Saturday, May 29, 2010


Gregory Klump, chief economist of the Canadian Real Estate Association, is telling Canadians to relax when it comes to housing. Read the report found on Mike Fotiou's blog and then watch this video.

So just relax, OK?

The report argues the diversion in home prices versus income is part of a normal housing cycle that will unwind with stagnate prices and rising income. The chart used to back up this premise shows these variables over time. Initially, I was surprised the gap between them appeared small enough to support this theory.

It turns out that the income statistic used here hides the difference between these two variables over time and conveniently implies a less severe downturn. Between 1980 and 2005 the index for income increases from roughly 55 to 225. An increase of over 4 times is definitely not typical over this 25 year period characterized by flat real wages.

I suspect the data used here is wages and salary and supplement labour income from statscan which matches what is charted above (between Q1 1997 and 2009). This is total income of a growing labour force and therefore will increase faster than individual incomes over time.

A reasonable source would be full-time, full year median earnings because it represents typical earnings over time, not aggregate earnings. Total labour income also hides the fact that virtually all wage growth over 25 years has been from top earners. Median income has barely changed in real terms for 25 years, moving from from $41,348 in 1980 to $41,401 in 2005. I converted to nominal using CPI tables from statscan.

full time, full year median earnings (nominal)
1980 $17,003
1990 $29,878
2000 $36,058
2005 $41,401

After this adjustment median income between 1980 and 2005 increased 2.4 times, much less than the 4.2 times of total labour income. I have added this series to CREA's chart to illustrate the difference below.

What is astounding here is even after selecting this convenient data for the study he has the nerve to criticize others who have looked into this relationship.

Warnings of a U.S.-style home price correction in Canada are unfounded. Such warnings are typically based on a limited analysis of the historical relationship between national average home prices and income which ignores Canadian housing market cycle dynamics.

It doesn't matter how sophisticated your analysis is when your data sources are distorted!

Sunday, May 23, 2010

Easy money for Canadian Lenders

This post will take a look at some ways banks and lenders are taking in some easy money in Canada.

1. The 5-year fixed corral

Right now there is a fairly hefty premium built into 5-year fixed mortgages and with the new qualifying rules it is the only option available to marginal buyers. If buyers choose a variable mortgage or a short fixed term they have to qualify for the posted 5-year rate. This rate now appears to be arbitrarily high compared to 5-year government bonds and this will move buyers towards fixed mortgages.

Even the spread on the discount rates appears elevated. Lenders are moving marginal buyers into fixed rate mortgages and collecting wider margins for it.

See the chart below to compare the fixed 5-year mortgage to the 5-year government bond. Before the financial crisis (2000-2006) the posted mortgage averaged 2.42% higher than the benchmark bond. The premium has been above 3% for about a month.

Read the article from Canadian Mortgage Trends about this situation.

It makes one curious about the logic that went into the final decision. Is there a real threat of sustained 4.35%+ higher rates? Or did the powers that be set the bar overly high to herd people into 5-year fixed mortgages—which just happen to be more profitable?

Maybe the latter is just too cynical a thought…

My opinion is that the restrictions to variable mortgages is a good thing and should be applied to discount 5-year fixed mortgages to avoid this problem.

2. Prime rate premium remains elevated well after financial crisis subsides.

Must be nice to collect an additional 0.25% off indebted Canadians, which the banks are doing since the prime rate remains inflated long after the financial crisis faded away. Normally the prime rate tracked the Bank of Canada's overnight target rate with a 1.75% premium. However back in 2008 they decided to withhold part of the central bank's rate cut. See article: Canadian consumers shortchanged on rate cut.

"Continuing market turmoil has steadily driven up the cost of borrowing for financial institutions. This makes it challenging to match the Bank of Canada rate cut at this time,"

Tim Hockey, president and chief executive of Toronto-Dominion Bank, said.

TORONTO - Toronto-Dominion Bank (TSX:TD) continued a string of big-profit announcements from Canada's largest banks Thursday, posting first-quarter results that handily blew away analyst estimates for both earnings and revenue.

The bank's net income rose to $1.3 billion in the quarter, essentially doubling the $653 million it earned a year earlier, helped by record earnings from its domestic banking operations.

It joined CIBC (TSX:CM) and Bank of Montreal (TSX:BMO) in beating Bay Street estimates, with only Royal Bank (TSX:RY) falling slightly short of expectations even as it posted a $1.5-billion profit.

To illustrate this additional 0.25% premium see the chart below with the BOC rate, prime rate and difference between them. There were the occasional spikes resulting from a small delay to move the prime rate, but this is the first sustained difference going back 10 years.

3. The renewal trap.

Canadian mortgages are typically renewed every 5 years or less. With the recent mortgage changes to decrease amortization from 40 to 35 years, require 5% down and qualify on 5-year posted rate (not to mention future price decreases) there will be borrowers who fail to qualify at renewal. Since a change in lender requires requalification while renewal does not, borrowers have no choice but to renew at the rate given to them. Good deal for lenders who can set price arbitrarily. Another unintended consequence of loose lending. Again, from Canadian Mortgage Trends:
The kicker is that you can’t change lenders at renewal without requalifying. Therefore, if you don’t have 20% equity at maturity you could be stuck in another 5-year fixed mortgage (possibly at your existing lender’s “rack rate”). If you instead want to switch to a variable or 1-4 year fixed term, your debt ratios will have to fit under the much stricter government guidelines at that time.

Saturday, May 8, 2010

Payment Engineering

When buying a home one should keep focused on the actual price as opposed to the payments. This is especially true today with all the ways the they can be made to appear lower in advertisements. Consider this builder ad I found in the "Homes" section of the Edmonton Journal.

$611 bi-weekly? In the example below I will start with a fairly normal mortgage and work towards this engineered payments of $611. This will highlight how much impact different choices can have on the appearance of financing costs.

The listed price is $383,800. With a 10% down payment, 2% of that would go to CMHC fees and the total mortgage amount would be $353,096. Assume a 25 year amortization, 10% down and a 5 year fixed rate of 4.5% the mortgage payments would be $1,954.30/month

This does not include property taxes typically rolled into the mortgage payment. I would not expect a builder to include this but it is useful to note among other things when comparing to rent. These payments are for the first 5 years only, the remaining 20 will be paid back at a different rate. For example if interest rates are 7% during the last 20 years of the mortgage the monthly payments for that time period would be $2,384.91.

How do we get to $611?

1. Bi-weekly payments

Bi-weekly is a clever way of showing payments because initially people think of it as pretty much two payments a month. However on average throughout the year it is actually 2.17 payments per month. $1954.30/month works out to $977.15 semi-monthly and $901.99 bi-weekly.

2. 35-year amortization

Bi-weekly payments are reduced from $901.99 to $767.06 when the amortization is increased by 10 years. No free lunch here as in the end the longer amortization costs a lot more in total interest.

3. Downpayment

This example started with a 10% down payment but this ad used a 20% down, or $76,760. A buyer enticed by low bi-weekly payments over 35-years probably would not be putting down that much which is why I picked 10% for this example. Using a 20% downpayment reduced the payment to $667.01.

4. Interest rate

There was no mention of the interest rate but clearly they are using less than 4.5%. With a 3.84% rate instead I arrive at payments of $611. This rate is now only available with a shorter term, probably 3 years, which neglects the remaining 32 years of financing at unknown (higher) rates.

All the above serve to reduce the payment shown but each carries some risk or cost. No matter how the debt repayment is structured buyers are on the hook for the purchase price.

Update: This ad has been updated with payments of $599. They must have lowered the interest rate used in their calculations.

Tuesday, May 4, 2010

Edmonton stats - listings remain high

New listings in Edmonton are coming on the market at a substantially higher rate than the 2nd half of 2009. Clearly many of these sellers will be unsuccessful and the market will be forced to find a new equilibrium. The question is on the path to equilibrium will there be a self reinforcing negative cycle and how long it will last. Listings will tell this story during the second half of this year.

Sales were off to a slow start this year and are now following a seasonal trend albeit at a lower level than the last half of 2009.

The seasonally adjusted sales to new listing ratio remains below 50% from which we can expect price decreases.

Looking back at historical figures listings are similar those that lead to peak inventory levels of 2007 and 2008 while sales are relatively low.

Sunday, May 2, 2010

Calgary Stats and the Edmonton Journal is misleading again

Bob Truman updated Old Criteria stats for Calgary so it's time to look at seasonally adjusted data to guage where we are at.

Listings stayed at the same level both in both raw and seasonally adjusted terms after big jumps in February and March. At this level of listings inventory is still accumulating rapidly, reaching 10,078 in April. This is double the low of 5,018 in December and up from 8,688 this time last year.

Sales decreased slightly while normally we should expect to see a slight increase from March to April. Sales have maintained their position above the dismal pace set during the financial crisis, labeled scorched earth, but still below the bounce we saw in the second half of 2009. From this I expect YOY declines in sales to widen going into the second half of this year.

Seasonally adjusted there was a slight deterioration in the market compared to March, with the sales to new listing ratio dropping to 42%. This is the sixth consecutive month of falling sales/list ratio. A ratio under 50% leads to falling prices.

On another note - Here is an article I missed last week from the Edmonton Journal: Yes, there's a housing bubble in Canada -- but only in three cities

Kevin at the Edmonton Housing Bust Blog does a good job discrediting the article. Read it. The article argues that the relatively slow price appreciation in Alberta last year indicates a healthy market. However, it neglects the rapid price appreciation that occurred during the boom.

In addition to that I noticed that the Edmonton Journal makes a faulty comparison to U.S. bubble cities in terms of affordability.
When U.S. house prices got to their most extreme levels during the U.S. housing bubble, the ratio of average household income levels to average local house prices got to 10 times or more in overheated markets like Los Angeles and Phoenix.
Wrong. While Los Angeles did exceed 10, other ground zero housing bubble cities such as Phoenix were not even close. The 3rd annual Demographia affordability survey shows that in 2006 some of the cities which have seen spectacular collapses had price to income multiples well under 10.


Los Angeles-Orange County 11.4
Las Vegas 6.5
Fort Myers, Florida 5.3 (see video)
Phoenix 5.1

Los Angeles-Orange County 5.7
Phoenix 2.6
Las Vegas 2.4
Fort Myers, Florida 1.9

2009 (Canada)
Vancouver 9.3
Toronto 5.2
Calgary 4.6
Edmonton 4.1

Just because Alberta is below some of it's Canadian counterparts does not mean the market is healthy. The Edmonton Journal states that affordability levels are "reasonable" in Alberta even though the demographia survey labels the market as "seriously unaffordable" using the price to income measure.