Whither the Weather [Extended Blog Entry]

Whither the Weather
The weather is getting a lot of attention for driving the pending results in December. But how do we know that the weather was responsible for the slip in contract signings as opposed to other market forces? Is there a way to measure the weather and its impact on the home purchase process? The answer is yes, there are tools to help measure weather’s impact. In this article we’ll look at how weather can impede the home buying process, measures of the weather in recent months, and look ahead to how meaningful December’s data are for determining the future housing market trends. While the December pending home sales figures signal notably weaker home sales figures in January and February, the next few months of data will be better indicators of the 2014 housing market.

Home Buying: How the Weather Can Get in the Way
First, let’s think about how the home purchase transaction might be vulnerable to the weather. In spite of the way that the internet has improved the speed and accessibility of information in the home purchase transaction, home purchases remain the largest purchases most buyers will ever make, and as a result, an in-person visit to a potential home is a key step. In fact, three-quarters of buyers who used the internet to search for a home later drove by or viewed the home and more than three in five buyers who used the internet later walked through a home they first viewed online [1]. Surveys also show that buyers do not just look at the property they ultimately purchase. In 2013, the typical buyer who purchased a home visited between 10 homes before completing their home purchase [2].

A buyer’s visit to a home typically occurs before they decide to submit a contract—an offer to buy a home. Thus, adverse weather that keeps potential buyers from getting out to look at homes can have a particularly bad effect on our measure of pending home sales—the number of contracts submitted by buyers that have been accepted by sellers—a leading indicator of existing home sales.

Winter Weather: 2012 vs 2013 Storms and Temperature
So weather can get in the way of the home buying process. How do we know that the weather in 2013 was any different than normal? The Weather Channel began naming storms systematically in 2012. Looking at the number of named storms is one way of accounting for differences, though it’s worth keeping in mind that not all named storms produce an equal amount of disruption.

In 2012, there were 4 named storms in December: Caesar, Draco, Euclid, and Freyr. Most of these storms hit the West, Midwest and Northeast—areas relatively accustomed to snow. Euclid had the southernmost reach, dumping snow as far South as Dallas on Christmas Day and also spawning tornadoes. These storms were concentrated in the second half of December; only Caesar,which dumped snow on the Dakotas, Minnesota, and Iowa on December 8 and 9, occurred in the first half of the month.

By comparison, in 2013 there were 5 named storms in December: Cleon, Dion, Electra, Falco, and Gemini. Most of these storms hit early in the month with Cleon and Dion affecting large swaths of the country through December 10. Electra and Falco concentrated snow on the Midwest and Northeast through December 17 and Gemini caused more havoc with its ice from Michigan to New York through December 22. In a contrast to 2012, the holiday week from Christmas to New Year’s was relatively quiet. Looking ahead, January 2014 had 6 named storms compared with 7 for January 2013.

Another measure kept by NOAA called the Regional Snowfall Index and Societal Impacts shows that nationally, 2012 saw 4 December storms while 2013 saw only 3, but the 2012 storms were in the latter half of December whereas the 2013 storms occurred earlier in December. By region, the Rockies and Plains and Ohio Valley saw higher category impact storms in 2013 compared with 2012, and the Upper Midwest saw a greater societal impact in 2012 than 2013.

NOAA also keeps information on temperatures that necessitate heat use called Heating Degree Days. While December 2013 was nearly normal, with only 13 more heating degree days than typical, it was significantly colder than December 2012 which had 122 fewer heating degree days than normal. January shows a similar differential; January 2014 had 53 more heating degree days than normal while January 2013 had 90 fewer heating degree days than normal.

Seasonal Adjustment and Winter Sales
Housing data is seasonally adjusted in an attempt to smooth out the fluctuations we know exist from month to month. For example, many families with school-aged children choose to buy and sell homes during the spring and summer so as to minimize disruption to children during the school year. However, these fluctuations are not uniform and the efforts to smooth are based on estimates of typical monthly patterns.

If this January was significantly different from the last January, our estimate of the seasonal factors could be revised in the future. Typically the year to year variations in seasonal factors are small, but over the years seasonal factors on December pending home sales have varied. A review of past factors suggests that a seasonal revision to an extreme historical value could raise the December Pending Home Sales Index estimate for the US by as much as 5 percent or lower it by as much as 2 percent. The actual revision is likely to be only a percent or two higher and will occur with the release of revised December data next month.

Non-seasonal Potential Sources of Weakness
The housing market is entering an interesting year. After record price gains in 2013 and near-record gains in sales many factors could cause slower growth in sales and prices for 2014: implementation of Qualified Mortgage rules that could curtail access to financing, uncertainty over government policy such as tax rules for short sales and flood insurance rates, a relative affordability crunch as the rapid rise in prices has outpaced income growth making homes affordable but much less affordable than they had been in recent years. These factors could be offset by an improving economy generating jobs for potential homebuyers. However, if job creation slows, this momentum will carry over into housing.

Right now, NAR’s forecast calls for sales to be roughly the same in 2014 as in 2013—slightly below the record year. However, after months of payroll job gains near or above 200,000, December’s job gains figure was a surprise on the low end. The months ahead will reveal whether the weak payroll figures and pending home sales figures are seasonal blips, or a sign of a broader weakening.

[1] Both of these figures come from the 2013 Profile of Home Buyers and Sellers by the National Association of Realtors®.
[2] Ibid.