Baby, we got a bubble!?

[interested]

Not really. Why would someone choose to pay a $40k CMHC premium if they had the extra downpayment available? Possible, but unlikely.



Yes, a low arrears rate is always good, but lets remember that the arrears rate in the US was also low as well prior to 2006. And it quite common to have a real estate correction without a high arrears rate. Canada only hit a 0.6% arrears rate during the early 90's correction.
http://www.cba.ca/contents/files/statistics/stat_mortgage_db050_en.pdf



This means that 8% of CMHC mortgages are underwater if they chose to sell (after sales costs). Consider that only 4% of the properties in the country are sold each year. Housing prices derive from this marginal sales activity, not from the average of all the properties in a market.



Agreed that very few CMHC mortgages are $2m+, although so too are very few houses at $2m+. Also, the incidence of $500k mortgages would be significantly higher in Toronto than Canada-wide.



CMHC is insurance, and serves to transfer risk from the banks to the taxpayer in exchange for a premium.

Supporters of CMHC's growth point out that it has returned $16b of profit to the taxpayer over the past 10 years. This is indeed admirable. But let's remember that AIG produced $100b of profit in the 10 years prior to 2008, when it then lost $100b in a single year. Such is the nature of insurance. When the black swan catastrophe hits, you lose a lot of money.

UNT, I don't wish to seem nitpicky in my comments on your points, and I appreciate you providing links to support your data. But I think that assessing the true risks here is sort of like identifying the size of an iceberg - its important to look below the surface. In our case, if banks are significantly less willing to lend mortgage money without taxpayer- backed insurance, then it is important to consider, why?

Great post Daveto. Agree totally. I think however that we are not that far off in our conclusions from UNT on most issues though clearly you and I identify a much greater risk to CMHC than UNT implies.
However, I think his point that the banks do download certain higher mortgages to have more lending room is probably correct. However, it opens the door to "moral hazard" and I have to wonder if the logic is the banks want to loan more than the money they have to backstop it with, this suggests above average risk taking. I don't think it is a quantum leap to assume that if the bank wishes to lend more than they really have money to do so that they may in the same breath turn somewhat of a blind eye to less than pristine mortgages since it will be off their books anyway.

 

[redfirm]

Can't find any posts by Condo George lately ... I'd like to hear his assessment of our current situation. What to expect over the next several months? And more importantly, what happens comes April/May next year?

 

[Riverdale Rink Rat]

CMHC is insurance, and serves to transfer risk from the banks to the taxpayer in exchange for a premium.

Supporters of CMHC's growth point out that it has returned $16b of profit to the taxpayer over the past 10 years. This is indeed admirable. But let's remember that AIG produced $100b of profit in the 10 years prior to 2008, when it then lost $100b in a single year. Such is the nature of insurance. When the black swan catastrophe hits, you lose a lot of money.

UNT, I don't wish to seem nitpicky in my comments on your points, and I appreciate you providing links to support your data. But I think that assessing the true risks here is sort of like identifying the size of an iceberg - its important to look below the surface. In our case, if banks are significantly less willing to lend mortgage money without taxpayer- backed insurance, then it is important to consider, why?

Dave -- I don't disagree with your overall assessment, but you can't use AIG as an example. The were done in by their derivatives group, not plain vanilla insurance. Also, of course banks are less willing to lend if they're not insured against loss. It's much more costly to do due diligence on credit risk than insure a mortgage. That's why CMHC's rules got tightened in the first place.

IMHO, this is not a Black Swan at all -- it's been forecasted and telegraphed for years and years. You cannot tell me that smart guys like Ed Clark and Al Jette at TD haven't modelled these things over and over in their ALM and Credit Risk departments.

 

[interested]

Dave -- I don't disagree with your overall assessment, but you can't use AIG as an example. The were done in by their derivatives group, not plain vanilla insurance. Also, of course banks are less willing to lend if they're not insured against loss. It's much more costly to do due diligence on credit risk than insure a mortgage. That's why CMHC's rules got tightened in the first place.

IMHO, this is not a Black Swan at all -- it's been forecasted and telegraphed for years and years. You cannot tell me that smart guys like Ed Clark and Al Jette at TD haven't modelled these things over and over in their ALM and Credit Risk departments.

One would hope that TD and our banks have done this:
A few rough numbers for thought though.
Mortgages by CMHC: Limit is 600 billion. Let's assume they are at about 575 billion but use 6 billion for calculations as that is the ceiling and they are rapidly approaching it(and it is simpler math for me).
Default rate so far has been covered of less than 0.5% and profit over 10 years is 16 billion and let's assume it is all there for use.
If defaults just went to 2% assuming the premiums covered the 0.5% and yielded 1.6 billion of profit/year; the residual 1.5% is 9 billion and this exhausts 7.4 billion (9 billion-1.6 billion this year) of the 10 year cushion.

Double that default rate to 4% and that leaves 3.5% (over the 0.5%)x 6 billion or 21 billion -1.6 billion for this year or 19.4 billion and now the 10 year cushion is exhausted and CMHC is 3.4 billion in the hole.

I realize this may be the definition of a black swan event but it is not such a large number (4%) and please remember the subsequent cascade of events that occur.
Properties drop 10%...say defaults go to 1% from 0.5%...that is OK but the problem is with a cascade when the neighbour now who had 20% equity now has only 10% and the one with 10% now has zero this starts the downward spiral of property prices and it is presumably those at the margin who will be forced to sell further influencing prices. In this scenario, psychology takes over and the downward spiral becomes a self fulfilling prophecy.

 

[bmiller]

The Bank of Canada is now sending out warnings about a surplus in the market and potential problems, yet real estate people see a drop, but nothing to fear.
Is it fear mongering? This could drive away potential investors and make the sales drop even more dramatic.
http://www.ctvnews.ca/video?playlistId=1.1069900

 

[rbt]

Properties drop 10%...say defaults go to 1% from 0.5%..

A price drop will not increase the rate of default. You can't turn in your keys to the bank and walk away in Canada. Banks have all kinds of ways of helping you cover a balloon payment when your term rolls over if you have income.

A signifiantly increased default rate would be the result of mass layoffs and/or a broad salary cut.

 

[interested]

A price drop will not increase the rate of default. You can't turn in your keys to the bank and walk away in Canada. Banks have all kinds of ways of helping you cover a balloon payment when your term rolls over if you have income.

A signifiantly increased default rate would be the result of mass layoffs and/or a broad salary cut.

I was thinking a bit longer term. Prices drop,(presumably this will occur because the economy is not doing well otherwise I do not expect a significant drop other than in certain select markets if we find out all the foreigners want to pull out at once from the condo market and they end up being a higher percentage of that market than people are guessing) people may have negative equity, others who don't want to buy because they don't want to purchase a declining asset hold off...people can no longer make their payments (because some people are losing jobs which further affects the overall psyche) or alternatively interest rates rise, new buyers can't get the down payments needed and this results in more jobs being lost in construction which percolates though the economy and people stop buying a depreciating asset.

I do appreciate however it will take a more general trend/turn and significant event for this to occur. I am just saying that if prices drop it may become a real snowball. I was not trying to imply that masses would walk away as you correctly point out they are still responsible financially. However, that does not preclude this from occurring if prices significantly drop

 

[cdr108]

One would hope that TD and our banks have done this:
A few rough numbers for thought though.
Mortgages by CMHC: Limit is 600 billion. Let's assume they are at about 575 billion but use 6 billion for calculations as that is the ceiling and they are rapidly approaching it(and it is simpler math for me).
Default rate so far has been covered of less than 0.5% and profit over 10 years is 16 billion and let's assume it is all there for use.
If defaults just went to 2% assuming the premiums covered the 0.5% and yielded 1.6 billion of profit/year; the residual 1.5% is 9 billion and this exhausts 7.4 billion (9 billion-1.6 billion this year) of the 10 year cushion.

Double that default rate to 4% and that leaves 3.5% (over the 0.5%)x 6 billion or 21 billion -1.6 billion for this year or 19.4 billion and now the 10 year cushion is exhausted and CMHC is 3.4 billion in the hole.

I realize this may be the definition of a black swan event but it is not such a large number (4%) and please remember the subsequent cascade of events that occur.
Properties drop 10%...say defaults go to 1% from 0.5%...that is OK but the problem is with a cascade when the neighbour now who had 20% equity now has only 10% and the one with 10% now has zero this starts the downward spiral of property prices and it is presumably those at the margin who will be forced to sell further influencing prices. In this scenario, psychology takes over and the downward spiral becomes a self fulfilling prophecy.

sorry ... maybe it's because i have a headache but i can't follow the math. could you please clairify?

 

[interested]

I will try cdr but please understand in doing this I realize the numbers are just rough approximations. Bear with me: you are much better with numbers than I am.
I assumed CMHC has $600 billion outstanding:
I understand the default rate is less than 0.5% and I also assumed that given that there is $16 billion of profit that it was earned straight line at $1.6 billion / year based on $600 billion. In fact CMHC is more profitable because they made this money when they had less than $600 billion so probably the latest earnings are closer to $3 billion/year and a default rate of less than 0.5% as is now based on $600 billion/year. This will improve the numbers somewhat.

Assuming that the default rate went to 2%: the correct way to do this would be to assume 2% default would result in 2% x $600 billion or $12 billion of loss. However, we know that currently even with about 0.5% default, CMHC makes money...with the above assumptions $1.6 billion/year ( or perhaps we should put in $3 billion with the above latter assumption ).

So this to me would suggest that 0.5% default which would be $3 billion dollar loss still resulted in 1.6 billion of profit or possibly 3 billion profit; then up to 1% or $6 billion CMHC would be in the black. However when we get to 2% loss which would be $12 billion -$3 billion for the 0.5% -the 1.6 billion profit (or - $3 billion with the latter assumption) the profit would evaporate and one would be left with with a loss of 7.4 (12-3-1.6) billion or possibly $6 (12-3-3) billion.

I hope this makes sense. Again it is very broadly based numbers with assumptions but it just demonstrates to me that CMHC can withstand on an operating basis no more than 1% total loss on their portfolio before they dip negative and at 2% they will exhaust their 16 billion cushion in 2-3 years.

I apologize but this is as clear as I can make it. It tells you 2 things....I am not great with math and second I am certainly not going to make it as a math teacher because if you can't follow this and I know from past posts you are good in math, I surely can't be hired as a teacher even if you have a headache as a partial excuse.

 

[Riverdale Rink Rat]

Interested - You are assuming a HUGE jump in defaults - 4x as many insured mortgages! That is a very bleak scenario. IMHO you're really projecting too much of the US experience onto Canada. The biggest offloading of risk was CMHC allowing portfolio insurance and that has been stopped. As the portfolio-insured mortgages mature, risk to the gov't will drop dramatically.

A drop in Canadian house prices does not bode well for those that purchased at the top, but it does not mean Canada in general will have a disaster on its hands.

 

[interested]

Interested - You are assuming a HUGE jump in defaults - 4x as many insured mortgages! That is a very bleak scenario. IMHO you're really projecting too much of the US experience onto Canada. The biggest offloading of risk was CMHC allowing portfolio insurance and that has been stopped. As the portfolio-insured mortgages mature, risk to the gov't will drop dramatically.

A drop in Canadian house prices does not bode well for those that purchased at the top, but it does not mean Canada in general will have a disaster on its hands.

I don't think this is a huge assumption at all. I do agree that you are right about those who purchased at a top not meaning there will be a disaster. However, since many of us are talking mainly about the TO condo market and in C01 in particular (downtown) I think it might be.

I don't recall the numbers but RRR recently there were surveys conducted with people asked how they could handle a 1%, 2% and beyond interest rate hike on their mortgage and a fair number, I don't remember how many, said they could not manage it but it was quite significant. Historic rates would be closer to 3-4% higher than present level and somehow regression back to the mean seems to be a recurrent theme over long periods of time though we have had a continuous 30 year decline essentially in interest rates. This trend has definitely run its course.

My fear is when interest rates turn around, and they eventually will, or at least we better hope they do because if they don't it will be because we are going down the road of Japan and interest rates stay low there but their economy is fighting deflation constantly...or the interest rates rise because things are going better. My fear is that the rise could be quite rapid, especially if my concern and others is that the low monetary policy which is very inflationary has to be reigned in. I won't even talk about stagflation where interest rates go up to fight inflation and jobs disappear due to high inflation and peoples incomes just stagnate. I could easily imagine a quadrupling if this happens.

I am not talking at all about the US experience though to think that can't happen here I think is not the case. Yes, I appreciate we don't have the NINJA loans, in the most inflated markets (other than TO condos and Vancouver as examples) prices tripled over 5-6 year whereas here they went up to double over a decade or more, and we have a better economy in TO than say a Phoenix. I understand all that but I do have concerns that people are complacent because interest rates have been down for so long and we have a generation that for 15 years in Toronto at least has only seen a steady price escalation. I was speaking recently to someone who bought condos over the past decade. He is a multimillionaire having about 15 mill of properties...rented out. He takes out his equity and has from 20-25% in all his properties. He has a good job as well. However, a 20% price adjustment would almost wipe out his almost 4 million of equity. He may/may not be able to hold on to all the properties...I don't know. And he is well capitalized and has a good job. However, what if rents drop....lots of properties and desperate landlords...what happens if there is job loss....or if the excess liquidity which drove up the prices is mopped up because the bond market decides en masse it will not give its money for such low rates?


I still remember in 1989 an individual who was sitting on 22 houses at various stages of construction(and there the timeline was 1 year, not the 3-5 years with PRECON condos)... multimillionaire on paper as he would flip each one and buy more. When the market stopped in 1989...it was February I remember it well....suddenly problems. He sold/lost a number of them...and went bankrupt. By the way he has done the same thing again and is a multimillionaire again though I suspect he is somewhat more cautious. The question is how many big time speculators are there...how many stretched individuals are there....and you see from the US experience that once the psychology changes, no one can/wanted to buy anything but at severe discounts and it was for a protracted time frame...4-5 years. After 1989 in Toronto it was 7 years. Real estate cylces can be quite long. In 2008 Toronto had a 15% drop in 6 months. The fact that it recovered totally was fortunate but the Canadian economy and the consumer are not as well off as they were back in 2008-2009.

In conclusion RRR, I do worry because I guess I am concerned about how any of the above events, let alone ones I haven't mentioned or thought about...could go wrong here.

So in conclusion...4x is a huge jump...but it would not take more than a normalizing of interest rates to create it in my estimation. I can't back this up with figures...sorry daveto that I can't reference it. I am expressing a feeling or guestimate and I readily acknowledge I may be out to lunch.

 

[CN Tower]

A drop in Canadian house prices does not bode well for those that purchased at the top, but it does not mean Canada in general will have a disaster on its hands.

It seems the minds at the BOC believe in the contrary:


Housing Market: Bank Of Canada Outlines Scenario For How Home Sales Slump Could Drag Down Economy

http://www.huffingtonpost.ca/2012/1...ada_n_2251445.html?utm_hp_ref=canada-business

 

[daveto]

CMHC Reserves

I'm an actuary and insurance executive, but my comments here below are based upon only a quick read of the CMHC report. The CMHC has two main types of "reserve".

$1b - Provision for Claims
$7b - Unearned Premium Reserve

See page 41 (and 61/62)
http://www.cmhc-schl.gc.ca/en/corp/about/core/upload/CMHC_2012_Q3_QFR_EN.pdf

Think of it this way. A homeowner pays a $10k premium to insure a $500k 25 year mortgage. The CMHC apportions that $10k over the 25 year period in proportion to its assessment of the underlying risk over the 25 years. So after the 1st year, it might have $1k as "earned" and $9k as part of the unearned premium reserve. The 100% $10k premium is included in income in the year paid, but a portion is offset by the $9k Unearned Premium Reserve (UPR) and released over subsequent years.

To accurately discuss the effect of an increase in defaults, we'd need to know

-how much of the $600b face amount exposure is from premium earned (ie reserved by the $1b)
-how much of the $600b face amount exposure is from premium unearned (ie reserved by the $7b)

Also, we'd need to know what are the assumptions for spreading the risk over the life of the mortgage.

And if we're in an environment with a lot of mortgages ending early (and thus releasing the CMHC from its insurance guarantee) then that is good for the CMHC bottom line. So we'd also need to know what assumptions the CMHC has for mortgages ending early vs going the full term, and how that might increase the CMHC exposure if we enter into a period of significantly lower home resales/moves.

Something else to remember is that any CMHC claim paid would be lessened by any recovery from selling the defaulted house to the market, etc. So its not like the CMHC pays 100% of the mortgage amount as soon as someone defaults.

How large a default spike would cause losses?
Something else to understand is that the full $8b is not currently available to pay claims. Only $1b is available to pay claims, and the other $7b is for the future over the duration of the mortgages which the premium insures. So if we had a modest spike in defaults (less than 1%), that could quickly cut through the $1b. It would also likely require their actuaries to increase both their reserves to reflect updated assumptions for their future claims

As a real estate bear, my personal opinion is that we will see losses at the CMHC, and probably in the range of a total of $10b-$15b over the next 5 years before the correction is finalized. But probably nothing until 2014. And I don't think the default spike will be very high, probably it will stay below 1%.

I think the bigger picture of the effect of a real estate price correction is not a possible loss to the CMHC, but rather the economic slowdown as people tighten their belts and reduce discretionary spending. Similarly, there would be a lot less economic activity in the housing/construction market.

 

[daveto]

One more point. In Insurance, we talk about "insureable" versus "uninsureable" risks.

Consider a fire insurance policy, where the person burns down their housee. That is an "uninsureable" risk because it is a not a random act that the individual was trying to avoid, but rather an example of a clear moral hazard. They burned the house down because they wanted the money. In this case, they would not be entitled to receiving their insurance benefit

Well, in mortgage insurance the same risk exists. But there is no penalty for willfully and recklessly defaulting (via bankruptcy). The belief is that the consequences of bankruptcy are a sufficient deterrent. However every one of us has a price, above which we would be willing to go bankrupt. For some it may be $50k, and for others $500k. But I think virtually all of us would eventually default and go bankrupt if we were far enough under water relative to our income/means.

Remember that a few years ago the bankruptcy laws were amended to exclude Student Loans for precisely this reason. Students were going bankrupt to avoid repaying relatively small amounts (significantly less than $100k in most cases)

In order to derive proper assumptions for the reserves, the CMHC actuaries need to assess how likely an individual is to perhaps declare bankruptcy and walk away from all their debts (including their mortgage) if they find themselves under water. In this scenario recourse vs non-recourse is irrelevant.

My guess is that when the dust finally settles, an examination of the CMHC reserve assumptions will show that it was these types of "moral" hazard (from how a small section of our population might respond to becoming underwater) for which they had not been adequately reserved.

 

[daveto]

Rough numbers, but the CMHC has about $2b claims a year, and $2b of income (after paying claims). So doubling the claims would reduce the income to zero.

If we assume that claims=defaults, and that $2b claims=0.4% defaults, then one can see that above 0.8% the CMHC is in the red.
Again, just very rough numbers.