William Watson: Playing the inflation average

The Bank of Canada’s current inflation mandate is to keep the rate of increase of the Consumer Price Index (CPI) at the “two per cent midpoint of an inflation-control target range of between one and three per cent.”

We can debate how much inflation “control” that actually provides. With two per cent inflation, what costs $1 now will cost $1.61 in 25 years, $2.64 in 50 years and $7.10 in 100 years. I know: those all sound very far away. Except that young people keep telling us we have to plan far ahead for our planet. Maybe for our planet’s money, too.

But, for better or worse, that’s the target — unless it gets changed by the every-five-year agreement between the Department of Finance and the Bank of Canada that’s due to be renewed by the end of this year.

How has the bank done in terms of hitting its target? There have been 142 months since January 2010, to pick an arbitrary start date. In 11 of those months (eight per cent of them) the year-on-year change in the CPI was exactly two per cent. That happened three times in each of 2014 and 2019, so: kudos to the bank for hitting the bull’s eyes. In six other years, however, it got zero bull’s eyes.

Bull’s eye is a high bar for this job, however. Only by fluke, really, does the CPI hit an exact value. More generally, and more to the point here, there were 86 months in which the CPI increase was less than two per cent and just 45 in which it was greater than two per cent  — though those 45 include each of the last eight months. In the last seven months, the CPI increase has been greater than three per cent and in the last three it has been greater than four per cent — a value not previously breached in the 142 months.

Over the almost 12 years, the bank has typically under-shot its two per cent target, missing it low 61 per cent of the time versus high just 32 per cent of the time. Those of us who like our money’s value stable — who prefer that a dollar in 2046 or 2071 or 2121 have roughly the same value as a dollar today — haven’t minded much that the bank has missed low.

On the other hand, people with an inflationary bias, who probably don’t actually enjoy inflation but believe it helps keep the economy operating closer to its capacity output and employment, have always complained about what they see as the bank’s systemic bias against inflation. And over the 142 months the inflation rate has averaged 1.7 per cent, not two per cent. (But, hey, the people who run the bank are central bankers: they’re supposed to be biased against inflation.)

The inflationists have lobbied for an explicitly higher inflation target — two to four per cent, say, with a target of three. But they’ve also lobbied for a policy of not letting bygones be bygones. Which means: if the bank misses low for a while, it should not worry overly about subsequently missing high, at least for a while. It should balance present highs against past lows before raising interest rates or making any of its other policies more contractionary. In fact, the U.S. Federal Reserve adopted this approach earlier this year — though it didn’t spell out the crucial ingredient of over how many months, or years, it would average.

Until recently, with a backlog of undershoots having built up, letting overshoots and undershoots balance in this way would have had the effect of making policy less hawkish. But (see chart) if you look at the last eight months, the high-side gap between the CPI curve and the two per cent line now represents roughly the same area as the low-side gap built up during the first year of the pandemic.

If we take the view that two years is the right span over which to balance things, even the (formerly inflationist) rule of “play the inflation average” indicates it’s about time to start tightening.

It has been taking an unusually long time for the Finance Department and the Bank of Canada to announce the renewal of their inflation-control agreement. Maybe they’ve been thinking that an averaging approach would be a clever way to avoid the tough choices David Laidler writes about elsewhere on this page. But the inflation surge has undone the cleverness, turning averaging into a sensor that now says “tighten.”

Financial Post