Moody's downgrades Manitoba's debt ratings to negative | GIC discussions | Discussion forum

Norman1 said

I can imagine credit unions running into trouble should their loan losses become serious. A serious downturn in the the province's economy would do that. No apocalypse necessary.

But, don't assume that the terms of that support would be painless, full restitution.

In 1987, the Principal Group in Alberta went bankrupt...

With Principal, whatever the regulatory failure, people were investing in unguaranteed notes, nothing to do with a deposit guarantee system. Credit union deposits are unconditionally guaranteed by the deposit corp, irregardless of where the depositor resides. To repay less than 100% or only to provincial residents, the government would first have to let the guarantee corporation fail. Say you are a member of a provincial credit union, possibly just having had your deposits made good by the government at less than 100% and only because you are a provincial resident. The credit union says "hey, even though the old deposit guarantee corp failed and you risked losses, your deposits are now guaranteed by this brand new guarantee corp, so you shouldn't worry". Do you stay with the credit union, or immediately flee to a CDIC bank? I thought so!

Deposit guarantee corp failure == near-fatal banking and confidence failure. They didn't let it happen in Ireland, Spain, Cypress or really even Iceland; I'm fairly confident a Canadian province won't either...

As for credit union health, remember, we aren't talking about making them less profitable, sick, or illiquid - you have to come up with a scenario where the system loses enough equity that it can't be managed, the guarantee fund is exhausted, and there are still losses to deposits. I'm not convinced you can get that point with a "serious downturn" that isn't more like "unprecedented kaboom".

Elsewhere, I ran a little bootleg stress test on the books of a large Manitoba CU. You can put an extremely unlikely loss rate before you destroy all the equity and tap the guarantee fund (although of course the CU would be toast). Remember, many mortgages are CMHC-insured and have essentially transferred the risk to the taxpayers. Others will have at least 20% equity, so only an enormous (unprecedented?) meltdown in Manitoba housing pricings is going to destroy equity if foreclosed.

If anyone wants to propose some actual numbers (house prices drop by X% with Y% defaults, Z% losses on commercial loans) I could run them against that CU's books, but just saying something like "serious downturn" really doesn't do much.

Norman1 said
I think two points are being confused here: (a) Manitoba credit unions are quite safe to deposit money into and (b) the money deposited into a Manitoba credit union is backed by the province of Manitoba and is therefore safe as a Province of Manitoba provincial bond that's a direct legal obligation of the province.

I'd certainly agree that (a) is true and (b) is false, but would insist on: (c) no other province (except perhaps Alberta) has a legal obligation to provide funds to the guarantee corp either, and (d) notwithstanding the lack of legal obligation, as a practical matter the costs of the government providing guarantees, loans or even un-repaid equity into the guarantee fund would be less painful than letting the fund collapse and accepting the consequences to the provincial economy.

It isn't a certainty, of course, but you can certainly see it in things like the bailout of Fannie Mae and Freddie Mac (the US government did not have a legal obligation), and I'm not aware of any reasonable example of where a government has let a guaranteed financial system collapse rather than helping it.

Norman1 said
I had a look at the stress test you mentioned, NorthernRaven. I don't believe the test numbers.

With a 50% drop in residential resale value implied by the uninsured mortgages losses, it would be very unlikely that the losses in consumer loans, consumer line of credits, and commercial loans would just be 10% to 15%.

That kind of drop in residential real estate usually comes from significant loss of employment income. No job means no money to make their consumer loan payments. There would be no money to spend. Businesses would suffer and may not be able to pay their loans either. There would likely be a draw on the Manitoba guarantee fund in that kind of situation.

I had no practical way of modelling a distribution of shocks among the various asset types, and merely put unreasonably high numbers in for the unguaranteed mortgages line, and sprinkled some nominal percentages for the rest so as to wipe out 90% of the existing capital of the CU. If you want to quibble with those, I'd insist on revisiting the mortgage numbers.

The 30% loss rate only implies a 50% drop in house values if you assume that all the mortgages are at the 80% origination limit. In practice, the book would have a range due to payments on older mortgages and larger down-payments, so the actual cushion would be larger. If you assume an average 70% ratio, 30% loss implies house falls of 60%, and 60% ratio a 70% drop in house prices. Even 50% would mean Manitoba would have to see falls as bad as the worst US cases (~60% in some Florida/California cities or Phoenix), which is stretching things.

Also, I had a 50% default rate, which is probably way beyond what you might have seen over any representative portion of mortgages in Ireland/Spain/worst-US, although I don't know if there's any handy cumulative default stats that could be plugged in from places like that. Also, the sub-80% unguaranteed mortgages might preferentially belong to more financially stable households that might be able to survive a temporary job loss without default. In any case, by tweaking the loss and default rates to more plausible yet still disastrous levels there would be more capital to raise the other losses to something you might approve. Obviously the credit union would be liquidated by the regulator long before it got to this point, but for our purposes we're only interested in drains on the guarantee fund.