With the Bank of England set to meet, markets are vetting the CPI report for the United Kingdom from the standpoint of what it will cause or permit the Bank of England to do next. The year-over-year inflation rate in this report has dropped causing some analysts to say it keeps the door open for a Bank of England rate cut.
But does it really?
I am not a fan of looking at short-term inflation indicators or for central banks to make policy based on what short-term inflation indicators are doing. However, central banks need to be aware of what these indicators are doing and what they're telling them about inflation. As I have mentioned many, many times, year-over-year inflation rates are well behaved, but we must be careful in vetting them. The best way to think of them is that they are the result of 12-month-to-month inflation changes over the past year. Each month the inflation rate has eleven of those same changes as the previous month. If 12-month inflation falls in the current month compared to the previous month, it means that the month-to-month inflation change in the current month is lower than the month-to-month inflation change and the oldest month that was just dropped out of the index. Is that a reason for a central bank to cut interest rates today?
The only answer to this question that stands up to scrutiny is this: it depends. And what it depends upon is the context and the trend. It depends on how broad the change in inflation is. It depends on whether that change in inflation is driven mostly by one category like changing energy prices. It depends on how economic performance has shifted recently (if at all). ‘It depends’ means there must be context for it.
In this case, the CPI-H month-to-month just rose by 0.5% in February, not exactly a very good number- a number that's going to annualize to an inflation rate of over 6%. That's not particularly good. On the other hand, the inflation rate year-over-year goes down because a year ago the CPI-H is dropping out a rise in the price index month-to-month of 0.9%. Clearly 0.5% is less than 0.9%! Viola! The year-over-year inflation rate falls. However, viewed on its own that 0.5% inflation rate is quite high. Also, the CPI-H core index that excludes food, alcohol, tobacco, and energy, rose by 0.4% in February; that's also a pretty strong increase. The headline month-to-month annualizes to 6.2% while the core annualizes to 4.9%. Neither of these strike me as performance that is good enough to accommodate a rate reduction by a central bank with a 2% target and a long legacy of overshooting its target. In fact, if we compare those annualized increases to the CPI-H over 12 months, the 12-month index increased 3.9%, the six-month increase was at an annualized rate of 3.3%, and the three-month increase in the CPI-H was at a pace of 4.9%. The one-month annualized change represents an acceleration compared to all these metrics (Yikes!). Similarly, the core for the CPI-H rises 4.8% over 12 months, rises at an annual rate of 3.7% over six months and at a 4.2% pace over three months. However, the annualized one-month increase is 4.9%, once gain stronger than all those metrics.
I'm not advocating that the central bank look at the annualized month-to-month number to make policy. However, the central bank should not ignore it either… The central bank needs to look at a sequence of these numbers and have some idea about how inflation is ‘trending’ if it's going to change policy. And since the Bank of England - like the Federal Reserve and the European Central Bank - has been over the top of its target for some time, it would seem most appropriate for the central bank to make sure that inflation rate is on the correct downward path and at a more acceptable pace before it begins cutting interest rates.
As this discussion made clear, almost all the focus on today's CPI report is based upon how it's going to fit into BOE policy and how the year-over-year rate dropping paves the way for the Bank of England to become more accommodative. However, in my view, that's not even close to right.
There is good news however... The good news is that the sequential inflation calculations for the headline and the core are not clearly accelerating and they're showing some temperance. The diffusion results from these calculations show that inflation over 12 months compared to the previous 12 months, over six months compared to 12-months, and over three months compared to six-months is demonstrating a step down across most categories persistently. That is very good news. The diffusion indexes are less than 50 (less than 50%) indicating that inflation is declining in more categories than it's increasing, period-to-period. And that's reassuring. However, diffusion calculations are executed across categories and summed-up without weighting. The actual inflation index employs weighted components, and the weighting is clearly an important part of the process. However, if we find that inflation is not broadly accelerating, that may be a sign that the inflation process is moderating and that it is also poised to put in some lower numbers in those categories that have higher weights that are presently still performing poorly.
Monthly diffusion...not so fast The monthly numbers are not as attractive as the sequential calculations for diffusion. In February, the diffusion calculation is at 54.5%; that’s up sharply from a 36.4% reading in January; January is down sharply from a 72.7% figure in December. The monthly numbers compare the month-to-month percent changes in inflation to those of the month before. Diffusion above 50 tells us that inflation is accelerating in more categories than it's decelerating. February and December show broad-based inflation acceleration while January brought respite with more deceleration than acceleration.
Diffusion calculations are important. They tell us about the breadth (not the intensity) of inflation. No central bank really makes policy based on the breadth of inflation, but it's still an important statistic to keep track of.
Global
