Welcome!

By registering with us, you'll be able to discuss, share and private message with other members of our community.

SignUp Now!

How do they calculate changes in home prices?

moniq

0
Registered
Joined
May 2, 2008
Messages
16
How do they calculate changes in home prices?

RICHARD BLACKWELL

January 13, 2009

Why do some statistics show house prices dropping sharply, while others show prices up?

The problem with home price statistics is that there are an infinite number of ways to calculate and compare them.

New Statistics Canada numbers released yesterday, for example, give an idea of what`s happening with newly-built homes (but not the resale of older homes). They show a 0.3-per-cent decrease in new-home prices in November, 2008, compared with the previous month, but a 0.7-per-cent increase compared with November of 2007.

That`s only the national average, however. In different markets there are wildly different numbers. In St. John`s, prices were up year over year by more than 25 per cent, while in Edmonton they were down almost 8 per cent.
Print Edition - Section Front

Section B Front Enlarge Image
The Globe and Mail

Is Statistics Canada the only group that compiles house price data?

No, and that`s another issue. Various organizations such as the Canadian Real Estate Association (CREA) and real estate brokerage Royal LePage also collect and publish regular sales numbers, and these don`t always agree.

Royal LePage often categorizes houses into different groups, such as detached bungalows or detached two-storey homes, so that adds another whole set of permutations. At the moment it is projecting a 3-per-cent drop in prices (nationally, on average) for 2009.

CREA`s numbers reflect all houses sold on the Multiple Listing Service (MLS), but again there are many ways to compare them. CREA`s most recent numbers, for example, show a 10-per-cent drop in average prices across the country between the month of November, 2007, and the same month in 2008. But its year-to-date numbers (which compare prices over 11 months in 2007 with 11 months of 2008), show only a 3.5-per-cent drop.

Are there any other important housing statistics?

Statscan also releases building permit numbers, while Canada Mortgage and Housing Corp. issues statistics on housing starts. Both of these give some idea of the health of the housing market.

If the assets in asset-backed commercial paper (ABCP) exist, why don`t the holders just cash them in? Many of the assets that back ABCP are mortgages and derivatives.

Those assets were chosen because they provide a continuous stream of income, which in turn flows through to the ABCP investors.

When the ABCP market froze, there was no point in selling off the property backing the mortgages or the assets backing the derivatives, as this would have locked in huge losses.

Making that move would have been analogous to selling off an income property that has lost value, even though it is still generating a good rental income.

There`s a further complication: Owners of ABCP don`t have first claim on most of the assets underlying the paper. A group of foreign banks that provided loans to buy many of the assets have first dibs, so in any liquidation owners of the paper would be left with almost nothing.

That`s why a group of the biggest ABCP investors, including National Bank of Canada and the Caisse de dépôt et placement du Québec, pushed for the creation of the Montreal Accord in August, 2007, to stop the liquidation of the assets and buy time to find another solution.Why do some stock prices on the TSX seem to change after the 4 p.m. (EST) close?

Finishing up trading at the end of the day isn`t as straightforward as it used to be, when the closing price of a stock was simply the last trade made before the markets shut down for the day.

In 2004 the TSX introduced a system called "market on close," which allows traders to place orders during the day that will be filled at the final closing price.

As part of this system, there are sometimes price adjustments that are not disclosed until 4:10 p.m., in order to balance the orders.

Do those changes affect all stocks on the TSX?

Only the stocks in the S&P/TSX composite index, the global gold index and the global mining index are subject to market-on-close or the 4:10 p.m. price recalculation. But that can mean that you won`t get a final price on a stock - or on the composite`s closing - until it is all over with. Also, many stock reporting websites have a 15-minute delay, so you might not see the final price until around 4:30 p.m.
 
This is one of the reasons that I always look very carefully at the actual sales numbers that come out from CREB every month rather than take the headline making trends as gospel.

CREB often compares sales stats using monthly snapshots taken 12 months apart. They will look at the average price in January 2007 and then also January 2008 and say that the market has risen or fallen depending on the change in metrics for these two intervals... be it sales or average prices, DOM or whatever.

Average sale price is much trickier to evaluate than it seems due to large variances in sample size (number of sales) that can be observed in each interval. For example, have a look at this graph for Calgary. Source data is from CREB however I put the graph together.

Citywide Sales Stats - last 24 months - Current to Dec 08

Each month has a number for average price... but that number is a function of of the sales observed in that month. Let`s look at December 2007. Average price is 413k, total sales are 1508. December 2008 saw an average price of 373k, with 806 sales.

At first glance it might seem that prices have plunged by 1-(413/373) or -10.7%... but look at the sales numbers... there are half as many sales in Dec `08 as there were in Dec `07. Now the reasoning behind that is a different matter altogether but for the purpose of evaluating how the market is doing I personally like to have equal sample sizes. People say that when sales are low it means there are no buyers but it can just as easily mean there are no sellers willing to accept offers from a large pool of buyers.

What I like to do is use what`s called a "Rolling Weighted Average". This method puts each value (monthly average price) on an equal playing field by assigning it a weighting commensurate with the number of sales in that month relative to the number of sales within the number of intervals (months) you want to include in your analysis. I have chosen six months as an appropriate time period I think it takes to view trends in a real estate market while also giving enough months to smooth out variability in sales volume.

So, when you compute the rolling average for prices you see that the rolling weighted average for December `07 was 431k. the RWA for December 2008 was 402k... for a change in 1-(431/402) = -7.4%. Still down, but not quite as bad as the spot average.

The weighted average technique is quite common in commodities trading. It is also much more useful for the analysis of specific communities where there aren`t as many sales and some smoothing is required. Have a look at this graph to see what I mean - notice how the actual monthly average has a lot of variablity (looks like it`s tanking) but when you smooth it by sample size it gives a much better idea of how the local market has been performing. (staying flat)

Ranchlands Single Family Home Dec 08
 
It is also important to remember that the numbers quoted are AVERAGE. There will be some above those number, as well as others below it.

And you need to know what is included to make up that average. Used homes seem to be selling better than new homes in Vancouver. Sales of new condos have all but tanked here. If you combine all of those numbers to work out an average, you figure may be very skewed. For that reason, Vancouver REB breaks down it sales numbers into different groups.

Example... if you have three sales, one detached home @ $1,000,000 and two condos @ $100,000, the overall average price will be $400,000. Clearly, in this simple example, the average is not an accurate measurement.

Brent said it well; you need to look "behind the curtain".
 
I totally agree with Brett and Dan. Thanking very much for taking the time to post your comments.

I sometimes wish that CREB does not publicize those numbers.... It confuses the majority of people I am dealing with.

Numbers do not lie that`s what they believe and it is very hard sometimes to explain it to them to look "behind the curtain" and study them carefully.

My point was that it is good to have some basic statistics in real estate but I find most of them are really redundant and too confusing.

Statistics just gives you an idea (a starter point) what the market is at the moment but do your diligence when investing a specific property in a specific neighborhood. In the end the result will maybe totally surprise you.
 
Back
Top Bottom