Those of you who regularly complain/mock/kvetch about the BLS methodology for measuring CPI prices — and I am as guilty as anyone else — should check out the “Billion Prices Project @ MIT.”

The idea behind the Billion Prices Project is that we can track inflation by collecting prices from hundreds of online retailers around the world on a daily basis. Thus, the  BPP currently monitors the daily price fluctuations of ~5 million items sold by ~300 online retailers in more than 70 countries.

That’s a pretty cool way to track real time inflation.

At the WSJ, Justin Lahart quotes Rutgers University economist Michael Bordo, who has reviewed the methodology: “It seems to me it’s a brilliant way of measuring the deep fundamentals of inflation.”

There are some caveats: The economists’ indexes are “tracking prices paid by the wealthy, who shop more at stores with an online presence, rather than ordinary shoppers.” And, it cannot pick up items whose prices are subject to haggling — cars and electronics come to mind. Services like legal, accounting, education, health care don’t exactly post a price list in the cloud either.

Justin adds that so far, the MIT approach has tracked the official price data fairly well, but it might stray in the future.

Meantime, having a real time inflation measure is a nice addition to the tool box — especially if we can measure the impact of QE2 . . .


MIT Sloan professors publish real-time inflation rates around the world in “Billion Prices Project”
November 8, 2010

See also: Daily price indexesMethodologyResearchAbout us

A Way, Day by Day, of Gauging Prices
WSJ, NOVEMBER 10, 2010

Category: Data Analysis, Inflation, Retail

Please use the comments to demonstrate your own ignorance, unfamiliarity with empirical data and lack of respect for scientific knowledge. Be sure to create straw men and argue against things I have neither said nor implied. If you could repeat previously discredited memes or steer the conversation into irrelevant, off topic discussions, it would be appreciated. Lastly, kindly forgo all civility in your discourse . . . you are, after all, anonymous.

39 Responses to “Inflation? Not According to a Billion Prices”

  1. Mannwich says:

    Am certainly not seeing higher prices in discretionary retail, but am quite certain I AM seeing higher prices in food, gas, education and health care, but what do I know? Maybe we’re getting deflation in the things we don’t need and price inflation in the things we do, but I don’t compile stats for a living, so, again, what do I know? Maybe it’s just MY region that’s experiencing these things but I doubt it.

  2. constantnormal says:

    GOLLY !! Somebody else who collects actual data, and on a far grander scale than myself. Glad to see that it comes to the same conclusion that my data shows — otherwise I would have had to struggle with changing my mind.

    But it will simply bounce off the mindsets of those who have their minds already fixated on defending their baseline biases …

    But as to how much of an impact QE2 is likely to have … what was the conclusion regarding the impact of QE1?

    Wasn’t it something like 0.17%?


    Rumors of the death of deflation are greatly exaggerated. At best it is merely biding its time, until all the over-stretched debt implodes and we see a tsunami of defaults from pensions, corporations, banksters, cities, states and nations. Or perhaps it will work in the other direction, emanating from the PIIGS.

    The likelihood that “muddling through” will succeed is not high.

  3. constantnormal says:

    I should add that I believe we will eventually see inflation, but only in the wake of a deflationary collapse, as currencies will fall faster than commodities, and wages fall faster still. I think that’s a ways down the road. Probably after the 2012 elections.

    And this is completely emotion-driven expectation, I am watching for any signs of it becoming reality. So don;t take it as a statement from me about any current inflation, as I believe there is none. Someday, prolly so, but not now.

  4. constantnormal says:

    Interesting to bounce this data regarding prices where the “rubber meets the road” with the swarm of charts showing commodity price inflation. Somewhere in the middle, there is some serious squeezing going on. But it is people who are being squeezed, with the impact squirting out into the public at large, resulting in less demand, which helps keep prices to the end consumer down.

    Eventually, this pain will travel upward through the income statements, impacting margins of retailers and manufacturers, and the diminished sales volume will begin to clobber commodity producers (assuming that the third world is unable to cover the shrinking first world demand).

    That does not bode well for stocks in general, at least at some point, no matter how much QE the Fed pumps into the system. But before we see any correction in equities, I think that valuations are going to get simply INSANE.

  5. constantnormal says:

    Game on.

  6. b_thunder says:

    Don’t know if a lot of people purchase rice and beans and potato online in Colombia. And especially Aguardiente.

  7. Invictus says:

    Have often thought that PDA barcode scanning apps would be ideally suited to the task of more and better real time price tracking. Crowd-sourcing of price data on virtually everything under the sun. Only need a repository/aggregator.

  8. Jojo says:

    But they are only tracking LIST prices! How many people actually complete a transaction and buy a product (nothing happens until the sale is made is an old sales adage)?

    When I am hunting to buy something online, I always find a wide range of prices for the same product. And of course, I always buy from one of the companies near the lower price end.

    Many companies show prices that are far greater than the prices of other vendors that I wonder how they continue to stay in business.

  9. surfboy says:

    BR: What do you make of Grantham’s CNBC Money Honey interview comments?

  10. constantnormal says:


    Quite right. If you dig around a bit, you can find aggregated sales data collected by the states –but this only shows aggregate amounts of purchases, and says nothing about prices. It does show quite clearly that consumer purchases are gliding lower, confirming the data on consumer spending at the national level. Some of the states are experiencing quite severe cutbacks in consumer spending.

    I’m pretty sure that A.C. Nielsen has an EXCELLENT handle on the direction of prices paid by the consumer, but unfortunately, they are not likely to give that information away for free. Also the credit card companies.

    The data is out there. Somebody knows. The rest of us are merely blind men attempting to define the elephant by touch, never having seen one before.

  11. DeDude says:


    I think you are right that we are having inflation in the things we need (partly driven by Fed funded speculation in commodities), and corresponding deflation in the things we don’t need (partly driven by the fact that consumers don’t have a lot left after they have purchased the things they need). When you average it out it looks like price stability. Current Fed policy is pretty ineffective because the “stimulus” from low rates and free money is countered by the “anti-stimulus” from speculation driven increase in prices of commodities based products. If they really wanted to stimulate they should puncture the current speculative bubble in commodities. The associated price fall in commodities based products would give a needed transfer of wealth from speculators to consumers and help create the inflation that this debt-ridden country so desperately needs.

  12. doug says:

    Doesn’t this leave out much of the same stuff as the CPI?
    I doubt a significant amount of food, gas, education and health care is sold online.
    As unrepresentative as the CPI based on more observations. Big Whoop. Still distorted…

  13. Jojo says:

    Secret Walmart Survey Shows Inflation Already Here
    Published: Thursday, 11 Nov 2010 | 2:59 PM ET
    By: John Melloy

    There might not have been a second round of quantitative easing, if Federal Reserve Chairman Ben Bernanke shopped at Walmart.

    A new pricing survey of products sold at the world’s largest retailer showed a 0.6 percent price increase in just the last two months, according to MKM Partners. At that rate, prices would be close to four percent higher a year from now, double the Fed’s mandate.

    The “inaugural price survey shows a small, but meaningful increase on an 86-item grocery basket,” said Patrick McKeever, MKM Partners analyst, in a note. Most of the items McKeever chose to track were every day items like food and detergent and made by national brands.


  14. Kort says:


    There’s inflation in the food, and the clothing and the bottom 20%, nee 80%, spend more of their income on the basic necessities than the upper crust buying $100K cars and $500K “boats”. Pain, inbound.

  15. nickthap says:

    @ Mannwich and dedude

    Education and health care costs have risen three times the CPI for the last 20 years. Do a search and you can find that info (CPI 1990-2009 71%, Education 274%, Health Care 250%) easily.

    Not sure how you can use the examples of education and health care to “prove” that QE2 is somehow creating inflation in these sectors right now. Housing costs are deflating, and this is arguably the top cost many of us have.

    The obvious perils of confirmation bias Mr. Ritholtz is always harping on are no more apparent than the absolute certainty that abnormal inflation is upon us when what we really should fear is deflation. This is why the Fed actually seeks some modest inflation, to avoid the deflationary spiral that hit Japan in the ’90s.

  16. gordo365 says:

    measures the price of anything that fits in a container ship. This approach proves that offshoring works. So what.

    Greenspan fell for this – and dropped rates too far too long to avoid this type of “deflation”. Fool me once…

  17. jus7tme says:

    Scouring the web for pricing data is a great idea, but it does have some limitations:

    1. it does not measure actual transactions prices (nor does CPI, I suppose)

    2. it does not measure food, energy and other consumables

    3. it does not measure brick-and-mortar prices.

    One of the reasons I mistrust CPI is that I do not know, admittedly, how they collect prices on food. Do they shop around, do they look for bargains, or is it all list prices from store shelves? List prices on store shelves is not necessarily the same as what customer actually pay. List prices may only increase by minor amounts, but the absense or presence of sales and discounted items may strongly affect what consumers really pay, and that is what the real price level.

    In other words, until I get to mine the internal database of store sales transactions, I don’t know whether I trust the numbers.

  18. jus7tme says:

    Also, web shopping is for many items that come and go, such as different models LCD televisions or other consumer electronics. Electronics is in deflation because for the last 100(?) years the real price/functionality has been on a never-ending decline.

    But what about Beef, Chicken and Pork? The functionality of a slab of beef has not gotten better, some people might argue it has gotten worse due to hormones and farming methods. How do you account for that?

    Maybe what I’m doing here is to argue for a Case-Shiller-style of consumer price research methodology. That is, a methodology that considers actual purchases and prices, and adjusts for quality/functionality in some way.

  19. JimRino says:

    Do you mean Sarah Pay-lin will have to find something Else to get Hysterical about?

  20. JimRino says:

    As for oil it’s down, from it’s high of $138.
    And food is also down from it’s high in 2008.
    So, that’s deflationary.

  21. constantnormal says:

    Yesterday’s Breakfast with Dave (Rosenberg) market letter had an interesting pair of bar charts, showing a smattering of real personal consumption expenditures, one showing the annualized 90-day percentage change in the volumes, the other showing chain price indices. Here are the numbers:

    volume changes – Real Personal Consumption Expenditures
    Casino 40.2%
    Motor vehicles 11.4%
    Books 10.9%
    Restaurants 6.3%
    Nursing Homes 4.9%
    Personal Care Products 3.5%
    Telecom Services 3.1%
    Hospitals 1.7%
    Education Services 0.0%

    price chain index changes – Real Personal Consumption Expenditures
    Personal Care Products 3.4%
    Motor Vehicles 3.0%
    Casino 2.7%
    Hospitals 2.3%
    Education Services 1.8%
    Restaurants 1.2%
    Telecom Services 1.2%
    Books 0.2%
    Nursing Homes 0.2%

    I thought the change in volume in Casino spending was illuminating. Apparently Joe Sixpack is taking the lessons of Wall Street to heart.

  22. The problem with this analysis (and the CPI) is that rising prices aren’t always indicative of inflation, and declining prices aren’t always indicative of deflation. If demand is falling, yet prices hold steady, that’s inflation just as much as when demand increases only a bit and prices sky-rocket (commodities just now).

    Everyone seems so concerned with a “deflationary spiral” such as Japan has experienced. Though Japan’s aggregate economic performance has languished over the last two decades, her per capita income has remained stable. Why? Because Japan’s deflationary spiral is really a demographic implosion. It has far less to do with money than it does with demand. Indeed prices have declined and continue to do so. But so long as gross economic output holds steady, the individual Japanese, on average, is gaining by dint of the price declines.

  23. Mannwich says:

    @nickthap: Of course QE2 can’t be causing those things to go up all of a sudden now, but it can be surmised that the Fed’s policies over the years and decades have been at least a contributing factor in these price increases, no?

  24. baychev says:

    you should be taking such number compilations with a hefty grain of salt:
    - this measures offered prices in often very limited quantities;
    - online stores do not cover even 20% of all sales;
    - many products at discounted prices are refurbished;
    - the product mix is heavily scewed towards discretionary spending;
    - shipping is not included in this index;
    - many times the item when you click on the link is a totally different thing, erroneously listed, especially on price comparing sites;
    - discontinued models are heavily discounted, etc.

  25. baychev says:

    and for the guys that are claiming there is no inflation (like Invictus): you cant have it all at the same time – growing corporate profits, increasing input costs, modestly increasing wages, growing GDP, not shrinking labor force.

    it looks to me that the lack of capital expenditures coupled with amortized equipment and employee layoffs have contributed quite a bit to the rise of corporate profits, but is this path sustainable and for how much longer given tepid demand and rising cost pressures. maybe we will see a significant slowdown in corporate earnings over the next year and markets will tank regardless of Ben’s QE63.

  26. constantnormal says:


    - this measures offered prices in often very limited quantities; — and the quantities available at the local retail outlet are not “very limited”?

    - online stores do not cover even 20% of all sales; — however, it is very rare that the online store pricing differs from that of the same company’s brick-and-mortar stores prices by very much

    - many products at discounted prices are refurbished; — and are advertised as such. I would be surprised if the software conducting this survey did not treat these as different products, just as a new BMW/Chevy/Ford/Honda is priced differently from the used models on the lot (both of which appear in the web ads for cars)

    - the product mix is heavily skewed towards discretionary spending; — so what?

    - shipping is not included in this index; True, and many products offer free shipping, such as Amazon’s free shipping for orders over $25. But for a lot of small-ticket items, the shipping costs come close to or even exceed in some cases, the cost of the product. But web purchases of those products are almost always because they are unavailable locally. Unless the shipping rates are changing at a faster rate than the product pricing, I’m not sure this puts a spin one way or the other on the trend results. All it says is that the pricing data obtained in this manner is inexact — it says nothing about the validity of the observed trends.

    - many times the item when you click on the link is a totally different thing, erroneously listed, especially on price comparing sites; Yeah, yeah, and “many times” the items in a brick-and-mortar store are mislabeled. So what is your point?

    I think that unless you are A.C. Nielsen or a major credit card company, any compilations of price data used to infer changes in consumer outlays should be taken “with a grain of salt”. Does that render the charts displayed wildly in error, in terms of the changes from year to year? If so, please explain why the items you have listed impart a bias to the observed trends?

    The intent of this tool is not to show precise pricing data — as has already been mentioned in this forum, it only shows listed prices, not actual sales. So by definition it is an approximation — that should be obvious. But approximations can reveal trends just as well as precise data.

    While it is possible that the bulk of the items noted represent phony prices for discontinued or unavailable of unpopular items that generate very few sales, logic tells us that this is not likely to be the case, there are costs (small costs, but costs nonetheless) to keeping large product catalogs packed with items that are not actually available. While from time to time I encounter items that are listed but unavailable, this is by far the exception rather than the rule.

  27. constantnormal says:

    @baychev — I quite agree, we can’t have it both ways indefinitely. The growing corporate profits are not accompanied by corresponding revenue increases. Sooner or later, they are gonna run out of people to lay off, or rising raw materials prices (especially for imported raw materials, where the shrinking USD is putting the hurt on corporate American’s purchasing departments) will put an end to the rising profits. And with the markets seeing rising valuations, things could turn very ugly very quickly. Just take a look at CSCO’s performance today.

    There are a lotta companies out there with sky-high PEs (Netflix?), and implied growth rates that would be nigh-unto-impossible in the best of times. When the tide goes out, it’s gonna get ugly.

    I just don’t know when the tide will go out, with Bernanke pumping and pumping and pumping some more …

  28. Mannwich says:

    @constant: Disney missed on both top and bottom lines today quite badly. I think your scenario is playing out.

  29. rktbrkr says:

    I think the Walmart survey is the most accurate, 4%/year easy.I generally shop online as much as practical but it isn’t real world. How do you account for the consumer scraps dealers on ebay?

    I see a case of Bud go up $1 – 5% inflation there.

    I see a roll of toilet paper so small I could put it in the front pocket of my jeans

    The Bernanke family tree must have roots in Weimar republic.

  30. VennData says:

    What’s really going up in price in the comment zone are tin foil, for their hats.

    Anybody out there heard of owner’s equivalent rent? Oh right, you bought a house so’s you could be part of the “Ownership Society.” Whoops.

  31. The key issue may not be the “real” % of inflation (about which no one can agree), but rather the more fundamental question: Does federal deficit spending cause inflation? Matching the above chart against U.S. deficits shows there is no relationship.
    Also, graphing the Fed’s measure of CPI shows no relationship with Inflation
    So whether inflation is 1% or 8%, the answer to the question, “Does federal deficit spending cause inflation?” seems to be No, at least not in the last 40 years.
    So what has caused inflation? Oil prices.

    Rodger Malcolm Mitchell

  32. royrogers says:

    food, gasoline, rent, how important are those items in this graph ??

    which are important for the working poor in terms of inflation

  33. SCTTD says:

    Two comments.

    1. Packaging sizes have shrunk and I do not believe the government statistics are picking this up.
    2. http://www.shadowstats.com/alternate_data/inflation-charts

  34. Jojo says:

    Here’s a simple search for a Canon G11 camera. Prices range from $329 to $524.95. Do you pay $329 from “saveheredigital” or $439 from Ryther Camera or $524.95 from Office Depot?


    Scanning the net for prices doesn’t prove much of anything. We need real transactions.

  35. The missing dynamic in this inflation thread encompasses the globe. Remember that thing called “globalization?” Well, there’s this financial element involving hundreds of billions of dollars in hot money flows and the recycling of these. Largely as a function of circumstance, first allowing, then demanding, the creation of ever-increasing quantities of mispriced risk the point has been reached where an illusion of “excess supply” in economic capacity is made to appear vexing. So, some trillions are thrown into a yield hungry wind and, suddenly, commodity input costs escalate, margins are squeezed and “excess supply” increasingly disappears out of necessity. Again, pressure toward this end reaches globally.

    In other words, the domestic inflation v. deflation debate generally excludes consideration of dynamics involving hyperinflation (and this now on a global scale) — a condition only possible when a mountain of [largely illegitimate] liabilities are made the master of debtors, rather than their servant. Increasing scarcity — jobs, housing, food, energy, domestic tranquility, you name it — have been, and will continue being, served up one way or another. QE — but the next act in the post-Bretton Woods age of wildcat finance — can only accelerate this course, and so its greatest probability likely will maximize wealth destroyed. Judging by widespread consternation over QE, this inflation-deflation debate taking a pre-globalization perspective assuming a national economic setting that is well-sheltered from foreign contagion is soon to receive its wake up call.

  36. not-affiliated-with-Wall Street says:

    Yeah, a billion items excluding food and energy. Useless.

  37. druce says:

    Consumer goods are paid for by credit card every day…it would clearly be possible to harness SKUs and location info and get practically real-time inflation measurement.

  38. You can argue ’til the cows come home, about how to measure inflation. The fundamental point is, however you measure inflation, it does not seem to correlate with federal deficit spending. Oil pricing does correlate with inflation.

    Rodger Malcolm Mitchell

  39. [...] Inflation? Not According to a Billion Prices (The Big Picture)–Inflation is such a hot topic that even Sarah Palin is chiming in on the matter.  Yet is there really cause for alarm?  Many question the validity of the CPI, so with that in mind, the Billion Prices Project decided to dig into a deeper database of price levels around the world.  Ah the wonders of the web, it just makes information so accessible.  Sure enough, by just about any metric, we are not suffering from an inflation problem. [...]