The strength of the housing market continues to make headlines. The April 24 headline of an article by CNBC real estate analyst Diana Olick says it all: Spring Housing – ‘Strongest Seller’s Market Ever’. A week earlier, Bloomberg reported that “sales of previously owned U.S. homes rose more than forecast in March to the fastest pace in a decade.”
Yet, amidst all the talk of a robust market, a story in the Denver Post on April 16 headlined “Denver, Front Range real estate party ends in late 2019” may have garnered more attention. Read Full Article Here.
The article cited a study by Location Inc., a real estate research firm based in Connecticut, warning “that years of rapid appreciation (along the Front Range) will shift into a stretch of falling prices starting in late 2019. More specifically, the forecast says, median home prices will fall by more than 20 percent during the following few years, due in large part to a growing gap in wages and home prices.”
That’s a pretty bold prediction. Certainly bold enough to generate some publicity for the firm who sells software products to real estate investors. But is it accurate? Should it be used as a basis for important life decisions, such as buying or selling a home?
At 8z, we are real estate advisors who look at things from every angle to help our clients make sound decisions. We are not simply cheerleaders for the real estate market who think “it’s ALWAYS a good time for everyone to buy or sell.”
We thought it would be helpful to take a closer look at the study in this month’s newsletter. In fact, our CEO Lane Hornung took on the project himself. Lane has an engineering degree from Stanford along with an MBA in Finance from CU. He’s no stranger to statistical modeling and predictive analytics.
According to Location Inc.’s website, the model is “built upon 200+ independent variables, 35+ unique data dimensions, and 20+ custom, geographically nested, hierarchical models to predict home prices in micro areas down to the neighborhood level.”
Peeling it back a bit more, the model is largely based on the relationship between home prices and people’s ability to pay those prices, predominantly determined by per capita income and prevailing interest rates. Sound familiar? That’s the 3-legged stool, jobs, listing inventory, and interest rates. Not as fancy as “hierarchical models,” but based on the same fundamentally sound concepts.
Like any statistical model, the output is only as good as the input. “Garbage in, garbage out” as the saying goes. Location Inc. offers few details (actually none) about the assumptions used to predict future income growth and interest rates.
There is no question that incomes must grow to sustain home price appreciation. That said, it is unclear whether Location Inc.’s model factored in the impact of folks moving to Colorado and bringing their high paying jobs, or businesses, with them. In-migration of high earners can cause per capita income to grow rapidly.
The 60,000 plus people moving to our state each year play a large part in sustaining appreciation when you consider that our prices are still relatively cheap for folks coming from the coasts, that only 90,000 existing homes sell along the Front Range each year, and these folks have to live somewhere.
Location Inc.’s model may have also underestimated per capita income growth for current Colorado residents in the next 5 years, based on paltry income growth in the last decade. The trend of flat incomes appears to be shifting. It was widely reported last week that unemployment in Colorado in March dropped to 2.6%, the lowest in the US and the lowest since the state began collecting unemployment data in 1976. Income growth should follow.
Another way to evaluate a model is to “backwards” test it to see how the model performed for earlier time periods for which the outcome is known. Location Inc. says they “tested its analysis on the most recent housing downturn, in 2006-11, correctly predicting that Los Angeles and Miami home prices would be hard hit…and the model also closely tracked with home prices from 2000 to 2016 in northern Colorado.”
However, their prediction that Boulder home prices will be the hardest hit in the next five years casts some doubt on how well their model would have performed in our local markets in the downturn. Specifically, Location Inc. predicts that Boulder is the most at risk and will see prices drop by 17%, with some neighborhoods seeing values cut by 33%, from now to 2022. That simply does not jive with historical performance in which Boulder home prices dropped less than 3% despite the nation experiencing the worst housing downturn since the Great Depression.
Considering prices are already on track to increase at least 5% in Boulder this year, a 20% cumulative drop would be an unprecedented reversal. Location Inc. does not disclose details of their backwards testing, but if their Boulder market predictions seem to run counter to known market dynamics, it’s fair to wonder if their projections for other Front Range markets are off.
In summary, we agree with the basic premise of the model that appreciation rates cannot continue at today’s torrid double-digit pace forever. It is highly unlikely that incomes will grow that fast, even with in-migration of high earners.
But it is a whole other thing to predict that prices will fall by double digits. Of course, it’s possible. It’s just not likely. We’d give it a 10% chance.
The more likely scenario is that appreciation rates will settle down to historically typical rates of 2-5%. Home values may even dip a percentage point or two as the market flattens out at this new high plateau and another cycle begins anew. Of course, that will come after prices rise another 5-10% in 2017. We’d give that increase a 90% chance.
I’m happy to talk through your specific situation. We can come up with scenarios and probabilities tailored to you. Most importantly, we will factor in your life needs and wants, not just the dollars and cents. Timing the market is a bit of a crap shoot. Timing your life decisions is not.
A Refreshing Look at the Question “What is my House Worth?”
Homes are hitting the market at a rapid pace. Let’s take a look at some of the stats for our area to get a better idea of what is going on.
In Boulder County for March 2017, the average sales price* was:
- $531,300 for Single Family Homes (up 8.87% from 1 year ago)
- $345,000 for Condos/Townhomes (up 7.8% from 1 year ago).
The graphic above shows that on average a buyer paid 100% of what the seller listed the home for last month. Also with low months of inventory, this keeps us solidly in a seller’s market for now until more homeowners decide to list their homes.
I have access to detailed stats across Colorado and can help you find out the worth of your property any time. I can also help you determine what your home is worth even if it’s in a different area. As always, I am here for you. If it’s time for you to buy or sell, let’s talk.
*Median sales price based on a six-month moving average