Big Data Means Small Margins in Mortgage Industry of the Future
Automating the mortgage process will force tighter margins, but drive higher volume, for lenders.
The big story in mortgages today is the rise in mortgage loan rates. For the first time in years, we’re seeing 30-year fixed mortgage rates consistently above 4%, and a 5% rate is in sight. Higher rates make sense if you look at it one way: the economy is strong, inflation is climbing, and it’s safe to expect Federal Reserve hikes in 2018 and 2019.
Industry veterans might be sighing with relief. In the 10 years since the burst of the housing bubble, we’ve seen a slow economic recovery, a federal funds rate stuck at 0, and 30-year mortgage rates in the high 3% range for fixed-rate loans and lower for floating-rate loans. Is this the long-awaited return to normalcy, even if that normalcy comes with stricter lending standards?
The answer, I’m afraid, is no. Macro factors are pushing mortgage rates higher, but another element is going to start influencing mortgage rates, too: technology. And the direction of technology’s pressure on rates only goes in one direction — down. It’s not inconceivable that, 10 years from now, we’ll be remembering a 30-year fixed rate of 3.75% as the “good old days” of mortgage margin.
As rates have risen in the last 12 months, many people have pointed out that rates are still low by historical standards. And they’re right: when I bought my first home in the early 2000s, I would have killed for a rate that was under 5%. But it’s worth asking if those historical standards still apply in 2018?
The fact is that the historical standards emerged from a real estate and mortgage industry that is based primarily on gut decisions — people fall in love with homes, or think they’ve got just the opportunity to make a quick buck to pay for next year’s vacation. The problem is that gut thinking applies collectively to everyone, not just individual homebuyers.
I got into the business of real estate in 2009. My goal was to help clean up the mess that all these gut-led investment decisions had created. There were investors buying up the homes of NINJA (no income, no job, no assets) borrowers from foreclosure and dealing with ripple effects from the residential real estate “bubble” popping.
It’s not like the lenders had no information about the properties or the buyers. Credit scores were available, and property details and records existed in a multitude of bank, city and state offices. But the technology to collect this information and make sense of it wasn’t available. So we relied on human judgment … and, well, we all know how that turned out. Today, the technology is here. The disparate data sources are beginning to collaborate and standardize how data is entered and processed in the cloud. Financial institutions have gotten smart about allowing access to data and systems through APIs. Most importantly, artificial intelligence is advanced enough to effectively analyze data and make objective, data-driven decisions.
This is all heading to a place that can be summed up with a single word: automation. Many of the tasks that humans performed to help a buyer obtain a mortgage — from assessing a property’s value to pinpointing the buyer’s creditworthiness to sharing documents for buyer and seller to sign — can now be done using software. That’s making the industry faster. Homeowners can get a loan in a few days, rather than weeks. More and better data will also mean more and better loans, as valuations become more accurate and faster appraisals send earlier alerts to lenders and buyers about any potential loan issues.
But this trend will also mean lower rates. Today, the supply chain of a real estate transaction eats up about 12% (3% buyers agent commission, 3% sellers agent commission, 2% closing costs and 4% loan origination costs) of the money exchanged. That 12% pays the mortgage broker, the real estate agent, and, of course, the lender. That number is going to fall by at least a third in the next 10 years — and mortgage lenders will see their shares fall the most.