A Discriminating Eye On CPI
August 4, 2005: Today's Events:
Thursday was a day of waiting, ahead of Employment on Friday. As we eastcoasters sweltered in oppressive heat and humidity (especially those of us who were schlepping around New York City in suits for part of the day) and wondered aloud about the capacity of the power grid to keep our air conditioners going, the bond market traded with the syrupy lethargy apropos of the weather. Treasuries sold off 2bps (10y: 4.316%) on the session, but TIPS were relatively firm and breakevens widened by 1-2bps.
Another beating of the greenback helped to push energy prices back to challenge recent highs and, in the case of Nat Gas, to exceed them. This, along with weak retail sales reports ex-Wal Mart, created a somber mood among equity traders, who shaved ¾% off the value of the S&P 500 on volume that was again reasonable at 1.45bln, but has now turned 5:2 in terms of down volume to up volume and 3:1 in terms of decliners to advancers. With oscillators topping, bulls had better hope for a good report from BLS tomorrow.
CPI and Price Discrimination:
CPI is supposed to measure the change in the price of a particular standard of living. However, and this is key, it is not supposed to measure the change in the price of your standard of living, or indeed of any given person's standard of living. It is not supposed to measure what it costs to remain in the middle-middle class, or what it costs to remain in the upper-middle class; all of these are different numbers. What it is supposed to measure is what it costs to maintain the same standard of living - meaning, you're buying the same stuff with the same features over and over.
No one lives like this, but that's not the point. The point is that the price index is supposed to capture changes in prices, not changes in values. Clearly, more valuable things cost more.
Part of the perceptual problem for people such as our readership, denizens
of Wall Street, and folks who have enough at stake in the markets to bother
with CNBC is that you guys are not the sorts of folks who buy Ramen Noodles.
You buy better stuff, because you can afford better stuff, when it comes on
the market. You replace old cars. You buy a new computer when the old one seems
slow. You pay Microsoft's
extortion licensing fees to upgrade to new
software. You eat at nice restaurants, and many of you own your homes rather
than rent. So when I tell you that Ramen Noodles have gone from four-for-$1
when I was in college, to six-for-$1 when my colleague Ronit was in college,
to eight-for-$1 today, you might be surprised at the deflation. You might also
be surprised to hear that the price of a loaf of generic bread, currently $0.89
at our local, overpriced King's Supermarket, has not budged a penny since I
was in school in the '80s in Texas. Leisure airfares, if you're willing to
fly during off-peak hours to destinations during their off seasons, are lower
today than they were ten years ago. Those are just anecdotes, of course, but
they have in common the fact that those items are not in the market basket
of folks who live in the Hamptons.
I think that this is another phenomenon of modern marketing - price discrimination has become a science. Recall from college economics that price discrimination describes the situation in which buyers with elastic demand curves (they're very price-sensitive) are charged lower prices than buyers with inelastic demand curves. Business flyers, because they're not paying for themselves and so are willing to pay up for the right times, seats, and airlines, always pay higher fares than coach class. Grocery stores in the good area of town always charge more than the grocery stores in the bad area of town, even for the same chains and even though the latter grocery store has more theft losses and higher security costs. And, as we get better and better at tracking customers, the real-time demand for airline seats and other products, inventory, etcetera, the number of different prices that can be charged and the potential gains from pricediscriminating could be quite large.1
What is the effect of behavior like this on price indices such as CPI? Well, if you capture the average price increases correctly, then you'll have the more wellheeled customers continually grousing about inflation being understated. The fact that we have just such a phenomenon doesn't prove my thesis, but it doesn't hurt it either. It's something to think about.
Everybody is ginned up for the monthly Payrolls data, which hits tomorrow at 8:30ET. Consensus calls for the Unemployment Rate to remain at 5%, for Average Hourly Earnings to rise +0.2%, unchanged weekly hours at 33.7, and a relatively tame 180K rise in Payrolls. The 3-month and 6-month averages for Payrolls are both 180,000, so this isn't exactly a strong vote of confidence although some observers are pointing to a possible hurricane effect. But the 180,000 average doesn't capture the interesting part of the data, which is the sawtooth: 124K, 300K, 122K, 292K, 104K, 146K. By rights, we're probably overdue for another number above 200K at least, especially if the economy was really surging last month, so I'm a bit guarded. A print in the mid-200s, combined with a decline in the Unemployment Rate to a 4%-handle, would be enough to shake the confidence of bond market bulls just a few days prior to the FOMC meeting.
The equity market seemed to establish today that the longs are a bit weakerhanded than bulls would probably like to believe; the bond market really didn't establish very much one way or the other.
While I would be a fool to think that the immediate path of the markets won't be dictated by the economic figures tomorrow, I continue to think stocks are "done" - absent a very strong jobs figure (in which case people will start to worry about more-aggressive Fed action next week), I doubt equities can make much headway.
With bonds, I still think the market is trying to consolidate. But while I think stocks would have much trouble moving very far north, I don't think bonds will have trouble moving south if the economic data is strong. My sense that this level of yields is a reasonable place for bonds to find some footing isn't getting the same sort of support from the price action as is my contention that stocks are done. A strong piece of data could crash notes to new lows. I'm not anticipating strong data, and indeed a status-quo employment report, in my view, is bullish...but if bonds do start heading lower I am less anxious to be a buyer immediately.
Question of the Day: Twenty people are in a room. Each man weighs 300 pounds; each woman weighs 150 pounds; each child weighs 50 pounds, and as a group the twenty people weigh 2,000 pounds. How many men, women, and children are in the room?
Answer to Prior Question: The question was, "What does the 'i' in the molecule RNAi stand for?" The i is for "interference"; see http://www.pbs.org/wgbh/nova/sciencenow/3210 /02.html for more.
1 Selling on the Internet is a gold mine for this sort of approach, and although it does not appear that most on-line vendors utilize this information well, you might make a reasonable argument for why, if Amazon really can track the buying habits of its customers and re-price books accordingly, it actually could be worth more than Barnes & Noble. Not 6.5 times more, as the current market caps have them after BKS returned 66% last year to AMZN's 22.5%, but a bit more. Of course, to make this argument you'd have to see AMZN's margins actually higher than BKS's, and in fact they're a good deal lower.