The Future of RMBS Repo

A proposed US winding down of Fannie Mae and Freddie Mac has direct implications for the MBS repo markets, particularly agency repo. We dig into what the change could mean for bank balance sheets and the impact for clients needing funding.

Bloomberg published a story yesterday that was picked up by a number of mainstream newspapers covering a Senate plan to dismantle Fannie Mae and Freddie Mac. Sens. Tim Johnson (D-South Dakota) and Mike Capo (R-Idaho), both members of the Senate Banking Committee, are the co-sponsors of proposed legislation to wind down FNMA and FHLMC over a five year period and replace the current market structure with a government backed insurer. The key to this plan is that the government backed insurer would only step in after private investors have absorbed the first 10% of losses. There is a qualification embedded within the proposal that could stay the death sentence of the two GSEs temporarily in the event of market disruptions. While this does leave some ambiguity as to the complete reform of the US mortgage market, there is value examining the proposed future structure of this market and the impact this could have on repo.

The Bloomberg article included comments by a number of market analysts about the potential social impacts of this bill. First and foremost, the loss of a government guarantee will raise mortgage interest rates; there is no doubt about this. Surprisingly, none of the analysts quoted attempted to quantify this increase. A possible indication of the incremental cost of a non-insured mortgage might be found in the differences in rate of jumbo mortgages and conforming mortgages.

Historically, jumbo loans (loans that exceed the FNMA or FHLMC maximum principal) have been 25 to 50 basis points more expensive than conforming loans (those eligible for packaging into FNMA and FHLMC pools). Recently this spread has collapsed and in some cases turned negative as banks searched for high quality assets post 2008. Suffice it to say that given the current low rate environment the impact of this loss of federal insurance on the first 10% of a mortgage’s principal should only have a modest impact on consumer rates. What is harder to overcome is the impact this structure will have on the subprime market. Without federal insurance private investors will demand tighter underwriting standards for pass through securities. This will certainly decrease the opportunities for people with lower credit scores to secure competitively priced mortgages.

Leaving aside the societal impact, how will this change the mortgage market and with it the mortgage repo market? The current structure of mortgage backed securities dates to the late 70’s and early 80’s when the idea of selling of ‘pools’ of mortgages from banks to institutional investors caught on. Hence, the creation of a new asset class; Mortgage Backed Securities (MBS). Despite the events of 2008 and 2009 most market observers would agree that this process added capacity to the mortgage market and created a new class of high quality assets. As a measure of the success of this asset class, before the recent expansion of government debt, in 2008 the total size of the US mortgage backed market exceeded the US treasury market; $9.1 trillion versus $5.7 trillion (Source: SIFMA).

MBS trading has its own unique set of market conventions that can often confuse the uninitiated. In order to accommodate the constant supply of new pools, mortgage dealers established a forward market for their product. Liquidity in the mortgage bond market generally revolves around these dates. The funding of MBS product has historically conformed to these dates as well. If a client was buying FNMA pass-throughs for April settlement he would most likely request funding from a dealer that would cover the term from the April settlement date to the May settlement date – approximately one month. Dealers could repo these assets into the market for the same term or run the mismatch. Agency backed MBS enjoyed a quasi-government status in the hierarchy of repo collateral as the government sponsorship allowed a broad base of institutions to take them in as collateral. The only change to this structure came with the proliferation of ‘private label’ MBS beginning in the late 1990’s. This market grew partially in response to the conservative limits FNMA and FHLMC placed upon loan principals as well as the market’s undying faith in the safety of residential mortgages. In the early 2000’s the private label market eclipsed the agency market in terms of issuance. Because these securities did not have government sponsorship their funding proved a lot more difficult particularly in times of stress.

This proposed mortgage reform will have a dramatic impact on the MBS asset class. The removal of government sponsorship will remove the quasi-government status this asset class currently enjoys. Hence, they will in all likelihood lose their broad-based acceptance and become a non-liquid asset. This will require longer term funding as these new securities are subject to the full impact of a bank’s liquidity coverage ratio. To put this in perspective, the Fed’s March 5 dealers’ statistics show a total or $359 billion on agency MBS reverses, the majority of which would become non-liquid assets under this plan. FNMA and FHLMC also provide a universal structure to the pass-throughs they create. Without a consistent set of underwriting standards, the task of evaluating individual pools will become more difficult. This change in status would most likely remove this new class of securities from the FICC’s plan to create a solution to the issue of fire sales.

Finally, MBS are a one way product for a repo desk. By that we mean that their presence on a dealer’s balance sheet simply reflects funding which has been given to a client. There are no off setting trades which could be used to net this position down. Relative value hedge funds maintain long and short positions which can be netted down on a repo book. Individual mortgages cannot be shorted. To express a bearish view a trader goes short in the forward market. Simply stated, mortgage repo can be a significant consumer of balance sheet.

As financial reform continues, it becomes apparent that different asset classes will be impacted differently in the funding space. This proposed reform presents a significant challenge for mortgage financing. Ultimately to accommodate client demands for funding, a solution which shifts business away from dealers’ balance sheets or which bypasses dealers entirely may be required to support this market.