The subprime lenders that worked: Part 2, why
All my blog posts seemed so far away
— we saw, in an era when national multi-billion dollar subprime lenders could collapse seemingly overnight, one category of lender — the state housing finance agencies — has been working in the same space, with the same target customers, in the same period of time, and yet these lenders have neither melted down nor sunk their clients. What are the HFAs doing right that the private subprime lenders did wrong?
I’m from the government and I’m here to help you
1. Originate-and-hold business model
Like the old-time savings bank, HFAs use a simpler business model. Instead of securitizing or selling off their portfolios, they simply hold on to the loans (or bundle them into HFA-backed bond issues):
An overriding factor is a difference in motivation. The interaction between borrowers and HFAs is more relationship-based, whereas for subprime lenders, the relationship is focused more on producing loans, which are then sold in the secondary mortgage market. HFAs hold on to their loans through maturity.
The specifics here deserve a bit of elucidation. HFA’s raise capital from the public markets by selling (usually tax-exempt) bonds.
Side note: The bonds’ exemption of their interest from Federal income tax enables HFAs to offer lower rates — in a perfectly efficient market, a discount equal to the marginal ordinary income tax bracket (today, 35% before considering state taxes). For instance, on 25 May, the ten-year taxable AAA stood at 5.58%, while the ten-year tax-exempt earned 4.05% — 27% lower. (The difference between 35% theoretical savings and 27% actual market represents a bit of inefficiency that is generally present in tax-exempts, for a host of reasons.)
They secure these bonds by assigning a particular pool of mortgages, plus other bond collateral (escrows raised in the bond issue), and an implicit brand-and-reputation commitment to keep the bonds from going into default.
Think of each of us as a happy individual mortgage!
(The rating agencies severely punish issuers who default, or even get close to default.) Though the HFA’s have thus raised money from the public, they have held on to those mortgages as a servicing agent, and the mortgages represent the main form of bond collateral.
As I’ve previously posted, bad lenders make bad loans, in part because they have agency risk between the lender’s agents and the lender, and between the originator and the eventual investing lender. Even a few horror stories of subprime origination certainly illustrate the power of the loan origination fee or brokerage fee in making the individual pro-transaction — at any level of implied or assumed future risk.
2. Less-aggressive underwriting
Lending less aggressively means that the HFAs got a smaller share of the market:
The subprime cats were after us
HFAs lost ground in terms of volume to the conventional mortgage market by not offering numerous loan options to less creditworthy borrowers. But in doing so, HFAs maintained a low to moderate risk profile that should limit any fallout from subprime lending.
Lending is a long-term business. Sometimes you win by losing.
During the refinancing boom of the early part of the decade, mortgage brokers and lenders experienced a large increase in volume. HFAs were often unable to compete and lost loans through prepayment. From 2002 to 2004, the 25 HFAs with Issuer Credit Ratings lost $5.3 billion in loans, more than 10% of their portfolio.
It’s also more than plausible that the HFAs were charging higher interest rates — or at least, higher starter payments — than their market competition. So not all of their runoff can be traced to conservatism … but a fair portion can.
Meanwhile, loan volume increased in the broader market, and these entities hired additional staff and offered an array of products to generate revenue.
3. Innovating program design to learn from market experiments
Even if not losing future money represents ‘real’ winning in the present, it feels like losing.
Consolation prizes in the mortgage business
HFAs continued to offer 30-year fixed-rate mortgages, using conventional insurance and guarantors like FHA and VA, or PMI (private mortgage insurance) loans using underwriting guidelines from Fannie Mae and Freddie Mac.
Further, the private sector innovates, which it usually does faster than the public sector. What’s a public body to do? Shamelessly copy — or partially copy!
Some HFAs expanded their product lines, but the new offerings did not involve the risk of many subprime loans. California Housing Finance Agency has a loan that pays interest only during the first five years, but the rate is fixed at closing for the entire term, which is extended to 35 years to permit an amortization period of 30 years.
Note that this product mixes some of the subprime innovations — longer-term, interest only — with a double safety valve: the rate is fixed now, and the uptick in payment happens five years out, and is probably manageable. (For instance, if the base rate is 6.0%, the increase is from 6.00% to 7.19%, a 12% increase five years out.)
Wyoming Community Development Authority has long offered loans to so-called subprime borrowers, but these borrowers must still provide full documentation verifying employment and income.
Finding the same customers, with different products
Some parts of the subprime sector were using ’self-documented’ underwriting, jocularly named ‘liar’s loans,’ a practice so fundamentally unsound that some of us are glad they went under.
The proper fate of self-documented originators
On this point, avoiding idiocy has served the HFAs well.
The only difference is that these borrowers can have a slightly lower FICO score.
That is, the HFAs used precisely the same due diligence and verification protocols, they just adopted a slightly more liberal/ Pelagian view of their applicant borrower’s credit history. That’s a conscious policy choice — one compatible with their mission and their capital structure (see below) — and a far cry from willful blindness to risk.
Delinquency on these loans in June 2006 was 4.8%, not much higher than the 3.8% for the indenture that holds the loans.
A delinquency rate of one in twenty is nothing to crow about, but it’s within the range of manageable.
We’re managing the delinquency rate
[Concluded tomorrow in Part 3.]
