The Decision That Saved Canada in 2008 Was Made in 1991

Transcript generated automatically from the video and lightly corrected. It may contain errors.

In October 2008, the government of Canada put $25 billion behind the country's mortgage market. Within weeks, that commitment had been raised to 125 billion. Most people read that as decisiveness. It was not. Canada came through the crisis without a single bank failure and without putting public money into a bank. The most important reason was a rule written in 1991, 17 years earlier, by officials who could not have named the crisis they were preparing for.

That is the thing almost nobody understands about how governments decide under pressure. By the time the crisis arrives, most decisions have already been made. I want to show you how that works using documents you can open yourself.

I served in Canada's Prime Minister's Office during the 2008 financial crisis and as chief of staff in the Senate of Canada during the pandemic. So, let me be direct about where the line sits. In my book, I do tell stories from those years. The ones that are mine to tell, obviously, the ones already on the public record. What I will not do is trade in confidences. The advice that was given behind closed doors. The deliberations, the positions people took when they believed the room was closed. That material is not mine to spend and it never becomes content. Not anonymized, not blended into a composite. That is not modesty, it's the job. People stop speaking candidly in front of you the moment they suspect you are collecting material. The discretion is not a limit of the work, it is a precondition for it.

Here is the better news. You do not need the private version. Canada's response to 2008 is one of the most thoroughly documented decision sequences in modern government. The Bank of Canada publishes a dated chronology of every crisis measure it took. The Department of Finance published the budget that framed the fiscal response. The Parliamentary Budget Officer published what was actually drawn down afterwards. The pattern is in the public record. That is what makes it teachable.

So, watch the order. The order is the actual lesson.

First, liquidity. In September of 2008, the Bank of Canada opened a swap facility with the US Federal Reserve, alongside the Bank of England, the European Central Bank, the Bank of Japan, and the Swiss National Bank. Coordinated, same day. Then, the price of money. On the 8th of October, the Bank cut its policy rate by 50 basis points, acting in concert with other major central banks. Two weeks later, it cut again. The Bank called the October move extraordinary in its own release. Then, funding for the banks. On the 14th of October, as part of the G7 action plan, the Bank introduced new liquidity measures. In November, it added a Canadian dollar term loan facility. Then, the mortgage markets. In October, the government began buying insured mortgage pools through the Canada Mortgage and Housing Corporation. 25 billion to start. Raised later to 125 billion. Then, and only then, demand. The budget did not arrive until the 27th of January of 2009.

Hold that last one against another one. The C.D. Howe Business Cycle Council, Canada's recession dating authority, puts the peak of the cycle in October 2008. The recession had already started when those emergency measures were rolling out. The fiscal response landed almost 4 months into it. That is not a failure. It is what the sequence costs. Fiscal measures need a budget, and a budget needs Parliament. Liquidity measures need a phone call.

Governments under pressure do not optimize, they sequence. And the sequence runs from the most irreversible failure backwards. So, if the payment system seizes, nothing else you do matters. So, it goes first. Even though it is invisible, even though no voter will ever thank you for it.

Now, the part that gets missed. Start in 1980. The average major Canadian bank was carrying about $40 of assets for every dollar of capital. Some individual banks were at 50. That is not a typo, and it's not the story most people expect. Canadian banks were not always the careful ones. They were by today's standards extraordinarily leveraged. So, Canada capped it. Not with risk weightings that banks could argue with. With a blunt ratio, total assets divided by total capital. The statutory authority came in 1980. A formal limit of 30 was set in 1982. With lower limits imposed bank by bank in practice. In 1991, the ceiling came down to 20. From 2000, a bank in good standing could apply to go as high as 23 and had to make a business case for it. The result? Since 1985, average leverage at Canada's major banks has stayed consistently below 20.

Now, watch what that constraint did during the boom. The Bank of Canada compared the four years to the third quarter of 2007. Over that stretch, Canadian bank leverage rose by about one and a fifth. In the United Kingdom, it rose by seven. In continental Europe, by nearly six. At US investment banks, which were not subject to a leverage cap, it rose by eight. The world's 50 largest banks reached an average of 30 to one. Canada barely moved. Not because Canadian bankers were more prudent by temperament, because there was a number they were not allowed to cross, and the regulator would cut their limit if they did.

Here is the uncomfortable version for anyone who runs anything. Your crisis performance is mostly determined before the crisis. What you do on the day is real, and it matters at the margin, but you are operating inside an envelope built while nothing was happening. The quiet years are when decisions get made. The loud week is when you find out what you chose.

People assume crisis speed comes from bold leaders cutting through process. Look at what actually deployed, and you find something a lot duller. Almost none of it was new. The swap lines with the Federal Reserve were an existing instrument, reopened and expanded. The list of securities the Bank of Canada would accept as collateral had already been widened in December 2007, nine months before the acute phase. The machinery for buying insured mortgage pools already sat inside the CMHC because Canada had been securitizing insured mortgages for years. The crisis response was largely about widening pipes that were already laid.

That is where speed comes from. Not from deciding faster, but from having already decided in advance, in calm conditions, about instruments you hope never to use. The practical translation? If a decision has to be invented during the emergency, you have already lost the time you needed. The work is to pre-authorize. Decide now what you would do, who can trigger it, and what the limits are. Then leave it on the shelf. Most organizations do the opposite. They build plans for what they expect. The useful thing is not a plan. It is a standing authority, a pipe already laid, that somebody is permitted to widen without convening anyone.

When Budget 2009 landed, the Department of Finance stated three principles in the document itself. Stimulus should be timely, targeted, and temporary. Timely was a given number. Measures had to begin within 120 days to be most effective. Targeted meant directing support where it would produce the most output and employment. Temporary meant phasing out to avoid permanent structural deficits. Read casually, that's a slogan. It's not. It's a filter. A budget in a crisis attracts every proposal that has been waiting years for a hearing. Each one arrives with a sponsor and a reason. Without a rule, you argue every case on its merits and you lose because under pressure, the person arguing hardest tends to win. A rule lets you decline without arguing. It is not that your project is bad, it cannot start inside 120 days. The rule takes the decision out of the room and off the relationship.

This is the piece I would most want a working executive to take away. Decision rules are not about making good decisions. They are about making ones you can defend at 3:00 in the morning in the fourth consecutive week of this. Set the rule before you are tired because you will be tired.

I want to be careful because this story gets told badly in both directions. It does not prove the response was optimal. The mortgage purchase program was authorized up to 125 billion. The Parliamentary Budget Officer reported purchases between 2008 and 2010 at about 69 billion, well under that ceiling. You can read that as prudent headroom or as a sizing that overshot what the market needed. The documents do not settle it. It does not prove Canada escaped. The Business Cycle Council dates the recession from October 2008 to May 2009, 7 months. Cumulative decline in GDP, 4.4%. That is a real recession. And being more resilient than the United States in 2008 is not the same thing as being unharmed.

The comparison is also flattering to Canada in a technical way worth naming. US accounting practice allowed banks to report derivative positions net. Canadian and international standards require gross. The Bank of Canada notes this understates US leverage relative to Canadian. The gap was real. It was also narrower than the raw numbers suggest. And I would rather tell you that than let you over-learn from it.

And the conservatism had a price. The structure that produced stability, a small number of large, tightly supervised, heavily domestic banks, also produces concentration and limits competition. A mortgage market that relies on a government insurer is stable precisely because the state is bearing the risk. You can hold the view that this was the right trade. You should not pretend it was free. I am describing how a system behaves under stress. I am not handing out marks.

You are not running a G7 Treasury. The patterns hold, and they scale down cleanly. So, what?

One. Your crisis capacity is set in the quiet period. Cash reserve, credit line, contract terms, who else can sign? None of it can be built during an emergency. Canada's binding limit was 17 years old when it was needed. Yours has to be, too.

Two. Sequence from the most irreversible failure backwards. Ask what breaks first. And if it breaks, what makes everything else moot? Protect that. Even though nobody will notice.

Three. Pre-authorize rather than pre-plan. Do not write a plan for the scenario you expect. Write standing authorities for the categories you fear. Who can spend what? Who can pause what? Without a meeting.

Four. Write your decision rules while you are calm. Three of them. In plain words. Before you need them. Their job is to let you decline things quickly without re-litigating.

The fourth one is the K in SPARK. Key decision-making and execution. It is the one most people skip because in calm conditions, it feels unnecessary. That is precisely when it is available to you.

The decisions that carried Canada through 2008 were mostly not made that year. They were made by people setting a leverage ceiling in 1991 and widening a collateral list in 2007. People who got no credit because a prevented crisis is invisible. That is the honest shape of decision-making under pressure. So, the moment is real. The moment is also mostly the settling of accounts on choices made when it was quiet. Which means the useful question is not what you would do in a crisis. It is what you are deciding right now while nothing is happening.

I put one of these out regularly, sourced through to the original documents. It goes to the Brief first. The link is below.