CMHC has fired back at Reuters for its Wednesday article criticizing CMHC’s “emili” underwriting system.
The nation’s largest default insurer states that Reuters’ story contains inaccuracies. Specifically:
“The article implies that emili relies primarily on information provided by home sellers, does not look at the specific property and that it accepts this information without verification. This is not the case.”
CMHC notes that it uses four tools to evaluate properties, including:
- the physical characteristics of the property (like square footage, lot size, age, style of the home, etc.)
- the municipal property tax assessment
- historical and current sales activity within the local housing market, and
- prior sales activity of the property being assessed, when available.
In many cases, data is crosschecked among these varying sources.
Reuters goes on to speculate (somewhat blindly) that emili “could raise the risk of a devastating U.S.-style crash.” CMHC replies point blank that emili is “objective and consistent” and “does not contribute to higher home valuations.”
Emili is “continually updated” and “independently reviewed” by a third party, it says. (The company doesn’t state who that party is.)
In addition, “The lender is contractually responsible for ensuring valid data is submitted into emili…”
In our past experience, branch staff at some lenders have significant control over the input of this data.
Related story: CMHC’s emili Under Fire
Rob McLister, CMT
http://finance.yahoo.com/news/northcore-enters-loan-agreement-222053052.html
This is a very weak rebuttal. CMHC should release the verification data that shows how close emili is to manual appraisal or subsequent sale price. Otherwise it’s just one biased source (the news industry selling papers) against another (CMHC protecting their reputation).
Will anyone here tell us what the inputs are that are required by emili?
The problem is not Emili, but rather the fact that banks are lending out amounts progressively greater than supported by the appraised values, enabling a vicious cycles of ever-higher home prices.
Like almost all AVM’s including MPAC, the inputs are proprietary.
It would be like trying to get Cadbury to tell you how they get the caramel into the Caramilk bar.
Disagree with gokou3 … banks are risk adverse … I have had several deals that because of a flip to another lender have been passed thru Emili and then Genworth (who quite often require appraisal) … all cases showed Emili and the appraisal to be within 2-3%.
To me this is just the rest of the world trying to get our housing market to crash. Except for Vancouver I can’t see that happening.
Hi Rob. You stated that: “In our past experience, branch staff at some lenders have significant control over the input of this data.”
What does that mean? Do they inflate total square footage the property? Exactly what control over inputs do banks that have?
Also, is that why sometimes a deal can be declined by CMHC when it is submitted by a mortgage broker, but then approved when a branch sends it in?
Thanks
Hi gokou3. How are the banks …”lending out amounts progressively greater than supported by the appraised values”?
Doesn’t CMHC limit them to 80% of the appraised value on Refis, or 95% on purchase etc?
Here is a discussion about it:
http://www.mortgagebrokernews.ca/mxcvjpicwrzxh/brokers-banks-abusing-emili/124534/
Banks are risk adverse? Here read a Noble prize written working paper on real estate, governments and banks. http://ideas.repec.org/a/bin/bpeajo/v24y1993i1993-2p1-74.html
“Bankruptcy for profit occurs most commonly when a government guarantees a firm’s debt obligations”
Hmmm that sounds familiar…
Nothing is different here. It’s the same fraudulent scheme again and again.
Hi MortgageMan,
I got a similar situation last week. According to the CMHC’s norms, you’re not suppose to take into account a rental income when a client buys a single-unit dwelling he occupies. In this case, it was a bi-generational cottage. I was stuck with a GDS of 42% and Desjardins was just below 39%. I finally learned that the branch’s banker used a ”normal rental income” since it was bi-generational. I called my CMHC bdm to argue the case and he told me that the financial institutions were responsible of data entry and fact checking. Basically, CMHC’s approval is on auto-pilot. However, should a default occur, the CMHC would look at the inputs in depth and could refuse to pay if it’s norms are not met.
Yes, Canadian banks want to bankrupt themselves. Now you’re thinking Watchdog. How boring life would be without your irrelevant analysis.
Okie dokie VJ. But when CMHC and Genworth’s insurance cap is breached and banks wind down loans (as they have been), just remember this post.
@Appraiser,
Surely anyone who works in the mortgage business would know what *inputs* go into emili… eg, number of rooms, square footage, condition of house, age of house, age of roof, number of rooms, postal code, etc are all things I’d expect would go in. (Perhaps not?)
If that was proprietary as you’ve stated, nobody would be able to write a mortgage application for an insured mortgage.
Is anyone willing to divulge what information goes into emili to get a valuation?
The silence here is deafening.
What I would like to know, is how a
property may fail AVS with a broker,
yet a financial institution is able to
get the same or an even higher value.
Do the banks use a different system for
their automated valuations?
Where was Reuters when US banks and Fanny Mae where underwriting mortgages in excess of 100% of marketvalue and appraised value to people with no jobs?