR basis would change as well.

        1. Yves Smith Post author

          Dave,

          You are really off base. If you ever again mislead readers with jargon-filled nonsense. I will ban you. I have very little tolerance for people posting false information and trying to pass it off as gospel truth. It’s one thing to volunteer a point of view and label it as such, or to indicate where you are sure of your facts (or can link to someone else and other readers might take issue with who you cited, but you can at least point to some support for your view). It’s quit another to say things that are just incorrect, like “bank loans are recorded at historical cost”. That is not true:

          A fair value option (FVO) has recently been adopted by the Financial Accounting Standards Board (FASB). Under FASB’s FVO, acompany can elect to measure nearly any financial asset or liability at fair value in its financial statements. This standard is effective as of the beginning of an entity’s first fiscal year that begins after November 15, 2007, with early adoption permitted. This election is generally available at inception of a financial instrument on an instrument-by-instrument basis; however, once the decision to use fair value is made for a particular instrument, it is irrevocable for that instrument. Loans and unfunded loan commitments would be eligible to be measured at fair value under FASB’s FVO.

          http://www.federalreserve.gov/pubs/feds/2007/200729/200729pap.pdf

          Because I lack the energy to shred your comment and I am in no mood to argue with you, I outsourced the task to world recognized derivatives expert Satyajit Das, who graciously responded. His comments in bold:

          Every balance sheet mortgage lender is required to use interest rate derivatives, whether the rate is fixed for 3 years or 30 years. This became a fact of life when interest rates were deregulated in the early 1980s. The same applies to all corporate fixed rate bank loans, but for retail lending it’s done on a portfolio basis. There is no requirement to hedge as such. It is a risk management decision. If bank’s did not gap ie mismatch short term liabilities with lonng term assets (both in a liquidity sense and interest rate risk sense) then this has profitability implications.

          Also, this is a hard fact about accounting for hedges and derivatives. There’s always a disconnect if one side of the hedge is marked to market and the other side isn’t. Interest rate derivatives are always marked to market, whereas bank loans are recorded at historical cost. This is not true. The rule is based around ‘available for sale’. many loan assets which are intended to be sold off or securitised must be placed in the tarding book and hence marked to market. This is the case in most jurisdictions that I am aware of. The lender will report mark-to-market gains and losses from derivatives, which tell you nothing about the cash flow, or economic, hedges. Derivatives are generally marked-to-market if they are in the trading book. However, wehre interest rate derivatives are used as hedgesfor banking book assets then they would not be marked-to-market. This is the case in most jurisdictions that I am aware of.

          A mark-to-market gain or loss may reflect a change in the yield curve, which may have nothing to do with the hedges locking in interest rate exposure. I don’t understand this point. When you hedge banking assets or trading book assets great care is taken to try to match the risk profiles. So if we are hedged why should there be a mark-to-market loss on the underlying asset plus the hedge unless it isn’t a hedge.

          That’s why the GSEs publish their duration gap each month to show how the tenor of their assets match up to the tenor of their interest rate obligations, (the GSEs actively swap the 90-day rate into the 10-year rate.) The real problem is actually pre-payment risk – the ability to prepay fixed rate mortgages without full economic penalty. GSE duration gaps can change wildely based on this. GSEs hedge this with swaptions and/ or issue of callable bonds. These can create mark-to-market problems.

          Again, the securitizations transfer interest rate risk but not credit risk, which is why they are profoundly different from private securitizations, which assume both risks. GSE balance sheet loans assume both risks, and again, that is actively managed, as it is with all balance sheet lenders. I don’t understand this. All securitisation shifts credit risk as well as interest rate risk. All credit risk is transferred other than any retention of tranches or explicit guarantee. If this was not the casse then the assets would remain on the sponsor/ originator balance sheet as lacking true sale or risk transfer characteriistcis. In the case of the GSEs, the real issue is whether theunderlying mortgages are guaranteed by the government.

          As the footnote says on page 39 of the Freddie presentation, that $61 billion potential exposure assumma, “an instantaneous 50 basis point shock to interest rates, assuming no rebalancing actions are undertaken and assuming the mortgage-to-LIBOR basis does not change.” The GSEs take rebalancing actions every business day. If the 10-year rate changed by 50 bps, I can assure you mortgage-to-LIBOR basis would change as well. As noted above, a lot of things go into this but a significant component is the prepayment optionality.

          1. vlade

            The rules of mtm or not to mtm are arcane and complex, and definitely nowhere as simplistic as Dave suggests. Not to mention, that its entirely misleading, as the derivatives for securitization are held by trust, not the sponsor, and the trust is not subject to banking regultion (althoughg it may be consolidated on sponsor balance sheet these days).

            Of course, if the sponsor is the derivtive provider (which is a complex issue unto itself), then the derivtives are in a trading book and hit by mtm volatility. Otoh, in such a case, the derivtive provider hedged this sale into the market, so any such vol is offset by the hedged mtm. BTW, it’s not unusual for the sponsor to ‘juice’ the securitization by taking their profit via the derivtive (as it crystlises the PnL pretty much immediately, unlike holding it in a banking book)

            As Das writes, prepayment is one of the main risks in mortgage securitization.

            Not touched in Das response is that prepayment has impact on IR hedging (well, he writes about swaptions, but these ade of limited use), as your principal changes have re-hedging costs. It’s actually worse for nonUS banks, or I’d say for banks that issue rmbses in different currency from the underlying assets (common in the uk).

            The costs of cross currency basis rebedging can be punitive, which is why the uk adopted the master trust structure with scheduled amortization and calls (which, incidentally, worked quite well during the crisis, and likely will work still ok evn with rising rates most uk mortgages have only about 2 years fix iirc)

            going back to the credit risk, this time in the uk – even Granite, the worst of the worst (loans originated by northern rock) had zero credit losses to investors on ANY tranche sold to investors, and given the size of sellers share in MT its unlikely that there ever will be any loss.

            1. Yves Smith Post author

              Thanks for your addition. I am sorry for getting stroppy with Dave, but I really try to do the best I can to inform readers, and get upset with myself when I am wrong or incomplete, and I get frustrated when others who comment here don’t seem to care about understanding and signaling the limits of their knowledge. Your further clarification is very much appreciated.

    4. Yves Smith Post author

      On your point re interest rate derivatives, the GSEs have hedged their interest rate risk, so I have no idea where your claim comes from. In fact, their hedging activities in 2002. when the GSEs were a much smaller % of the mortgage market than now, was so large that it was systemically destablizing (the hedging is pro-cyclical). Greenspan was very concerned about it and that is probably why he was pushing ARMs around that time. The accounting scandal at Freddie, which led to restrictions on balance sheet growth being imposed on them, gave Greenspasn a break on a front he didn’t expect.

      As to underwriting, the GSEs hog the best credits (including the houses in a price range with lower resale risk than jumbos) so if they weren’t getting good results with their market focus, something would be very wrong. And despite your over-the-top praise, their loan losses post crisis were 2x what they’d allowed for in their models.

    5. Ryan

      “Loan performance on 30-year and 15-year fixed rate mortgages is vastly superior to ARMs, even in a low interest rate environment.” For loans that pass through DU or LP (fannie/freddie’s underwriting engines) this is absolutely untrue. 6 years ago, absolutely. Not today.

  3. washunate

    …hoover up the earnings from Fannie and Freddie for Treasury, rather than give shareholders the proportion they were due…

    I’m curious what that means? Why would shareholders be due anything? GSE debt was private debt that was socialized onto the backs of taxpayers to keep fraudulent housing prices artificially inflated. The government had zero obligation to backstop Fannie and Freddie. In addition, the process of socializing losses leads directly to more malinvestment that makes society overall even worse off over time.

    The result would be much higher interest rates, shorter mortgage terms, and more borrower assumption of interest rate risk. All of those would mean more costly mortgages, which translates into much lower housing prices. That’s not acceptable from a policy standpoint…

    Of course, from a different policy perspective, affordable housing is superior to expensive housing :-)

      1. washunate

        The housing subsidy show must go on.

        Longer term, bigger forces are at work, but in the short term it’s fun watching the squabble. Basically one side said hey, lookie, some money. We’ll take that. And the other side said, hey, you can’t do that. We stole that money fair and square.

        And the first side chuckles and says, uh, you do know we wrote HERA and those other signature bills back at the tail end of the Bush presidency, right? You want to tell us that we’re doing it wrong? Now that’s funny.

  4. Clive

    TV sitcom “The Golden Girls” taught me all I (or almost anyone else) ever needed to know about the U.S. housing market:

    Phony Medium / New Age Crank: “This house is possessed by an evil spirit! You must leave this house at once!”

    Blanche Devereaux: “Actually, it’s possessed by Miami Federal and at 3%, you couldn’t blast me out of here”

  5. MH

    * A 30 year mortgage is shorter in duration than, say a 30 year Treasury or corporate bond, since the bond amortizes over its life.

    Should this read “since the mortgage amortizes over its life”?

  6. debitor serf

    most mortgages aren’t held for 30 years, they’re only held by the mortgagor for 7 or less before refinancing. and the first 7 years of a loan are the most profitable. and during this 7 years mortgage will often be sold and resold by banks or investors as part of their portfolios. as a practical matter, few borrowers these days actually take a 30 year mortgage and pay it for 30 years. even a refi’d mortgage reamortized over 30 years gives a lender 7 more good years of interest. it’s like perpetual and lifetime debt. that’s the american exceptionalism – keeping homeowners in debt forever.

    1. Yves Smith Post author

      1. We said in the post that the average life of a mortgage has been 5-7 years and that mortgages are priced off 5-7 year Treasuries.

      2. We ALSO said that if the Fed really does increase rates, you can guarantee that everyone will hold onto today’s super low interest rate mortgages. They will NOT be refied, and no one’s modeling is anticipating what will happen. For example, we had a discussion of interest rate hedging. The hedges will be all wrong if the effective maturities extend a ton.

      3. Most mortgages are securitized. They are conveyed through a series of bankruptcy-remote entities to a trust. They stay in that trust. They are not resold. The bonds created from the cash flows in the mortgage pool in that trust will be traded, but not the mortgages.

      Jumbos (loans too large to be Fannie/Freddie “conforming” loans) are often held on bank balance sheets and may be sold, but not most mortgages.

  7. Tim Duncan

    F&F have scale which allows them to create efficiencies that are beneficial to American homebuyers. This is particularly evident in their ability to develop effective credit models and policies and to sustain the TBA market – which effectively manages the duration risk of 30 year mortgages. Both F&F did remarkably well and validated their underwriting ability even under the tremendous strains of the financial crisis. Their huge losses were attributable to tax credit write-offs which were subsequently reversed when the market leveled out.
    In my humble opinion, the GSEs, warts and all, are shining examples of how government can work and, as noted, these public efficiencies would be transferred to banks and mortgage companies from consumers if the government mortgage guaranty (implicit or explicit) was allocated other than through the GSEs. The mistake was made by Johnson in making the GSEs private and the cat has now been put back in the bag – including the prohibition of lobbying and political contributions by the GSEs.

    One point that is often missing in discussions relating to affordability is the contribution of the GSEs and FHA to housing demand and the resulting increase in home prices. Is it possible that the infusion of excess financing into the market ultimately does more harm than good to potential middle-class buyers by keeping the price of homes at an unnaturally high level? In other words, pumping government guarantied money into the housing market to promote affordability may have negative consequences at a certain point by causing prices to rise to rapidly. Has this been explored sufficiently?

  8. aa

    First you take Manhatann, then you take Berlin.

    First you respect the Rule of Law, then you take whatever you want.

    Or you become a Banana Republic.

  9. David in NYC

    Yves

    I certainly didn’t intend to be needlessly inflammatory or provocative, and I’m sorry it came across that way. I think you and others raised important points worth mentioning. I should have been clearer to avoid confusion.
    You are absolutely right, beginning on November 15, 2007, lenders had the opportunity to assign fair values to loans on their books. I should have made clear that I was referring to the pre-2007 rules.

    I should have also made clear that I was not referring to loans held on the books to be securitized; yes those loans are not hedged, only the loans held on the books until they are paid off.

    I also should have made clear that federal regulated institutions are required to hedge their interest rate exposures under safety and soundness regulatory guidance; it’s not a set in stone accounting rule.

Comments are closed.