Monday, July 18, 2011

Are Prices Going Up or Down?

Article Courtesy of Michael Orr
The Cromford Report

July 16, 2011

I don't think we have ever had such a confusing set of pricing data so it's not surprising that opinions on price direction are all over the map. Even the facts are all over the map, which is a much rarer condition.

I'm going to wade into the issue and try to explain what I see going on. Hopefully it will shed light rather than adding to the confusion.

Before I start, let me make it clear that market direction and price direction are not the same thing. Sometimes the market gets worse while prices rise. This happened between April 2005 and August 2007. Throughout this period the market deteriorated - supply was increasing and demand falling and although prices reached a peak in June 2006 they did not start to decline significantly until September 2007 (see here). In contrast, since November 2010, when the market was in poor shape, supply has been falling while demand has been rising, but prices have shown little to no interest in responding by moving higher. 

Nevertheless the improvement in market conditions is very substantial. Let's look at some key indicators for all areas & types:
  1. The Cromford Market Index™ is higher now than at any time since October 2005.
  2. The inventory of active listings is lower now than during the buoyant market of July 2003 and a far larger proportion of the current inventory has a contingent contract than was the case in 2003.
  3. Monthly sales rates are setting new record highs.
  4. Recent contract ratios are in the low to mid 90's, the highest we have seen since September 2005.
  5. Listing success rates are approaching 75%, the highest recorded since October 2005.
  6. Pending foreclosures and REO inventories are declining faster than at any time since 2005.
  7. Foreclosures are off to a very slow start in the first week of July.
I could go on...
But let's get back to prices. The first thing to point out is that much depends how you measure them. First we have to choose between averages and medians. Why do we have two different measures? Because neither is entirely satisfactory. Average price, which you obtain by adding all the money spent on homes and dividing by the number of homes sold, can sometimes be unduly influenced by a relatively small number of large and expensive luxury homes. So instead, analysts often refer to the median sales price which is hardly affected by luxury homes at all. The median sales price is the price of the home that sits half-way between the lowest and the highest priced homes when you arrange them in an ordered list by price. You can add fifty multi-million dollar luxury homes and the median sales price won't move much, and possibly not at all. The disadvantage of the median sales price is that if you have a huge number of extremely cheap homes being snapped up by bargain hunters, it can drop to alarmingly low levels completely unrelated to the value of the bulk of the houses in the area. This is what happened in the first half of 2009 when investors were snapping up low-priced REOs in West Phoenix so fast that the median sales price for the City of Phoenix fell to $60,000 in March 2009 from $95,000 in January. By October 2009 it had recovered to $100,000. So while the median is insensitive to the high end it is sometimes over-sensitive to bargain buying at the bottom of the market.

It follows that we have to treat both average and median sales prices with a good deal of caution. At the Cromford Report we tend to focus on average sales price per sq. ft. which dilutes the disproportionate effect of luxury homes a little, but certainly not completely. It also goes a long way to avoid the problems associated with medians when REOs are selling like hot cakes. However $/SF is still highly susceptible to changes in the market mix, or "blend" as Tom Ruff likes to call it. This is especially true of changes in the blend at the high end of the market.

Today the average $/SF for monthly sales across all areas & types fell to a new extreme low of $80.32. We know it was $84.11 on June 9, so that's a big drop of 4.5% in a single month. Yet the monthly median sales price went up by 1.8% over the same period.   

Apparently prices are going up and down at the same time. Hitting a new bottom, at the same time as rising. Can this be true? You see what I mean about the facts being all over the map?

The facts are not really to blame. We must remember that averages and medians are not only different, they are not even closely related and often move in different directions, especially over short periods like months. It is only over the long term that they tend to follow similar trends. 
 
To find out what has really happened, let's break down sales into 5 broad price ranges to see if we can detect what changed in the 30 days between June 9 and July 9.

List Price Range Sales per Month June 9 Sales per Month July 9 % Change in Sales per month Monthly Avg $/SF June 9 Monthly Avg $/SF July 9 % Change Monthly Avg $/SF
Under $100K 4,488 4,539 +1.1% $43.48 $42.88 -1.4%
$100K to $200K 3,325 3,560 +7.1% $71.48 $71.14 -0.5%
$200K to $400K 1,547 1,607 +3.9% $104.28 $102.63 -1.6%
$400K to $800K 406 446 +9.8% $154.00 $154.39 +0.2%
Over $800K 154 121 -22.4% $279.92 $238.51 -14.8%

Voila! The problem jumps out at us. It's in the range over $800,000.
The average $/SF in the price range above $800,000 fell nearly 15%. This is very unusual. Sales volume also fell over 22%. The average sales price for homes listed over $800,000 fell nearly 20%. This drop in volume and price from the luxury home sector had a huge effect on the overall average sales price and the average price per sq. ft, but it had no effect on the median sales price. Remember that the median sales price is essentially immune to any changes in the luxury sector. In fact the median sales price improved because of the growth in the price ranges between $100K and $800K.
Let's home in on the price ranges over $800,000 and see if we can identify more detail about the price changes.

List Price Range Sales per Month June 9 Sales per Month July 9 % Change in Sales per month Monthly Avg $/SF June 9 Monthly Avg $/SF July 9 % Change Monthly Avg $/SF
$800K to $1M 44 49 +11.4% $192.30 $184.16 -4.2%
$1M to $1.5M 51 37 -27.4% $231.18 $236.26 +2.2%
$1.5M to $2M 25 24 -4.0% $281.38 $243.26 -13.5%
$2M to $3M 18 8 -55.4% $356.33 $365.48 +2.6%
Over $3M 16 3 -81.2% $428.08 $451.17 +5.4%

Now we can see that average pricing for luxury homes did NOT really fall by 15%. In fact the monthly average $/SF in 3 of the 5 sectors went up. The main cause of the large fall in average $/SF was the huge drop off in sales volumes over $2M - down from 34 to 11. Buyers of high end luxury homes had been active during the spring and seemed to have stopped buying as soon as the hot weather set in. The sudden absence of 23 sales over $2 million has a huge effect on average price and $/SF measurements, not just on the luxury sector, but on the market as a whole. 

So where are we really?

Well from my perspective, prices aren't really going up or down much at all. They are still bumping along in essentially the same price range they have been in since October 2010. In situations like this, changes in the blend can make all the difference in whether the average or median goes up, down or stays flat. Our advice is not to pay too much attention to these movements. They will gain their rightful insignificance over time.

When a real price trend forms, we will know about it for sure, because all the price indicators will then be moving in the same direction with obvious momentum.  And when that happens you will hear about it here first.


Note: By the way, you may have noticed that we did not use median sales prices in reviewing price ranges. You may also be wondering why. It's because median sales prices offer no useful information at all when applied to a price range. The price range has already been limited by definition so all you will discover is that the median sales price is pretty much in the middle of each selected price range. It will stay in roughly the same spot no matter what happens to the market. Hence our advice is: don't waste any time calculating medians when dealing with price ranges. Averages are not much better. You really need to study the average price per sq. ft. (or median price per sq. ft.) if you want to derive useful pricing information about price ranges. An alternative approach is to create ranges by living space sq. ft. instead - if you do this then median price and average price become meaningful measures again.

No comments: