SEPTEMBER 04, 2011
CMHC vs. Fannie Mae: Round 2
It’s nice sometimes to take a back seat and get some insights from some of the great analysts out there. Over the weekend, I received an insightful email from Jesse over at Housing Analysis, while Jonathan Tonge, author of the AmericaCanada blog posted an interesting comment on a previous artile. Both related to concerns over CMHC, and both are worth sharing.
Shades of Fannie Mae:
We heard recently from Chris Horlacher, a chartered accountant, who examined CMHC’s books and compared them to Fannie Mae, with some troubling conclusions. Officially known as the Federal National Mortgage Association, ‘Fannie Mae’ was a US government sponsored entity and a publicly traded stock that was essentially taken over by the feds in 2008 after its loan portfolio experienced significant losses.
The tale of mortgage insurers in the US ought to cause us to pause and closely scrutinize the books of CMHC, Canada Mortgage and Housing Corporation, the government entity with an analogous, though not identical role as Fannie Mae.
Let’s remember that no analyst worth listening to would suggest that the careless and blatantly fraudulent underwriting practices that went on in the US have been replicated in Canada to anywhere near the same degree, a topic I recently discussed when I looked at subprime lending in Canada.
That being said, it should be equally clear to anyone who looks critically at the data that house prices in many parts of Canada are at concerning levels. For those not convinced, or who may be new to the site, refer to the primer articles for more on that. The bottom line is that how we arrived at the current level of overvaluation may prove to be irrelevant. A realignment with underlying fundamentals will be painful, regardless.
Speaking of such a realignment, I have little doubt that a housing correction would be associated with a significant rise in delinquencies. This is in part due to the role that persistent negative equity plays as a motivating force behind mortgage delinquencies and foreclosures. Furthermore, as I recently discussed, the real estate boom has played a significant role in directly and indirectly stimulating the Canadian economy and labour market, meaning that a house price correction would also most certainly be associated with a nasty recession and persistently high unemployment for several years.
Bottom line: Comparisons between Fannie Mae’s and CMHC’s balance sheets may offer little in terms of predictive value, but they are interesting nonetheless. They may ignore hedging strategies and other factors that may help mitigate widespread loan losses, but they are not without merit. At the very least, they should highlight how quickly a stable balance sheet can erode in the aftermath of a housing correction. Furthermore, they do call in to question the true risk that CMHC poses to the Canadian taxpayers.
Inspired by Jonathan Tonge’s comment, I put the following chart together comparing some key ratios from CMHC’s mortgage portfolio in 2011 and Fannie Mae’s mortgage portfolio back in 2007.
Data for CMHC can be found in their latest quarterly financial results. Data for Fannie Mae can be found in this 2007publication from their website. Note that the ratios for Fannie Mae were already deteriorating in 2007 as the bubble in the US had already been deflating for a year and a half at the time of publication.
Again, let me stress that the Fannie Mae data was from well after their bubble had burst. Note that the Loan-to-Value (LTV) ratio had zero value in predicting the housing crash, a topic I discussed recently when we compared average owner equity in Canada and the US.
Fannie Mae confidently reported their strong balance sheet and prudent risk management back in 2007 using a tone that sounds eerily similar to CMHC today. Then 2008 hit. We would be wise to keep in mind just how quickly these balance sheets can erode when house prices fall and the economy sours along with them.
Thoughts from Jesse:
I have to highlight a fantastic email I received from Jesse, the author of the Housing Analysis website, regarding CMHC. Here is his email in its entirety. It is WELL worth the read!
I'm writing here with some general qualitative thoughts surrounding the ongoing CMHC "issues" being discussed in both the MSM and the blogosphere. I think you know I think CMHC has unfavourably contributed to overvaluation of housing in certain Canadian markets -- a sheep that has lost its way -- however I do take some issue with hyperbolic insurance reserve ratio criticisms meted out by some -- $600 Billion insurance under force and only $12 Billion in reserves -- but we can leave that aside for the moment.
I have been wanting to run some "stress tests" on what we know of CMHC's balance sheet for a while but, like others, have found required variables missing from public scrutiny and have been forced to put on my Holmes hat and cape to give it my best guess. But even then there were a few things nagging at me about what I was doing. All the numbers I was running were showing CMHC to actually be reasonably capitalized (and by "reasonably" I am speaking relatively here), based on current insurance under force, such that the net drain on taxpayers would be in the order of a few tens of billions of dollars spread over several years.
This indeed matches the claims made by CMHC of its current state of financial health of late. Horrible, yes, but nothing to the scale of the distress witnessed by insurers in the US. Leaving aside the "Canada doesn't have a subprime problem" platitude, I thought I'd put into words a couple of nagging concerns I have with "the books" that seem in some ways unique to Canada's housing finance market, and not in a good way I'm afraid.
My first concern relates to future liabilities. We can trace the accounting of this to see the problem: low-ratio loans are provisioned by banks with a relatively small capital reserve and the assumption that most plausible equity collapses can be recouped after a 5 year duration (i.e. falling prices won’t eat up the down payment over the loan duration), at which point the bank gets its money and tells the borrower to f-off or laterals to CMHC or private MI (mortgage insurance).
So far, no major risk shows up on the banks' books. But now CMHC faces a loan application from a desperate mortgage holder who needs to refi with little to no equity — potentially not even the minimum 5% — and CMHC makes them pony up the MI premium and assumes the liability, now using an insurance accounting model. The problem is that if prices are falling, historically defaults will happen close to a magnitude higher frequency typically with a 15%+ drop in prices.
CMHC has provisioned for uncorrelated risk when mortgage defaults are anything but. But the issue I'm highlighting here is that I haven't seen any stress test on what CMHC and the government will need to do to handle the banks' dumped loans in order to stave off a significant foreclosure spike. If you believe prices can fall 15% nationally of course...
The second concern has been highlighted before by CMHC critics -- lenders don't want high-ratio loans on their books because they have to provision significant, if any, capital for them. Someone once told me banks actually prefer loans with MI because under accounting rules they don’t need to provision any capital because it’s perfectly hedged by a government-underwritten body; with non-insured mortgages they are required to have a capital reserve. (On a side note, having fully-insured mortgages makes banks’ leverage ratios look worse if you don’t know how to read the reports properly.)
The trouble is again in the low-LTV segment of the market, where some lenders have been qualifying low-LTV borrowers under shorter duration and discounted rates, and using various favourable add-to-income ratios for secondary rental income, to get their DSRs (debt-service-ratios) under the requirement, but high-ratio CMHC-insured loans require approval at the 5 year rate. (Central parts of Vancouver CMA are often financed with suites 100% add-to-income. Not sure, but some lenders may still be using 80% rental offset when calculating payments.)
So if prices do drop there is an immediate disconnect between a borrower who was qualified at the favourable discount rate and what CMHC requires. That’s a big problem only solved by recasting the mortgage in an even higher LTV offering, if possible. Otherwise… As with my first point, this problem is currently off everyone's books.
My third and final concern of this letter is stating the obvious, that this is all in the context that the government is attempting to "cool off" borrowing using marginal buyers insured through mortgage insurers as the lever. What’s interesting about Vancouver, where I live, and other regions of Canada, is that, even after the CMHC rule changes earlier this year, sales are not “40% less” as cited, they are about the same as last year with prices ostensibly 7% higher as measured through the Teranet House Price Index. So much for crimping first-time buyers. The federal government should, I am hoping, be awaking to the fact that this market may not be what fire marshals call “self-extinguishing” and are indirectly picking up the tab for reasons previously mentioned.
(I haven't mentioned an impending interest rate spike here; the point I'm attempting to make is that there are problems even if rates remain low ad infinitum. Suffice it to say that if rates do spike, all the above problems are exacerbated and policymakers will sell out the brown pants section of local department stores that much quicker.)
As a bit of background: I highlighted some interesting insight into CMHC from House of Commons testimony here. Regarding the condition of 100% underwriting of CMHC-insured loans, where other private insurers are only 90%, meaning banks don’t require any capital provisions on CMHC loans, CMHC representative Karen Kinsley stated the following:
The issue of the differential in our mandate and the cost of that really gets to the nub of the difference in the guarantee between CMHC and the private insurers. We are, by virtue of being a crown corporation, 100% guaranteed by the Government of Canada. Recognizing that private insurers can select the markets they choose to be in, and obviously they will not serve those that are less profitable, the government has set the guarantee for private insurers at 90%. That 10% differential in the guarantee, in order to create a level playing field between us, compensates us for that difference.
We have been able to operate successfully on that basis, as is evident by our annual returns, and the over $12 billion that we’ve been able to return to the government.
This model concerns me. 'nuff said. I thought these issues important enough that their absence from discussion in the media would be doing Canada a bit of a disservice.
Indeed. Thanks Jesse!
For more on CMHC, check out the following posts: