Follow Insights Blog

CoreLogic

CoreLogic Econ

LATEST CORELOGIC ECON TWEETS

Purchase Mortgages, High LTVs May Up Fraud Risk in 2017

2016 Ends With Increase in National Fraud Index

Bridget Berg    |    Mortgage Performance

Mortgage fraud has been relatively low since strong lending controls were spurred in the aftermath of the financial crisis. But there are trends we expect to emerge in 2017 that may change that outlook, and the level of risk, significantly.

At the end of the fourth quarter, the CoreLogic Fraud Index rose to 122, matching its highest level from three years ago (the baseline of 100 is taken from Q3 2010). The fraud index increased at the same time that we saw a 20 plus percent drop-off in Q4 application volumes after the unexpected volume surge in Q3.

The long-term increase in the risk levels is related to increasing risk in purchase transactions and a greater share of purchase transactions in the industry. The loosening of credit policies at GSEs has helped boost higher LTVs. The share of high-LTV purchase loans has increased from 58 percent to 62 percent of all purchases between Q4 2013 and Q4 2016.

 What Title of figure 2 is

 What Title of figure 2 is

In figure 2, you can see a 32 percent increase in the high-LTV purchase segment over time. The Fraud Index value was 153 in Q2 2013 and 202 in Q4 2016.

It makes sense that refis are generally less susceptible to many types of fraud, because there are fewer players involved and it’s harder to manipulate the outcome. Purchase transactions on the other hand are more complex and distribute proceeds outside of the closed loop of financial systems – to property sellers, builders, real estate agents, etc. – instead of the funds going from one financial institution to pay off another. This creates more opportunities and motives for fraud. Examples of opportunities for fraud in purchases include falsified down payments and straw-buyer schemes, while motives for purchases include real estate and loan commissions, contingent transactions, seller profits, and seller distress. Historically, a rise in purchase activity is often mirrored by a rise in mortgage fraud.

CoreLogic Chief Economist Dr. Frank Nothaft expects purchase dollar-volume to rise by 6 percent in 2017 and to account for 68 percent of the total market. Nothaft, and others like the Mortgage Bankers Association, expect a similar purchase to refi mix in 2018 as well. At least one forecast, the MBA’s, puts purchases above 70 percent of the market in 2018.

Higher interest rates, which seem to be a given for 2017, and continued home price appreciation will create affordability challenges and may reignite “fraud for housing” schemes.

A smaller overall origination market, perhaps $1.5 trillion, in 2017 will most likely be accompanied by credit expansion as lenders try to capture a bigger share of a smaller pie and qualify more fringe borrowers. During the past several years, lenders have pulled back overlays that made lending policies more restrictive (and less risky) than agencies allow. One example of an overlay change is the reduction in the use of IRS income verifications. This may have driven our income fraud risk alert to increase 12.5 percent year over year in 2Q 2016.

One final trend that we’re watching: the return of small, non-bank players into the market. These lenders had a larger presence prior to the mortgage crisis. The crash, the end of subprime securitization and hefty buy-back demands winnowed down the number of small players, and for the past decade most of the lending was done by larger, more-regulated banks. But now smaller lenders are coming back into the market.

It will be interesting to see what the Fraud Index looks like in the months to come. Stay tuned.

This blog is based on analysis of loan-application fraud risk the mortgage industry is experiencing as measured quarterly by the CoreLogic Mortgage Application Fraud Risk Index, which is based on residential mortgage loan applications processed by CoreLogic LoanSafe Fraud Manager.

© 2017 CoreLogic, Inc. All rights reserved.