Earlier this week, one of the deputy governors gave a speech explaining the Bank of Canada’s new approach to measuring core inflation. Pretty boring stuff if you are not an economics geek, or if you have a life. However, the shift in their approach in determining the underlying inflation trend could make it more difficult for market participants to get a timely read on the Bank’s thinking.
Since 2001, the Bank used CPIX as the measure of core inflation, a proxy for the underlying trend of inflation. This measure excluded the same 8 components that had been statistically determined to be the most volatile. However, over time, the Bank became disenchanted with this measure as some of the categories that were not excluded and remained in the core showed that they could also exhibit large transitory shocks. This could result in core inflation temporarily deviating from the true underlying trend. Worse, from a modeller’s perspective, this weakened the correlation between their measure of the output gap, whether the economy was in excess demand or supply, and their measure of underlying inflation.
Setting policy off a misleading operating guide could ultimately be costly to the economy. So the Bank, wanting to improve its ability to separate the signal from the noise, undertook to find the best way to manipulate the CPI data to get a trend rate of inflation more consistent with their theoretical construct of how inflation gets generated.
The good news for geeks is that they didn’t find just one measure, they found three. CPIX will now be replaced by CPI-trim, CPI-median and CPI-common. In short, two of the measures kick out price changes that are considered outliers and the other looks for price changes that are similar across various categories.
As with CPIX, history shows wide divergences between the three new underlying measures of inflation and the headline number. The gaps are usually driven by changes to the price of energy. No surprise there. What becomes more challenging for central bank watchers is that history also shows episodes where the three new measures diverge from each other. For example, in the fourth quarter of 2016, the three measures showed that underlying inflation was somewhere in a range between 2.1% and 1.4%. Apparently, this type of wide divergence is seen by the Bank, not as a weakness, but as a way of giving them additional insight into what is happening to the trend. They say that focusing on just one of these measures would have led them astray in their assessment.
This leads to one of my issues. The transparency around how the Bank will arrive at its verdict on the underlying trend of inflation is poor. They have told us of at least three of the ingredients that will be thrown into the pot, but no idea about the quantities that will give them the mixture that they are looking for, nor do we know if the same recipe will be consistently used. The cynic in me suggests the recipe followed behind closed doors will be the one that best suits the narrative that the Bank is looking for at the time. In reality this is not likely to be the case, but how will anyone ever know? In the example above, with concern about the relative strength of the currency, should we expect more weight to be given to the 1.4% than the 2.1%?
The “read our minds, don’t worry, we know what we are doing” approach versus a verifiable, consistent, publicized operational guide seems like a way that will lead to confusion in the market. It is also inconsistent with a Bank that complains that the market does not do enough analytical work on their own and relies too much on the Bank. This new approach appears to give the market no other choice than to rely on the Bank’s analytics.
The Bank has also decided to no longer publish an outlook for core inflation. Going forward in the MPR, they will only publish a projection of total CPI, as they want to reinforce the headline number as their official target. I did not realize that there was mass confusion out there in the real world. Instead, they will include a discussion on the key forces underpinning the inflation process at the time of their analysis. There will be no profile of projected underlying inflation given to the market to allow them to make any ongoing assessments of whether the Bank’s view on core is developing as expected between MPRs.
I am probably stretching things, but I will make one final point. It may be relevant to the market to note that these three new measures of core prices respond less than CPIX to any currency devaluation. Theory dictates that a move in the exchange rate will have an impact on prices but, since it will be transitory, it does not necessarily warrant a policy response as long as inflation expectations remain well-anchored. Kicking out any leap in the price of cauliflower that could jar inflation expectations, and numbing the underlying trend of inflation to a currency move, is one more quiver the Bank can lean on to tilt the case towards more desired currency depreciation.
No comments:
Post a Comment