One of the subtler lessons of the financial crisis is that offloading one kind of risk can mean taking on another. At Freddie Mac, Kevin Palmer is taking this to heart.
Fannie Mae and Freddie Mac offload most of the credit risk on new mortgages that they acquire through the capital markets, getting taxpayers off the hook for losses. Their two biggest programs, Freddie's Structured Agency Credit Risk and Fannie's Connecticut Avenue Securities, have a wide investor base, allowing them to price this risk efficiently. But these two programs transfer the risk only after loans have been seasoned for a year or so, leaving taxpayers exposed in the meantime.
Mortgage lenders would like to use private mortgage insurance to "de-risk" loans before they sell them; this would allow them to negotiate lower fees from the government sponsored enterprises for assuming the remaining risk.
Originators and private mortgage insurers argue that de-risking loans before they are bundled into mortgage-backed securities is a better way to limit taxpayer exposure to potential losses than capital markets transactions. They point out that Fannie and Freddie typically hold onto loans for a year or so before selling securities linked to their credit performance.
Freddie Mac doesn't see it quite that way. According to Palmer, the company's senior vice president for credit risk transfer, timing is less important than what he terms "reimbursement risk," which is essentially counterparty risk. This risk can be found with either type of transaction, he says.
Palmer says that the housing finance system would function better if lenders retained a portion of the risk on loans that they service and originate, rather than purchase insurance. Several large lenders have sold loans to Freddie or Fannie in which they held onto some of the credit risk in exchange for paying a lower g-fee. But few lenders are able to come to terms with the GSEs on the appropriate level of these fees.
In an interview in early January, Palmer discussed proposals by the Mortgage Bankers Association and U.S. Mortgage Insurers for a pilot program that would use private insurance to take effective loan-to-value ratios on mortgage down to 50%. Below is an edited transcript of the conversation.
There's been a lot of discussion about the relative merits of transactions that offload risk before loans are sold to the GSEs and those that offload risk after the GSEs acquire loans. Is one kind better than the other, or is it important to have a variety of ways to offload risk?
KEVIN PALMER: One concept that is commonly misunderstood is the use of the terms "front end" and "back end." This terminology is only in reference to when the risk transfer transaction was arranged. Transferring the risk as soon as possible is important.
More important than the timing is ensuring that when that credit risk is transferred, we are comfortable that we will be able to get our recoveries in full and on time. On both front-end arranged and back-end arranged transactions, Freddie is still the guarantor on the credit risk. When the losses occur, we need to cover the credit losses and then seek reimbursement. We often look to mitigate the reimbursement risk through full or partial collateralization of our CRT [credit risk transfer] offerings. The reimbursement risk has to be taken into consideration on any CRT transaction.
The Federal Housing Finance Agency's 2016 Scorecard opens the door for private mortgage insurers to offer deeper coverage, but it stopped short of approving a pilot program. What do you think of the U.S. Mortgage Insurers' proposal to provide $50 billion of deep coverage insurance?
We are actively working with the FHFA and mortgage insurers to respond to the USMI proposal to further deepen mortgage insurance coverage. We are also going to be working on a request for information on different kinds of front-end arranged transactions.
Greater use of private mortgage insurance would put Fannie and Freddie at greater exposure to potential financial problems at the insurers themselves. Is that a good or bad tradeoff?
Currently, 30% of risk transferred on new business is done via primary mortgage insurance. This is based on how much stress loss recoveries we would get from mortgage insurance companies versus other private capital providers. So we already have a large amount of what's being called front-end risk transfer via mortgage insurance companies.
Before we did our first STACR transaction, mortgage insurance was our sole means of risk transfer. Since the introduction in 2013 of STACR, we've been building out other credit risk transfer capabilities to diversify and bring in additional capital while not reducing our use of mortgage insurance. We continue to transfer a significant amount of risk through the mortgage insurers and they continue to be strong partners.
You want a diversification of investors. In the past, we had little diversification; now we've improved our diversification of credit investors through a range of different CRT offerings. We internally, or others outside, can say that the 30% concentration of risk transfer to mortgage insurance companies is too high or low, but it should be acknowledged that there is a significant amount ongoing via front-end primary mortgage insurance.
Most all mortgage insurance that is purchased today is deeper coverage, or coverage well beyond what is required by our charters. This has been the case for many years.
A handful of larger lenders have entered into transactions in which the sell loans to the GSEs and keep some credit risk on their own books, at least initially, rather than using private mortgage insurance. How do you view these kinds of transactions?
The system would function a lot better if all lenders retained a portion of the risk on loans that they service and originate, often referred to as keeping skin in the game; unfortunately, due to Basel, holding this risk is expensive.
We have, over the last two years, worked with small, medium-sized and large originators on arrangements for them to keep the risk on loans they sell to us in exchange for a g-fee reduction. We've only done a couple of these transactions to date. That's primarily because banks are not interested in taking this risk due to Basel capital rules. That leaves nonbanks. We have done a couple of these transactions, and are looking for ways to expand this to more. Fortunately if we don't come to an agreement with a lender, most all of the risk will flow into one of our other CRT offerings.