Analysts at Amherst Securities ask the “big question in the mortgage market”: “Will the next wave of defaults be driven by unemployment or by negative equity?”
Ask and answer. In a report published this week, Laurie Goodman and her team demonstrate that negative equity is far more predictive of loan defaults than is unemployment. The question is critical, because the policy response depends on the answer.
That is, if coming defaults are caused by unemployment, then the relevant response, says Goodman, would be to subsidize mortgage payments. On the other hand, if negative equity triggers defaults, then principal reduction must receive a higher priority in modification program waterfalls.
Let me put Amherst’s analysis in context. Goodman is addressing the current design of HAMP (Home Affordable Modification Program). It doesn’t address unemployment; in fact, borrowers must document income to participate. Moreover, HAMP procedures focus on reducing payments by lowering interest rate and extending term. Servicers may also forbear or forgive principal, but they are not required to do so. As a result, principal forbearance and forgiveness are rare in the small universe of HAMP modifications completed by September 1, 2009. According to Treasury data provided to the COP (Congressional Oversight Committee), of 1711 permanent HAMP modifications, only 261 had principal forborne and only 5 had principal forgiven.
This data is summarized in the COP’s October oversight report, An Assessment of Foreclosure Mitigation Efforts After Six Months
. However, in the COP’s analysis borrower’s equity is a contributing factor, while unemployment “now appears to be a central cause of nonpayment, further limiting the scope of the program”.
Unemployment, according to the COP, is presently driving a fifth wave of foreclosures. Not so, according to the Goodman’s evidence.
Goodman presents two pieces of evidence to make her point. The first is a matter of timing: default transition rates picked up long before unemployment. (To be precise, Amherst defines a “default” as a loan that becomes 60+ days delinquent for the first time. A 60+ delinquent loan has a low likelihood of cure.) Even among prime loans, small increases in transition rates were evident before the increases in unemployment.
But this simply sets the record straight on the timing of defaults and unemployment. To demonstrate that CLTV matters more than unemployment, Goodman constructed a matrix of transition rates (from 3 month previous) by sixteen unemployment/CLTV buckets within each product sector, Prime, Alt-A, Option ARM and Subprime.
Three conclusions emerged. One, even in the lowest unemployment column, higher CLTVs are associated with markedly higher default transition rates. Two, when borrowers have positive equity, unemployment plays a negligible role. Three, unemployment does impact borrowers with substantial negative equity, but far less than does CLTV. For example, for Alt-A borrowers with current estimated CLTVs > 120%, in a low unemployment area, the transition rate is 2.97%/month, compared to 3.5% in a high unemployment area. For a baseline, consider that the lowest unemployment/lowest CLTV bucket (<8%/<80% respectively) has a transition rate of 0.65%.
Carrying the analysis a step further, Amherst broke the data into owner-occupied and non-owner-occupied (investor and second home) categories. The same patterns emerged. Even for owner-occupied houses, unemployment only has a large impact where CLTV > 120%. For borrowers with very negative equity, unemployment is a catalyst
that can kick defaults up, not a cause.