Hi,Anyone know any sources of monthly seasonality for EUR, FRF, GBP, USD, JPY inflation indices?i.e. Reuters/Bloomberg pages, brokers etc.I know MarkIT have a service which covers the first 4.Thanks.

Jon,It would be interesting indeed to find a Bbg or Reuters source but I must admit it would be surprising.I thought that seasonality calculations i.e. historical range, deviation from the mean ect... were left to the bank's discretion (for risk purposes essentialy)?Sorry to reply to your question with another question but: - How often would these seasonality figures would change?thanks,K.

Does Tullett Prebon not have these?

Have calculated seasonals from historic data for EUR HICP exT, but would be intereseted to know if a market standard has developed as the market has matured?

TomK,I agree, I have done my seasonal calculations for Gilts but that would be interesting to have the market standards. I dont know where/what is the consensus.

I am also working on this issue. If there is no market standard, can we come up something most people will agree?Here is my idea:A)Regression method1) assume the CPI index is linear with time, then we can do linear regression on the index, and get the averaged seasonality with sum of 0. The seasonality adjust is on the absolute CPI number2) assume the CPI rate is linear with time or CPI is expolential with time, then we can do the same thing as 1) on the LN(CPI). The seasonality is on the log of CPI.B)X11 method. in stead of do regression, we can use X11 method. this can be done either on CPI or log of CPI.Any comments? or which one you prefer?

For the US, you can get seasonal factors from comparing the nonseasonally adjusted cpi with the seasonally adjusted cpi. I think the BLS uses X12 to do the seasonal adjustments.Many European contries don't publish a seasonally adjusted inflation index. I don't think the seasonals for EUR HICP are very stable, because some of the member countries make methodological and compositional changes to their indexes--but that's what makes trading inflation fun.A simple way to estimate seasonality is to regress the m/m pct change in the inflation index on 12 seasonal dummies and some explanatory variables, but this method assumes that the future will be like the past.

QuoteOriginally posted by: u418936For the US, you can get seasonal factors from comparing the nonseasonally adjusted cpi with the seasonally adjusted cpi. I think the BLS uses X12 to do the seasonal adjustments.Many European contries don't publish a seasonally adjusted inflation index. I don't think the seasonals for EUR HICP are very stable, because some of the member countries make methodological and compositional changes to their indexes--but that's what makes trading inflation fun.A simple way to estimate seasonality is to regress the m/m pct change in the inflation index on 12 seasonal dummies and some explanatory variables, but this method assumes that the future will be like the past.Firstly, there is only one european HICP from eurstat which is independant from the EUR countries.Secondly, there are two methods to estimate the seasonality factor: - X11 (not X12 because we have 12 months not 13); - Stochastic estimation.Most of the market players use X11.

QuoteFirstly, there is only one european HICP from eurstat which is independant from the EUR countries.Secondly, there are two methods to estimate the seasonality factor: - X11 (not X12 because we have 12 months not 13); - Stochastic estimation.Most of the market players use X11.Good god. I pray that you are or will be trading inflation at a big, real money shop, because guys like you create opportunities.I'll make some points of my own.[1] The HICP IS a weighted average of the CPI of the countries in the euro zone. Link is here: http://www.ecb.int/stats/prices/hicp/html/index.en.html. If it were "independant (sic) from the EUR countries", whose inflation rate would it be measuring? If, for example, you heard that inflation was high in France, it should give you no information about inflation in the euro zone?[2] The "11" or "12" that appears behind the "X" has nothing to do with the number of months in a year. Why on Earth would you think that? Did you feel sorry for the guy who created X12, because it's useless on our planet? The BLS does in fact use X12. Here's the link: http://www.bls.gov/cpi/cpisapage.htm.[3] What the hell is "stochastic simulation"?