by Tom Ruff of The Information Market
November was both very typical and very interesting. When we looked back at the historical data, most metrics landed exactly as we expected, but something odd was brewing with the median sales price. In our last issue of STAT, we were scratching our heads when we reviewed our prediction for prices:
“Our math on the PPI is projecting lower numbers than logically acceptable, leaning more on the model than the logical prediction. The ARMLS Pending Price Index projects a median sales price of $206,400 for November. Using more ARMLS secret sauce and my own logic, I predict a median of $210,000. We’ll see if our model holds up.” — STAT November 2015 (analysis of data for October 2015)
Our median sales prices landed at $209,000, which confirmed that something went haywire in our model. Conversely, our secret sauce prediction was off by a tiny 0.5% difference. What is in the ARMLS secret sauce that is missing in our model? TRID!
The data in November points out a temporary disruption in volume and the median sales price. This was a surprise as the conventional wisdom found through various Facebook groups was that a few closings were taking longer, but the overall consensus was much ado about nothing. In reality, we believe TRID has pushed back several hundred closings causing lower than expected sales volumes as well as a decline in the median sales price.
Just like measuring the weather, many of our metrics are daily. Two of these daily metrics involve the median sales price for pending home sale contracts as well as monthly homes sold. It’s a little bit counter intuitive but the daily measurements of the median sales price of all pending contracts will trend slightly below the daily measurements of the monthly closed contracts, as evidenced in the chart below:
As you can see by the chart, the gap between the monthly median sales price and the median sales price for all pending contracts narrows in October with the lines crossing in November. The disruption in these numbers can be attributed to TRID.
Why? TRID only affects purchases with a mortgage, in some instances delaying them. This causes a temporarily higher cash to mortgage ratio. Public records for November in Maricopa County tell us the price of a median resale home where the buyer paid cash was $173,000 while the median price for a resale home where the buyer obtained a mortgage was $215,000. This negative pricing pressure from the increase in cash sales in the cash to mortgage ratio accounted for the “lower numbers than logically acceptable” drop we talked about last month in the PPI. This will be a temporary disruption with numbers returning to their normal patterns.
If our hypothesis is correct we should expect to see the percentage of cash sales increase and they did. Here’s where it gets tricky! There are seasonal patterns that also affect the percent-age of cash sales, however, the chart below shows that the percentage increase this year is clearly higher than normal seasonal trends.
Percentage of Cash Sales
ARMLS Pending Price Index (PPI)
We expect the lagging sales from November to record in December as the TRID effect wears off. We predict sales volume will improve to 6,750 and the median sales price will land around $214,000 in December 2015.
Posted on April 22, 2016 at 5:12 pm by Tamara Weber