Press Room
 

June 26, 2006
JS-4338

Remarks of Emil W. Henry, Jr.
Assistant Secretary for Financial Institutions
U.S. Department of the Treasury

Before the Housing Policy Council of the
Financial Services Roundtable

Washington, DC- Thank you very much for that kind introduction.  I am very happy to be here.  The Roundtable was one of, if not the first, place that I visited after being sworn in at Treasury back in October.  I still remember my pleasant conversation with Steve Bartlett.  In that conversation, he asked if Treasury planned to get "more involved" in the GSE reform debate.  I think it is safe to say that we have and I think it is wonderful that I am back at the Roundtable to talk more about why there needs to be strong GSE reform legislation.

There are times when certain words or phrases are used or overused to such an extent that they verge on losing their meaning.  As anyone in the room today with teenage children understands, the impact of a parent's drumbeat of cautionary words, on any topic, can over time, diminish with repetition.  As the public GSE reform discussion crescendos, I fear we may be at such a point with the phrase "systemic risk".  If you have followed the arc of the GSE debate for the past few years then you will note that the terms "systemic risk" and "GSEs" are inextricable.  However, such repetition could make the likelihood of a systemic event occurring seem more the stuff of intellectual musing than hard reality.

There also appears to be significant misunderstanding about what it means and why we need to be concerned about it.  So, the purpose of my remarks today is to clarify what is meant by "systemic risk" as it relates to the GSEs, why we think it exists, what might transpire in a GSE-initiated systemic event, and why these are unnecessary risks that can and should be easily avoided.

At the outset, let me be clear on the meaning of systemic risk: it is the potential for the financial distress of a particular firm or group of firms to trigger broad spillover effects in financial markets, further triggering wrenching dislocations that affect broad economic performance.  Perhaps a useful analogy is to think about systemic risk as an illness that can become highly contagious.  

It is important to note that these types of concerns are not simply theoretical.  Like the case of a single gunshot setting off an avalanche, there are times when even seemingly modest or localized events in particular financial markets can trigger adverse consequences of enormous proportions.  One recent example of this type of event is global financial turmoil in 1998.

In August of 1998, Russia's external debt amounted to roughly $100 billion--a tiny fraction of global debt.  And yet that event led to panic and volatility in financial markets that ultimately triggered the Long Term Capital Management (LTCM) implosion and a period of significant financial distress.  LTCM pursued a convergence trading strategy.  It established very large positions across many markets, many of which were essentially bets that liquidity, credit, and volatility spreads would return to more normal levels.  Instead, spreads widened sharply in the financial turmoil following the Russian default and LTCM suffered losses greatly exceeding that predicted by conventional risk models.  The LTCM crisis laid bare the dangers of excessive leverage and perhaps more importantly, put a white hot light on creditors' and counterparties' over-confidence in the "hedged" nature of that fund's portfolio and strategy. 

If the Russian default could have such wide-reaching ramifications, you can understand why this Administration is so deeply concerned about the potential market repercussions of a deterioration in the financial conditions of a GSE, which collectively have more than $2 trillion of outstanding debt.

The hard lessons from LTCM include: i) the danger of investment decisions which rely upon the presumption of liquidity, ii) the importance of transparency and disclosure, iii) the extent of the interdependencies of our global markets, financial firms, investors and businesses, iv) the fact that complexity is sometimes the enemy of stability, v) the danger of complacency and false confidence in hedging strategies which, by definition can never hedge out all risk and which can produce the opposite of the desired effect in the absence of liquidity.

So, we at the Treasury are confident we are not simply "crying wolf". Before LTCM few, if any, would have guessed that it could have imposed significant systemic consequences for the financial markets.  And sadly, some are not heeding the important lessons from this experience in the GSE reform debate. 

To address such a looming problem, the Administration has consistently argued for meaningful reform of the regulatory structure of the GSEs. This reform must include mechanisms to protect the broader financial markets and our financial firms and counterparties from unnecessary risks.  The core basis for our policy of reform is the systemic risk presented by the size of the GSEs' mortgage investment portfolios and the corresponding concentration of risk in these two federally-chartered enterprises. Simply stated, our financial markets would be safer if these assets and associated risks were broadly redistributed.  And to add insult to this potential injury, these huge investment portfolios are much larger than what is necessary to accomplish the GSEs' mission.

The risks of the mortgage investment business are complex and far more difficult to manage than the risks of the GSEs other major business – the credit guarantee business. 

There are numerous levels of risk presented by the mortgage investment portfolios, but at a basic level the risk is created as follows: GSE portfolios are comprised primarily of fixed-rate mortgages, either held as whole loans, mortgage-backed securities (MBS), or other mortgage-related assets.  While mortgages in the U.S. typically allow borrowers the option to prepay at will, the aggregation of fixed-rate mortgages requires that the investor develop strategies to mitigate risks presented by these uncertain cash flows – both prepayments and extensions.    Unless the portfolios are hedged properly, in a period of significant interest rate movement, there is the risk to the GSEs that their assets and liabilities will be quickly become broadly mismatched which can lead to insolvency --much like the dynamics of the S&L crisis.

To properly hedge against such a dislocation, the burden rests on the GSEs to construct complex models of, among many things, borrower behavior, attempting to divine how and when borrowers will adjust behavior as interest rates change.  Once models are created the GSEs must, in addition to other things, deploy highly complex derivative-based strategies and other risk transfer mechanisms.  However, the risk, of course, never disappears. 

We all know that there are many large companies investing in mortgages that are exposed to similar risk.  So what makes the GSEs different?

There are three primary ways that the GSEs uniquely impose systemic risk on our financial system.  Taken individually, each reason might not be a cause for dramatic action.  However, aggregating each of these attributes under a single entity that also carries with it the broad misperception of a government backstop or guarantee creates a perfect storm scenario. 

The first element is the sheer size of the GSEs' investment portfolio.  Since 1990, the mortgage investment business of both of the housing GSEs has grown rapidly.  From 1990 through 2003, Fannie Mae's mortgage investments increased from $114 billion to $902 billion.  Freddie Mac's growth in mortgage investments was even more dramatic. From 1990 through 2003, Freddie Mac's mortgage investments increased from $22 billion to $660 billion.  Today's combined GSEs' mortgage investment portfolios still total almost $1.5 trillion.  By any standard, these are huge investment portfolios.

Secondly, the GSEs are not subject to the same degree of market discipline as other large mortgage investors.  That lack of market discipline is reflected in preferential funding rates that result directly from the market's long-standing false belief that the US government guarantees or stands behind GSE debt.  Of course, it is this funding advantage which drove the expansion of the portfolios in the first place.

To underscore this lack of discipline, imagine for just a moment if some of our most prominent complex financial institutions announced major accounting improprieties, significant restatements and serial failings and shortcomings in risk management and internal controls, and then further announced the cessation of annual reports and other standard disclosure materials.  Does anyone in this room doubt the ferocity of "market discipline" that would sweep down upon these institutions in the form of higher borrowing costs for market-based funding and heightened counterparty scrutiny?

Further complicating the external discipline picture is that the GSEs operate with less capital, meaning they are more leveraged than other financial institutions.  A non-GSE firm would have to have considerably more capital to access capital markets at anything close to the rates the GSEs are granted.  Greater leverage provides less of a capital cushion to absorb losses and it enhances the ability of the GSEs to grow.     

None of these obvious market-based checks have reined in the GSEs' growth.  Simply put, traditional market discipline has not applied for the GSEs.

The third element is the level of interconnectivity between the GSEs' mortgage investment activities and the other key players in our Nation's financial system. By way of example, as of December 31, 2005, commercial banks held $264 billion in GSE debt obligations (while not specifically broken out on call reports, given the relative size of the GSEs, the bulk of these obligations are likely those of Fannie Mae, Freddie Mac, and the FHLBanks). In comparison to bank tier-1 capital, GSE debt obligations exceeded 50 percent of capital for 54 percent of these commercial banks, and GSE debt obligations exceeded 100 percent of capital for 34 percent of these commercial banks.  In addition, the GSEs' interest rate positions are highly concentrated and pose significant risks to a number of large financial institutions. 

What I just laid out forms the basic framework around how the GSEs pose systemic risk:  large size; lack of market discipline; high degree of connections throughout our financial system.  While the U.S. financial markets are highly efficient and resilient, they are not infallible.  Now let's look at this issue even more closely.

Systemic events can unfold by direct and/or indirect spillovers.  Direct spillovers arise when the failure of a particular firm creates substantial losses for those who carry direct exposure with such firm, such as its creditors.  Indirect spillovers typically develop, not from direct exposures to the firm at the epicenter of the crisis, but when this firm causes a lack of confidence leading to a sense of panic and turbulence that results in action that generates substantial losses for firms that were not directly exposed to the impaired firm.  Such spillovers – not the initial event -- typically take the greatest toll on economic activity and, in the case of the GSEs, the potential for both direct and indirect spillover effects is nothing short of breathtaking.

How could such a systemic event begin?  They are many possible sparks but an unexpected sharp or volatile swing in interest rates, or in the parlance of risk managers, an interest rate "shock" would certainly be a distinct possibility.   Of course the GSEs claim to attempt to hedge their exposure for these types of events, though I note the following:

  • the OFHEO report suggests that the GSEs' focus on hedging such events have been, at best, lacking--at worst dangerously irresponsible; 
  • hedging is not an exact science and models are only as good as the judgments and expectations reflected in their inputs -- they are often wrong.
  • the LTCM episode provides a case study for the elements of a financial crisis; a number of these elements are present here:
  •  highly leveraged entities
  • presumptions of liquidity
  • enormously complex derivatives portfolios
  • abundant publicly-stated confidence in being properly hedged fostering a fundamental misperception as to the risks in the business
  • heavy reliance on risk-sensitivity models which, in the case of LTCM were, of course, wrong.

If such an interest rate shock occurred in a way that was not captured by the models, the results could be without precedent.  The immediate implication would be actual and mark-to-market losses.

The resulting actual or perceived inability of a GSE to meet its debt or MBS obligations or a significant decline in the market value of the GSEs debt obligations could be transmitted throughout the financial sector and the broader economy through a couple of channels. 

An obvious transmission mechanism is through direct losses to the commercial banking system, derivative counterparties, or other creditors.  If these key financial intermediaries suffered losses related to their GSE exposures, this could lead to a broader contraction of credit availability – for example fewer loans being made or more restrictive loan terms - that could have adverse implications f or overall credit availability and U.S. economic performance. 

In addition to the direct impact on the GSEs' creditors, consider, for example just a few of the other consequences.  A sharp deterioration in a GSE's financial condition would almost certainly increase risk premiums and boost yields on GSE debt and MBS relative to swap and Treasury yields.  Even if the rise in GSE yields might not fully reflect the true financial condition of the GSEs, institutions that are particularly exposed to GSE spreads to swaps and Treasuries in their ordinary course of business would be at risk in this scenario.  In particular, institutions such as large banks, hedge funds, and securities broker dealers that might hedge the interest rate risk in their MBS positions by establishing short positions in swaps and Treasuries could suffer substantial losses. 

To give you a sense of the potential scope of this one aspect of transmitting a GSE's financial problems, consider the group of primary dealers.  The Federal Reserve's primary dealer report indicates that the 22 primary dealers--a group that includes many of the dealer subsidiaries of the most important banks and investment banks in the United States--in aggregate typically maintain net long positions in GSE straight debt and MBS of about $130 billion and $30 billion, respectively.  These long positions are hedged in part by short positions in Treasuries on the order of about $130 billion.  Therefore any widening in GSE debt and MBS spreads over Treasuries would likely result in dealer losses that could be very substantial, especially relative to their capital.  Such losses might cause dealers to rein in their positions and market-making activities in the GSE debt and MBS markets and in many other markets as well.   Losses sustained by some primary dealers could well be large enough to reduce capital below regulatory minimums.   Risk spreads for many private firms would likely widen substantially and banks could choose to tighten credit availability.  Financial markets across the board would likely become very illiquid and volatile as firms with significant losses attempted to unwind their positions.

Unfortunately, that might not be the end of the story.  The GSEs make use of a considerable amount of short-term funding, that is then hedged to some degree to replicate long-term funding.  For example, short-term instruments account for more than 20 percent of all outstanding debt for both Fannie Mae and Freddie Mac.  In a financial crisis, the GSEs might face difficulty in accessing debt markets.  This difficultly might force the GSEs liquidate some MBS holdings, putting excessive downward pressure on prices in a market that the GSEs are supposed to be stabilizing. 

Those asset sales, in turn, would likely undermine confidence and exacerbate the sense of panic in the market and add to the losses of the GSEs and other entities that are major holders of GSE obligations.

I could elaborate further with various scenarios of how such a meltdown might play out in various corners of our world's capital markets.  I have not mentioned, for example, the potential volatility and unraveling of emerging markets that might ensue as they tend to do when a crisis results in a "flight to quality" mentality.

And of course, there are scenarios that could play out with foreign investors, a group that might not appreciate fully the GSEs' relationship with the U.S. Government and who own nearly $1 trillion of GSE debt and MBS.  Indeed, GSE obligations held on behalf of foreign official institutions at the Federal Reserve Bank of New York have been increasingly rapidly over the last 18 months and now exceed $500 billion. 

So there are virtually limitless scenarios. But you get the point.  We already know the lessons here. There is no need to endure the test.

I assume that some might contend that, despite the GSEs' current accounting, corporate governance, and risk management problems, what I have just laid out in terms of GSE getting into serious financial trouble is unlikely.  Past history reminds us that serious financial problems in the GSEs are not only a possibility, but an unfortunate reality.  And, I feel compelled to remind you that the federal government has taken steps to assist a troubled GSE in the past. 

Do we really want to be faced with unwarranted and irresponsible calls for bailing out another failed GSE?  

In fact, has it been so long that we have forgotten Fannie Mae's significant financial troubles in the late 1970s and early 1980s?  During this time period, Fannie Mae's balance sheet looked a lot like a savings & loan.  As interest rates rose, Fannie Mae's cost of funds rose above the interest rate it was earning on its long-term, fixed-rate mortgages. Like many S&Ls, Fannie Mae became insolvent on a mark-to-market basis.  It lost hundreds of millions of dollars.  Only a combination of legislative tax relief, regulatory forbearance, and a decline in interest rates allowed Fannie Mae to grow out of its problem.

 In the mid-1980s, the Farm Credit System (FCS) fell victim to a sharp drop in land prices, deterioration of agriculture market conditions, and increased interest rate volatility.  These economic factors coupled with poor interest rate risk management resulted in $2.7 billion in losses in 1985 followed by a $1.9 billion loss in 1986.  In the end, the federal government provided $1.26 billion to the FCS in financial aide. 

While I suppose those expectations were correct in the 1980s, as I noted recently, past government bailouts or assistance should not be viewed as a good predictor of future government actions. 

What I hope you ask yourself after hearing this is "Why?" and "What can we do about it?"

The answer to the first question is unsatisfying.  Ignoring all the rhetoric and spin, the simple truth is that there is no need for our financial markets to be exposed to this risk.  Passionate statements made by the GSEs to the contrary, the GSE investment portfolios are not necessary for them to stay true to their mission.

The answer to the second question is much more satisfying – we can address this risk rather easily.  As long as the portfolios of the GSEs are reduced gradually and responsibly, the overall impact to the housing market should be trivial.

I would be pleased to answer any questions you might have on this important topic. Thank you very much.