Unintended Consequences of Aggressive Regulation

"In case you're looking for some light reading this weekend, FRB-Philadelphia working paper 21-08, Does CFPB Oversight Crimp Credit?" 

~ Robert Morro (@bmorro44)

So went a tweet from one of my Twitter connections. So I listened to him, and dialed up the FRB-Philadelphia WP 21-08. Forty three pages later, with formulas like:

CFPBLoanSharelt = αl + β · Postt + εlt

And a 27 page appendix with subtitles like:

Alternative difference-in-differences model

I got my answer.

Note that the study was limited to Federal Housing Administration (FHA) loans. For the uninitiated, an FHA loan is government-backed mortgage insured by the FHA. It is popular with first-time homebuyers, low to moderate income (LMI) homebuyers, and those with relatively low credit scores... as low as 500. It normally has a low down payment requirement, as low as 3.5%. You can see how it accelerates home ownership, which many economists and policy makers feel is critical to improving net worth.

The study analyzed FHA lenders that were subject to CFPB oversight, which were non-bank FHA lenders and bank FHA lenders with greater than $10 billion in total assets. The period analyzed was immediately prior to the commencement of CFPB oversight in 2011, and afterward. The study also analyzed the impact of the change in oversight intensity brought about by the 2016 election. 

Conclusions

The study revealed intended and unintended consequences. First, the intended consequence was that CFPB oversight is associated with improvement in mortgage servicing practices, leading FHA mortgages from CFPB-supervised banks to become less likely to transition from moderate to serious delinquency. Poor servicing practices were an important driver of the foreclosure crisis during the Great Recession; the study results suggest that tighter regulatory oversight may help reduce inefficient foreclosures during economic downturns. This is good.

There were two unintended consequences. The first was regulator arbitrage, where a bank decides to have an activity regulated by one entity rather than another because of the perception that the chosen regulator will be less risky to them. In this case, bank holding companies (BHC) with mortgage subsidiaries in the holding company had a tendency to put the mortgage sub under the bank where it would be regulated by the bank's primary regulator and not the CFPB so long as the bank was less than $10 billion in assets. If it was in the BHC, it would be a non-bank mortgage lender, and therefore subject to CFPB oversight, even if the bank was under the $10 billion threshold. Meaning: banks chose to avoid CFPB oversight.

The second, and presumably more actionable unintended consequence: Banks subject to CFPB oversight decreased FHA lending and replaced that activity with jumbo mortgage lending. To avoid the reputation and regulatory risk of doing FHA lending, banks chose to do less of it, and therefore reduced the amount of credit advanced to first-time homebuyers and LMI families. Jumbo mortgages are typically for larger homes for wealthier and more credit-established families. There was no conclusive evidence that non-bank FHA lenders and under $10 billion FHA bank lenders increased FHA lending to make up for the shortfall.

Let that sink in a bit.

Further, after the 2016 election, banks subject to CFPB oversight began anew with FHA mortgage lending. They must have perceived lower reputation and regulatory risk in doing so, and therefore came back into the market. 

This is actionable information for both lawmakers and regulators. Particularly given the aggressive rhetoric coming out of the new-teeth CFPB.

Maybe they should read the study. And modify their approach. One can hope. But as Red from Shawshank Redemption said:

"Let me tell you something my friend. Hope is a dangerous thing. Hope can drive a man insane."

Add FRB-Philadelphia WP 21-08 to your night time reading list!

~ Jeff

Don't Bank. SoFi

After its most recent capital raise in September, SoFi, a marketplace lender that focuses on millennials, has raised nearly $1.5 billion in equity capital since its founding in 2011. By comparison, over 100 year old and $7.7 billion in asset Union Bank & Trust in Virginia had $1.1 billion in equity capital, with a market capitalization close to book value. 

Since 2011, SoFi has funded over $6 billion in loans (through December 15, 2015). And today, they are embarked on a campaign against banks.

Warren Buffett was right when he said "You never know who's swimming naked until the tide goes out." SoFi started in 2011, so the tide has not yet gone out on them. Like most marketplace lenders, SoFi claims a borrower risk rating system that is better than the FICO score. And in fact, is claiming a FICO score free zone. How good is their system compared to FICO, or other FinTechs that feel they are more evolved in credit risk management? We don't know. And I suspect we won't know. Until the next time the tide goes out.

Remember during the depression when the National Housing Act of 1934 created the Federal Housing Administration (FHA). Perhaps not. But during the depression, the typical mortgage was a five year balloon. So when 1/3 of borrowers lost their jobs, they were unable to re-finance when their balloon came due. Forcing them out of their homes. In came the FHA, with mortgage insurance to protect lenders, and the beginning of what is now a 30-year mortgage with a significant secondary market.

The secondary market removes unpalatable interest rate risk from financial institutions' balance sheets. Similarly, SBA guarantees reduce risk for small business lending. Reducing risk by transferring it to someone else is not new.

Marketplace lenders reduce risk for their loans similar to how financial institutions do it for mortgages and small business loans. They transfer it to investors. Many of the investors are other financial institutions! 

Sure, the originating institution retains some risk. I'm not privileged to see the contracts and know if there is recourse back to marketplace lenders for their loans. For financial institutions originating and selling residential mortgages, they retain some risk for early pre-payments or early defaults up to a certain point in time. And for fraud without a time limit. Absent those items, mortgages sold in the secondary market are at the risk of the investors.

So, too, I would suspect is the risk to SoFi and other marketplace lenders. Meaning the lion's share of SoFi swimming naked when the tide goes out is born by those that buy those loans. It is an agency problem that we saw before, in the 2007-08 mortgage crisis, when mortgage brokers and bankers, knowing they were transferring risk, pushed all sorts of loans through the system, so long as they adhered, however loosely, to the investors' underwriting criteria. What was best for the borrower was an after thought, in many cases.

Each recession is different. And certain asset classes are affected differently. In 2007-08, the first asset class impacted was residential mortgages. Banks likely felt it more in their investment portfolio as they were chock full of collateralized mortgage obligations and mortgage backed securities. Fannie Mae preferreds, anyone? 

Next recession may be different. And it will be interesting to see how the portfolios generated by these marketplace lenders perform. And who would actually own those portfolios should they tumble.

Maybe we can create another agency like the CFPB! Something to look forward to.

~ Jeff