Hickam v. Occam The Mortgage Industry Data Paradox
Some people spent the 4th of July holiday, and other holidays for that matter, enjoying the sun, the beach, picnics, family, a bottle of beer, a glass of wine, something stronger, and generally taking it easy. I am not most people. While I did all those things, I also spent some time reviewing historic and current mortgage lending performance data.
Why? Part of it is natural curiosity about an industry where I have spent most of my adult life. From mortgage analyst and accountant to investor, originator, servicer, technologist and consultant, I cannot help but be interested. And with yet another refinance boom coming to an end, there is, all of a sudden, a lot more to be interested in.
Another part of it, honestly, is frustration with long-term lending performance. Originating mortgages, periodic refinance booms since 1980 aside, is not a very profitable or attractive business. In fact, the 10 or so years that followed the housing crisis of 2008 were abysmal. Let’s call that the lost decade. But you knew that and besides, it is well-documented.
What has my attention now, and hopefully has the attention of everyone responsible for originating mortgages profitably, is avoiding a repeat performance of that lost decade.
Breaking the mortgage performance cycle
Our history of avoiding repeat performances, sadly, is not good. There have been roughly five major refinance boom / bust cycles since about 1980. Profitability tanks after each boom for a number of years, usually until another refinance boom comes along and often just in the nick of time. Isn’t it interesting that nothing, not new processes, new technology or an entire new approach to the business of originating mortgages has come along to ‘fix’ this?
And here we are again.
And where are we, exactly? The last two years, 2020 and 2021, will go down in history as two years of lavish abundance in everything related to mortgage origination. The most loans ever, rejoiceable profitability, relatively low production costs, the highest productivity in almost a decade and enough data about lending performance to satisfy even the most insatiable mortgage nerds.
It is on this last area of abundance, mortgage performance data, where I believe the answer to long-term profitability lies hidden. The good news is that the actionable information, the data every mortgage lender must have to assess their performance, is available and plentiful. Let’s not mistake available and plentiful for useful or actionable. The two don’t have to equate nor do they necessarily correlate. And that’s the bad news: performance data is so abundant we fail to see the forest for the trees.
Which of the multitude of available data points, if I want to manage my business, do I focus on? How do I put the data to work?
Making sense of the mortgage data
We’re confident we have the answer. I have played many roles in the mortgage industry, some of which I mentioned above. My favorite role, regardless of my actual job, has been analyst. I was fortunate, as many of us are, to have worked with really smart people who were willing to coach and mentor me and who nurtured my analyst tendencies when I started working in mortgage origination.
We had very little in the way of sophisticated tools (OK, we didn’t have any, really). So, I learned how to manage a mortgage origination business with a legal pad and an HP-12C by focusing on a handful of metrics.
If you’ve used a 12-C you know it was pretty groovy in its day. If you haven’t used one or haven’t heard of them, even better. We need fancier tools if we’re going to make this business long-term profitable.
I still have that 12-C, by the way, and I still use it regularly. I still use legal pads, too. Both are still good for thinking about concepts. The real work, though, gets done in spreadsheets and databases, which leads us to the data paradox.
There are only five metrics mortgage originators need pay attention to, a minimal number. That’s where Occam of Occam’s Razor comes in. No need for the history, philosophy and/or theory lesson on Occam. Most people know Occam’s Razor as, “the simplest solution is often the best.” Never mind that’s a way-over simplification and is just a very loose and not entirely accurate rendering of the Razor, but let’s go with it.
Five metrics to manage the mortgage origination business is an Occam’s Razor approach. A minimalist, commonsense approach. William of Occam would approve.
The trouble is that getting to those five metrics, understanding how they interact and putting them to work takes data, a lot of data, even more analysis, some geeky statistics stuff and a deep understanding of what makes this business tick. And therein lies the paradox. We have a lot of data yet not a lot of clarity. That’s why I relate to Hickam’s Dictum.
Hickam was a doctor who said, “a patient is entitled to as many darned diseases as they want to have.” The medical profession has been debating Hickam v. Occam for decades. Should a diagnosis be based on a commonsense, minimalist approach or should every possible diagnosis be considered? The debate has been going on for decades and continues to this day. The argument still isn’t settled.
I am obviously starting a similar debate in the mortgage industry. I side with the Occamists. Please join me. By taking the minimalist approach to assessing and managing the origination business, real, long-term improvements are possible.
I’ll go even further, and I will show you that one metric in particular requires your constant and everyday attention. Manage effectively to that one particular metric and overall lending performance improves for the long-haul. But that’s for another post.
Back to where are we, exactly: At the end of yet another boom.
More than 11 million residential first mortgage loans closed last year. More than 10 million will close in 2021, according to the MBA’s projections. How many will close in 2022? 2023? Six million and change in each of the next two years. A forty percent drop in volumes in a single 12-month period carries with it frightening implications. Not to be a downer, but this is exactly what is staring us in the face.
But wait! You can’t make that doom and gloom prediction then sign off! What exactly are the five metrics? How do they interact? How do you put them to work? Future topics. And you’ll find them right here. Comments and thoughts are welcome in the meantime. Drop me an email at firstname.lastname@example.org.