FreeTrader Posted February 14, 2011 Report Share Posted February 14, 2011 I was doing some work last week on some RPI indexing stuff and just out of interest I compiled a table of category inflation over various time periods. The following tables show the annualised 5-year and 10-year inflation rates for the product categories as measured by the RPI. I’ve also included the weights used in the RPI for each category (which are always out of 1000). Just to be clear what I mean by this, here’s an example: In December 2000 the index for bread was 137.3 and in December 2010 it was 221.6. So over 10 years the price of bread has risen by 61.4%. This is equivalent to an annual rate of inflation of 4.9035%, which is rounded to 4.9% in the table. I’ve split the table into three images for convenience: Here’s the sorted table for 5-year annualised rates: ...and the sorted table for 10-year annualised rates: Quote Link to post Share on other sites
the_duke_of_hazzard Posted February 14, 2011 Report Share Posted February 14, 2011 Very very good work. Thanks. Quote Link to post Share on other sites
swissy_fit Posted February 14, 2011 Report Share Posted February 14, 2011 Very very good work. Thanks. Yes, it's going to look quite nasty when the cheap goods from the Far East are no longer cheap. Quote Link to post Share on other sites
FreeTrader Posted February 14, 2011 Author Report Share Posted February 14, 2011 Yes, it's going to look quite nasty when the cheap goods from the Far East are no longer cheap. According to the RPI measurement, women’s outerwear clothing is almost 40% cheaper today than it was in 1987. Yesterday though the FT carried an article warning of large price rises in clothing later this year due to rapidly increasing costs feeding through the supply chain (cotton hit a new record on Friday). Here’s a similar article from USA Today: Clothing prices to rise 10% starting in spring The era of falling clothing prices is ending. Clothing prices have dropped for a decade as tame inflation and cheap overseas labor helped hold down costs. Retailers and clothing makers cut frills and experimented with fabric blends to cut prices during the recession. But as the world economy recovers and demand for goods rises, a surge in labor and raw materials costs is squeezing retailers and manufacturers who have run out of ways to pare costs. Cotton has more than doubled in price over the past year, hitting all-time highs. The price of other synthetic fabrics has jumped roughly 50% as demand for alternatives and blends has risen. Clothing prices are expected to rise about 10% in coming months, with the biggest increases coming in the second half of the year, said Burt Flickinger III president of Strategic Resource Group. (More at link) Quote Link to post Share on other sites
Misanthrope Posted February 14, 2011 Report Share Posted February 14, 2011 Very very good work. Thanks. +1 Quote Link to post Share on other sites
winkie Posted February 14, 2011 Report Share Posted February 14, 2011 +1 Very interesting.....what I am seeing, put into an easy to read visual format. Quote Link to post Share on other sites
Democorruptcy Posted February 14, 2011 Report Share Posted February 14, 2011 I was doing some work last week on some RPI indexing stuff and just out of interest I compiled a table of category inflation over various time periods. Very interesting. The one thing that caught my eye was purchase of motor vehicles -1.7% & -2.2% I find that hard to believe. Quote Link to post Share on other sites
Jadoube Posted February 14, 2011 Report Share Posted February 14, 2011 Clearly laid out and clearly explained figures. Very nice, indeed. Thank you. Quote Link to post Share on other sites
scepticus Posted February 14, 2011 Report Share Posted February 14, 2011 The proportion of overall expenditure each item makes up is also important . Here is graph to that effect, not sure whether it is US or UK. Quote Link to post Share on other sites
FreeTrader Posted February 14, 2011 Author Report Share Posted February 14, 2011 Very interesting. The one thing that caught my eye was purchase of motor vehicles -1.7% & -2.2% I find that hard to believe. The RPI index doesn’t use new car prices for its measurement of motor vehicle inflation. Instead it samples used car prices, and creates a proxy for new car prices from these data. The reason new car prices aren’t used (according to the technical manual) is because it’s difficult to construct an indicator which monitors vehicles of constant quality over time, and also because it’s hard to determine actual prices paid due to discounts offered from list price. The rules for the CPI index don’t allow for proxy measures, so they have to use new car prices and try and make adjustments for quality changes (e.g. air conditioning added as standard when it was previously an optional extra). The index is based on the list prices of a sample of around 35 cars covering a range of manufacturers, with a trade guide used to obtain the list price. CPI shows the same sort of numbers for used cars that RPI does for all cars: -1.8% annualised over 5 years, and -2.4% annualised over 10 years. However CPI is showing a higher rate of inflation for new cars: 1.8% and 1.2% respectively. This gives an overall annualised rate for car purchase in CPI of 0.5% over 5 years and -0.1% over 10 years. In theory the RPI measure ought to be fairly accurate (if trade guides on used car prices are accurate). The CPI methodology is more open to error I would have thought, due to complex adjustments for quality and the lack of info on discounts. Quote Link to post Share on other sites
Democorruptcy Posted February 14, 2011 Report Share Posted February 14, 2011 The RPI index doesn’t use new car prices for its measurement of motor vehicle inflation. Instead it samples used car prices, and creates a proxy for new car prices from these data. The reason new car prices aren’t used (according to the technical manual) is because it’s difficult to construct an indicator which monitors vehicles of constant quality over time, and also because it’s hard to determine actual prices paid due to discounts offered from list price. The rules for the CPI index don’t allow for proxy measures, so they have to use new car prices and try and make adjustments for quality changes (e.g. air conditioning added as standard when it was previously an optional extra). The index is based on the list prices of a sample of around 35 cars covering a range of manufacturers, with a trade guide used to obtain the list price. CPI shows the same sort of numbers for used cars that RPI does for all cars: -1.8% annualised over 5 years, and -2.4% annualised over 10 years. However CPI is showing a higher rate of inflation for new cars: 1.8% and 1.2% respectively. This gives an overall annualised rate for car purchase in CPI of 0.5% over 5 years and -0.1% over 10 years. In theory the RPI measure ought to be fairly accurate (if trade guides on used car prices are accurate). The CPI methodology is more open to error I would have thought, due to complex adjustments for quality and the lack of info on discounts. Thanks for that. When I posted my reply I was thinking something like a Ford Fiesta's list price couldn't possibly be dropping year on year but then wondered if they might try "relatise" it for added features. I don't think the technical manual reasons for not including new car prices hold much water. If they used list price each year what does it matter whether a discount was secured or not? It seems to me like cars are something used to give a false inflation figure. I was deflated at Asda today it was 8 salmon fillets for £9.97 now just 6 and the previously £1.37 coffee is now £1.78 Quote Link to post Share on other sites
OnlyMe Posted February 14, 2011 Report Share Posted February 14, 2011 The RPI index doesn’t use new car prices for its measurement of motor vehicle inflation. Instead it samples used car prices, and creates a proxy for new car prices from these data. The reason new car prices aren’t used (according to the technical manual) is because it’s difficult to construct an indicator which monitors vehicles of constant quality over time, and also because it’s hard to determine actual prices paid due to discounts offered from list price. The rules for the CPI index don’t allow for proxy measures, so they have to use new car prices and try and make adjustments for quality changes (e.g. air conditioning added as standard when it was previously an optional extra). The index is based on the list prices of a sample of around 35 cars covering a range of manufacturers, with a trade guide used to obtain the list price. CPI shows the same sort of numbers for used cars that RPI does for all cars: -1.8% annualised over 5 years, and -2.4% annualised over 10 years. However CPI is showing a higher rate of inflation for new cars: 1.8% and 1.2% respectively. This gives an overall annualised rate for car purchase in CPI of 0.5% over 5 years and -0.1% over 10 years. In theory the RPI measure ought to be fairly accurate (if trade guides on used car prices are accurate). The CPI methodology is more open to error I would have thought, due to complex adjustments for quality and the lack of info on discounts. Hedonic adjustment. What a load of ********. A base model hatchback is a base model hatchback. Mid range Ford, is a mid range For, likewise for executive. Just compare ranges / brands like for like. The utility value for each is almost the same it is a car, it is deisgned to get poeple from A to B, mostly work. Quote Link to post Share on other sites
InternationalRockSuperstar Posted February 14, 2011 Report Share Posted February 14, 2011 I'm just a bit curious as to how housing cost weighting has increased by 20% over the past 10 years, when housing inflation has exceeded general inflation only marginally over the same timespan. Any ideas? Edit: Housing weighting up from 19.5% to 23%+, yet housing cost inflation was 3.7 versus 2.9. Just seems a little odd. maybe because RPI is just a number that a government statistician pulls out their ****? Quote Link to post Share on other sites
onesmallstep Posted February 14, 2011 Report Share Posted February 14, 2011 thankyou for this info. the thing that stands out is the cost of energy, surely the cost of energy is going to double every 7 years. Wages will double every 20 odd years. I would be interested in the rate of change of a products inflation as some appear to be gathering pace, can we expect them to level out or drop back down. I suspect not, this would mean that we in this country will sooner or later have to source our food in a different way. Quote Link to post Share on other sites
FreeTrader Posted February 14, 2011 Author Report Share Posted February 14, 2011 Somehow this doesn't seem to work out: The total basket price has risen by 2.9% pa over the past 10 years, or from 1000 to 1331 Out of this, housing costs have risen 3.7% pa, or from 195 in 2000 to 280 in 2010, implying a weighting increase from 19.5% to 21.0%, yet in the 2010 stats the weighting of housing costs is now given 23.8% Seems that housing costs are now over-represented in the RPI All Items basket. I think you’re assuming a constant pattern of expenditure. Just because something isn’t inflating as fast as other prices, it can still consume a growing proportion of consumer spending and thus receives a higher weighting on annual adjustment. For example, telephone/telemessaging prices have fallen 11% since 1988, but the weight has increased from 16 to 23 (the weight started increasing when the mobile spectrum came into use). The weights don’t change to reflect price movements, they change to reflect the relative proportion of expenditure by consumers on an item in relation to total expenditure (but these dynamic ratios are influenced by price changes of course). Quote Link to post Share on other sites
FreeTrader Posted February 15, 2011 Author Report Share Posted February 15, 2011 I get what you're saying, but if weighting increase isn't down to price, it must be down to increased volume of consumption. I'm not sure there are that many more people making housing cost purchases, unless we are talking about smaller household sizing and other effects like that (although those trends seem to be reversing along with house prices) Yeah, I appreciate your point, but it’s not easy to determine any disparity without having access to the ONS expenditure figures. FWIW, here are the weight changes since 2000: The rent weighting has increased considerably since 2006, and we might have expected the mortgage interest payment weighting to be lower. By the way, you may or may not recall that the ONS issued a consultation document in 2009 which proposed to change the way that mortgage interest payments are calculated in RPI. They also had to consult the Bank of England when they did this because such a change might be considered materially detrimental to the holders of IL gilts. Anyway, it was pretty obvious that this was a complete stitch up, not so much with the rationale for the change, but because they were making the change at the most advantageous moment for the Government. The consultation document was called "Measurement of Mortgage Interest Payments within the Retail Prices Index". This document is listed on the ONS’s Closed Consultations page, but if you click on the link it says it has been removed. The response document has also been removed, although a cached version is still available through Google. Quote Link to post Share on other sites
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.