Much is made in economics of “sticky wages” that prevent the price of labor from falling to meet the quantity demanded and thereby provide full employment. But the root cause of our prolonged recession is not insufficient consumer demand or overpriced labor; it is due to mortgage debt that is excessive compared to borrower income and house value. Reducing wages would cause even fewer people to be able to afford their mortgage and exacerbate a cycle of foreclosures, falling house prices, lower consumer spending, and lower business investment. It is more appropriate to say that “sticky interest rates” are the cause of our current economic problems.
The housing bubble left behind elevated mortgage debt that continues to sap consumer spending and business risk taking. Savings are increasingly directed to the safest of investments, causing interest rates to fall to historically low levels. Although this benefits higher credit quality borrowers, those with poor credit or equity are unable to refinance.
Yet these are the same borrowers who are constraining the economy through the fear of their default and their lower spending patterns. I would guess that 1 in 4 Americans fall into this category. Some of these chose high-risk mortgages, but many were simply unfortunate on when and where they bought a house.
The reasonable approach would be to modify the borrowers who can’t refinance at current low market rates. Unfortunately, creditors became believers in risk management only after the bubble burst and are now zealots who are inappropriately applying underwriting standards and risk based pricing to their existing high risk borrowers. These borrowers can’t possibly meet the new standards – some never could, and never should have gotten the loan. But poor past credit standards are largely a sunk cost that can never be recovered.
The purpose of loss mitigation is not to meet a return on capital target or low default rate; it is simply to minimize losses. Workouts do not require an additional investment of capital – that investment choice was already made (it was a bad one). If 3 out of 4 loans in a workout program re-default, then it was a success merely by preventing one of the four loans from causing a loss. Cutting losses by 25% doesn’t sound great, but on the bottom line it is similar to increasing profits by 25%. The correct questions are if any additional loans could have been saved, and if the optimal workout program was chosen. The cost of workouts is usually small compared to the cost of default.
If we are trying to mitigate losses on high-risk mortgages, then we can extend the term, forgive principal, and/or cut the interest rates. Extending the term is fine, but if a mortgage originated in 2006 has a 30-yr term, then most of their current payment is going to pay interest. Extending the term at the same interest rate won’t reduce the payment much. Forgiving principal will certainly reduce defaults, but it will cost the creditor a lot, create a moral hazard incentive for new delinquencies, cause creditors to limit future lending for fear of more forgiveness, and anger everyone who didn’t get forgiven.
Consider the components of a mortgage interest rate and how they now apply to high-risk borrowers:
- Risk-free profit: Short-term LIBOR & US Treasury yields are near zero.
- Inflation risk: Low by most measures.
- Refinance option: High-risk borrowers can’t refi regardless of market rates.
- Origination costs: Sunk costs, if they haven’t been recovered by now…
- Servicing: Defaults have much higher servicing cost than performing loans.
- Credit risk: Very high potential credit losses.
The credit risk component of the interest rate is admittedly extremely high. But consider that the higher the interest rate, the higher the mortgage payments, and the more likely it is that the borrower will default. If one were to charge an interest rate proportionate to the borrower’s risk, then they couldn’t possible afford it and would certainly default. One should be happy to continue to get the current credit premium.
When I add up the components of interest rate on a high-risk borrower, I come up with a modification interest rate of 1-2%. I like 1% because it grabs headlines and should spur action. Is this how I want to invest my savings, on 1% crummy mortgages? No way. But if I already own the crummy mortgages, then it sure beats paying 50+% in losses every time one defaults.
One can easily make the 1% mortgage a quid pro quo by making the modification a 1-month ARM. That way the borrower is taking the risk of higher future interest rates in exchange for lower interest rates today. Since the Fed says they want to keep rates low for at least the next two years, this isn’t that bad a bet. And when rates do start going up, hopefully employment and house prices will too.
The common fad is to discuss workouts in terms of Net Present Value (NPV sounds sophisticated to people in DC, and naive to people in finance). If only people really understood the mouthful they say with “NPV”. It assumes an accuracy of prepayment, default, and loss severity models over a variety of different workout alternatives. It assumes an accuracy on future house price and interest rate paths. It sounds good in theory. In practice, the NPV estimate will be incredibly rough. People will tend to choose assumptions that favor whatever outcome they desire anyway. Complexity doesn’t equal accuracy. Complexity does equal operational disaster, at least in the mortgage servicing world.
If we want to solve the mortgage, housing, and general economic problems then we need to “un-sticky” mortgage interest rates. We don’t have to do 1%, or even 2% modifications. We don’t have to do it to all delinquent borrowers. But we do have to do something.