Tag Archives: Department of the Treasury

Mortgage REITs and Other Frights

Posted on by

The Office of Financial Research in the Department of the Treasury has released its 2013 Annual Report. It describes a number of things that should scare you as you put your head on your pillow at night and dream of the financial markets. It also describes some important steps that OFR is taking to get a handle on these potential nightmares.

One of the nightmares, relevant to readers of this blog, are Mortgage REITs. Mortgage Real Estate Investment Trusts (REITs) are “leveraged investment vehicles that borrow shorter-term funds in the repo market and invest in longer-term agency mortgage-backed securities (MBS).” (16) OFR identifies serious problems in this subsector:

Mortgage REITs have grown nearly fourfold since 2008 and now own about $350 billion of MBS, or 5 percent of the agency MBS market. Two firms dominate the sector, collectively holding two-thirds of assets. By leveraging investor funds about eight times, mortgage REITs returned annual dividend yields of about 15 percent to their investors over the past four years, when most fixed-income investments earned far less.Mortgage REITs obtain nearly all of their leverage in the repo market, secured by MBS collateral.

Lenders typically require that borrowers pledge 5 percent more collateral than the value of the loan,which implies that a mortgage REIT that is leveraged eight times must pledge more than 90 percent of its MBS portfolio to secure repo financing, leaving few unencumbered assets on its balance sheet. If repo lenders demand significantly more collateral or refuse to extend credit in adverse circumstances, mortgage REITs may be forced to sell MBS holdings. Timely asset liquidation and settlement may not be feasible in some cases, since a large portion of agency MBS trades occurs in a market that settles only once a month . . ..

Although their MBS holdings account for a relatively small share of the market, distress among mortgage REITs could have impacts on the broader repo market because agency MBS accounts for roughly one-third of the collateral in the triparty repo market. Mortgage REITs also embody interest rate and convexity risks, concentration risk, and leverage. For these reasons, forced-asset sales by mortgage REITs could amplify price declines and volatility in the MBS market and  broader funding markets, particularly in an already stressed market. (17)

Sounds like systemic risk to me.

Happily, the report also contains policy proposals to address some of these systemic risk concerns. First and foremost, it proposes the adoption of a Financial Stability Monitor tool to track financial threats. The OFR also proposes mortgage-specific tools. Reiterating the findings in a recent OFR white paper, the report calls for the creation of a universal mortgage identifier so that regulators and researchers can more quickly identify patterns in the mortgage market. Predicting financial crises is still more of an art than a science but it is a good development that OFR is trying to improve the quality of the data that regulators and researchers have about the financial market.

Social Security Numbers for Mortgages

Posted on by

McCormick and Calahan have posted Common Ground: The Need for a Universal Mortgage Loan Identifier, a Department of the Treasury Office Financial Research Working Paper (#0012). They argue that

The U.S. mortgage finance system is a critical part of our nation’s financial system, representing 70 percent of U.S. household liabilities. It is also highly complex, with many finance channels, participants, and regulators. The data produced by this system reflect that complexity; unfortunately, no single identifier exists to link the major loan‐level mortgage datasets. The establishment of a single, cradle‐to‐grave, universal mortgage identifier that cannot be linked to individuals using publicly‐available data would significantly benefit regulators and researchers by enabling better integration of the fragmented data produced by the U.S. mortgage finance system. Such an identifier could additionally serve as the foundation of a system that could benefit private market participants, as long as such a system protected individual privacy. (1)

This is a very important initiative, although the privacy concerns are very important to address. Regulators have been many steps behind the private sector in tracking developments in the mortgage markets and a cradle-to-grave identifier, like a Social Security Number for an individual, will help them (and private sector analysts for that matter) to track patterns among  borrowers and loan products.

The authors identify a number of serious privacy concerns:

a mortgage identifier would have to be designed to prevent market participants from re‐identifying individuals. No links from public documents to mortgage identifiers should be allowed. Otherwise the identifier could be used to identify individuals, rendering all datasets containing the identifier personally‐identifiable information. Such a designation would create concerns about the use of individual data in the private sector and trigger burdensome requirements for government researchers using the data. (3)

Researchers have proven resourceful at mashing up data sets to identify supposedly anonymous individuals, so the privacy protections that are ultimately implemented would need to be airtight. That being said, there is a lot of value in working toward the goal of a universal identifier.