By George Strickland
Inflation is dangerously close to becoming uncontained. In fact, inflation is probably already understated by the Consumer Price Index.
How much has the CPI been understated? I do not have the exact answer — that would require a much more detailed study than I am able to undertake. But I do have a very simple way of estimating the potential magnitude of the problem.
I estimate the rate of change in the CPI reported over the last five years has been understated by as much as 1.43 percentage points, thereby lowering the real rate on bonds of all types. Without this understatement, Treasury bonds appear to be significantly overpriced.
Investors have always been mindful of the threat inflation poses to fixed-income portfolios. They should be skeptical when viewing CPI data.
A major culprit in this understatement of inflation is the use of subjective data to impute housing costs. A data series embedded in the CPI, called owners' equivalent rent of primary residence, makes up 22% of the index. It is the CPI's largest single component. The Bureau of Labor Statistics started using this component in 1983 because, in its words, "the asset price method can lead to inappropriate results for goods that are purchased largely for investment reasons."
So rather than measure the price of homes being purchased, BLS uses rental equivalence to measure what is described as "the change in implicit rent, which is the amount a homeowner would pay to rent, or would earn from renting, his or her home in a competitive market." The BLS analysis revolves around how the participants in a BLS study answer the following question: "If someone were to rent your home today, how much do you think it would rent for monthly, unfurnished and without utilities?"
The process is very subjective. The question seeks to measure what one could rent one's house for, not what one could, would, or did pay for a house to live in. These are two very different things. First of all, most Americans would rather own their home than rent one — in essence making an owned home more valuable than a rented home. So, in asking someone what they would rent their home for, one would expect to get a value less than what they would be willing to pay to own the same home. Secondly, with low interest rates and more access to financing, home prices have been going up much faster than rental rates over the last few years. Are home buyers not paying higher prices because rental rates (and owners' equivalent rent) have been rising at a slower rate? Of course not! Housing price inflation is very real and it has been understated in the CPI through the use of owners' equivalent rent calculations.
The Office of Federal Housing Enterprise Oversight maintains an index called the House Price Index. The OFHEO uses data compiled from more than 28 million repeat housing transactions to estimate changes in the House Price Index. It covers all areas of the country, and is considered by many to be the most accurate and comprehensive gauge of housing price changes.
If we strip out the owners' equivalent rent data from the CPI and replace it with data from the House Price Index, we end up with an approximation of what the CPI would look like if it measured real home prices rather than owners' equivalent rent. With that substitution, the CPI rises much faster from Oct. 1, 1999, through Sept. 30, 2004. In fact, in the five years ended September 2004, substituting the House Price Index for owners' equivalent rent adds 1.43 percentage points to the average annual change in the CPI, taking it from a modest 2.52% to a somewhat alarming 3.95%. The one-year data are even more dramatic, where the use of the House Price Index increases the change in CPI to 5.82% from 3.41%.
By using owners' equivalent rent instead of real housing prices, the BLS is decreasing its relevance in reporting the inflation rate, resulting in overpricing of bonds. Therefore, I believe long-term bond yields are too low.
George Strickland is managing director of Thornburg Investment Management Inc., Santa Fe, N.M.