Wealth Manager

The CPI, in short, packs a punch. But while no one disputes its
entrenched status, there is a lack of absolute consensus when it
comes to the index’s reliability. Indeed, there’s a vigorous discussion
in some circles about whether CPI is an accurate measure of
inflation.

Given its reach into people’s wallets, the stakes in the debate
are more than academic. Assume for a moment that the measure
is flawed by more than a little by understating inflation, as some
critics charge. If so, investors hedging inflation with CPI-indexed
Treasuries (and some inflation-linked munis and corporates) are
exposed to more risk than is apparent. That may not seem like a big
danger these days, considering the reportedly low level of inflation,
as per CPI, of course. But if inflationary momentum rises in the
future, so, too, will the risk of relying on a flawed benchmark.

Unfortunately, hedging inflation with strategies that sidestep
CPI-linked instruments isn’t always practical for individual investors.
Commodities, for instance, are said to be an alternative
defense against inflation. That’s debatable as well, but even if
it’s true, there’s a catch. Commodities can and do move independently
of CPI at times. Owning raw materials (and their proxies)
means holding an inflation hedge that may not seem to be working
at times, relative to the CPI. The Dow Jones-AIG Commodity
Index, the basis for several mutual funds and exchange-listed
securities, has suffered fairly lengthy stretches of losses in the
past while CPI continued to rise. Commodities may be inflation
sensitive, but the asset class is subject to speculative forces,
which may or may not boost its inflation-fighting credentials in
the short run.

So while some may question CPI’s merit as an inflation gauge,
most investors are stuck with it. Until and if there’s a better alternative
for non-institutional investors, CPI-linked TIPS remain the only
game in town for directly immunizing a portfolio against inflation.

That’s just fine for most investors, or so a casual observer
might conclude based on the market’s embrace of Treasury
Inflation Protected Securities whose principal is directly tied to
CPI. Critics can rail about CPI’s flaws, but there’s scant evidence
that the complaints are taking a toll.

Since the market’s launch in 1997, TIPS’ outstanding value
has grown to $411 billion, as of 2006’s close, according to the
Treasury Department. That’s still a fraction of the conventional
Treasury market, but the 10-year-old TIPS market has been growing
quickly. The outstanding dollar value of inflation-indexed
Treasuries nearly doubled over the previous three years, for
instance. Meanwhile, investors can choose from 35 mutual funds
and an ETF that target inflation-protected bonds, according to
Morningstar Principia.

Compelling or not, the CPI is the standard for general price
trends. But do investors really understand the index, including
how it is designed and the resulting strengths and weaknesses?

Like any gauge that claims to quantify a complex and nuanced
economic phenomenon, the task of measuring inflation involves
a fair amount of subjectivity. Consider the evolving mix of factors
that collectively determine CPI. There are eight major component
groups in the index, tracking prices across a wide variety
of products and services. In addition, the weightings of those
eight groups continually change, as per the U.S. Bureau of Labor
Statistics, which sees fit to adjust the design from time to time.

Constancy, in other words, is not part of CPI’s design. At the
close of 2006, housing had the biggest weight of the index’s eight
major groups at nearly 43 percent, up sharply from the end of 1997,
as the table on the next page shows. At the opposite extreme was
apparel, whose influence on CPI fell to 3.7 percent at last year’s
close, down from a weight of more than 16 percent in 1998.

The underlying methodology that defines CPI is also a moving
target. In fact, debates about formulas break out from time to
time, including one in 2004 when Bill Gross, manager of Pimco
Total Return—the world’s largest bond fund—wrote that CPI was
a “con job.” In his monthly letter to clients, Gross charged that
CPI understates inflation by 100 basis points. The use of what’s
known as hedonic adjustments is the culprit, he counseled. That’s
a statistical treatment that factors qualitative changes into prices.
As Gross sniffed, “The government says that if the quality of a
product got better over the last 12 months, that it didn’t really go
up in price and, in fact it may have actually gone down!”

A parallel dispute has arisen in recent years over the details for
measuring real estate prices in the CPI. In the 1990s, CPI’s design was
altered so that housing costs were redefined as rental prices. The
current formula incorporates the so-called owners equivalent rent
index, which is calculated from homeowners’ reports on what they
can charge for rent. The formula has been criticized because it softpedaled
the recent boom in actual prices paid for homes. The bull
market in real estate over the past several years wasn’t captured in
the CPI, critics say, leading to another round of complaints that the
government is minimizing the true extent of inflation’s pace.

“The current CPI understates inflation as the average person
would look at it,” charges John Williams, an economics consultant
who also edits the Web-based newsletter ShadowStats.com,
which analyzes data published by the government. By his reckoning,
CPI’s official rise of 2.4 percent for the year through this past
February was really over 5 percent—if the index was calculated
under the old, pre-1990s revision. “As a result, the cost-of-living
adjustments are shy,” says Williams. “TIPS don’t give you an
adequate return on your money if you’re looking to at least keep
your capital base neutral in terms of being adjusted for inflation,
or even making a little money on it.”

Williams may not represent mainstream thinking, but he has
some support in concept—if not in detail—from some surprising
corners. The Federal Reserve Bank of Cleveland publishes an
alternative measure of general inflation that’s been running well
above CPI of late, although not as high as Williams’ estimate. For
the year through February, for instance, the bank’s median CPI
rose 3.6 percent vs. 2.4 percent for the standard CPI.

Nonetheless, the financial establishment and most dismal
scientists generally accept CPI at face value. The CPI “pretty
much gets it right,” says Jared Bernstein, an economist with the
Economic Policy Institute in Washington, D.C. There are a lot of
arcane discussions about its underlying methodology, but all
such measures are educated guesses at best, he counsels.

Lakshman Acuthan of Economic Cyclical Research Institute, an
economics consultancy in New York, also says that CPI is fine as
far as it goes. But it’s hardly perfect, he adds. In fact, most if not all
of the government’s measures are informed estimates, he notes.
As for the CPI, “In theory it should measure the change of a basket
of goods that the average person would consume. It probably
doesn’t do that perfectly.”

The CPI’s subjectivity should be recognized, but it’s not fatal,
Acuthan explains. In fact, CPI’s primary value (at least from an inflation-
forecasting perspective) flows from changes in the index’s
broad trend, which offers clues about the future path of prices.
When the general direction of the CPI reverses, that may be a crucial
signal about what’s coming. Arguing whether the index is rising
by 2 percent or 3 percent is trivial by comparison, he suggests.

Precision does indeed seem to be one of CPI’s lesser attributes.
The more people you talk to about the index, the wider the viewpoints.
Lee Hoskins also says CPI is subjective, but the former
Cleveland Fed president tells Wealth Manager that he’s convinced
the index overstates inflation by as much as 1 percent. Inflation,
in other words, is lower than the government tells us. “It’s not a
perfect measure, but it’s the one that’s most recognized, and it
seems to me to be a perfectly acceptable [inflation] target,” says
Hoskins, a senior fellow at the Pacific Research Institute and an
advocate for a monetary policy that aims to keep CPI to a zeropercent
inflation target over time.

CPI, for all its influence on real dollars, is but one gauge in a
sea of alternatives. Perhaps it’s telling that the Federal Reserve
reportedly favors personal consumption expenditures (dispensed
monthly by the U.S. Bureau of Economic Analysis) as its preferred
measure of inflation. The reason is the design. PCE tracks consumer
decisions, and so inflation derived from this index reflects
the human factor. CPI, by contrast, reflects prices drawn from a
fixed basket of goods and services.

“The CPI measure is based on a constant basket of goods and services
whereas the PCE is based on actual consumer expenditures,”
explains Tom Higgins, chief economist of Payden & Rygel, a Los
Angeles-based money manager. “The reason economists prefer the
PCE is because it’s a moving basket and it’s a more accurate reflection
of people’s experience with inflation.” Consumers, in other words,
don’t limit purchases to a fixed basket of goods. If steak becomes
expensive, shoppers may switch to hamburger. In fact, trying to capture
the human factor in inflation measures is the rationale for CPI’s
hedonic tweaks, albeit tweaks that Gross criticized.

No matter which inflation measure you prefer, the broadly
designed ones all suffer from what might be called a macro bias.
An inflationary climate for one investor may be disinflationary—
or even deflationary—for another.

A 65-year-old woman of modest means with a recurring out-ofpocket
prescription medicine bill is likely to have a different view
of inflation than a single, healthy, 25-year-old with a budding
white-collar career. Indeed, medical care inflation rose by 4.3 percent
in the 12 months through February, or nearly twice as much
as inflation generally, according to CPI.

Like the proverbial blind men describing an elephant, everyone’s
inflation is different. True, but it doesn’t change the fact
that all roads for hedging inflation lead back to the same index.
One size fits all, like it or not.