Inflation data deeply flawed: economist

By Steven Lamb | December 12, 2006 | Last updated on December 12, 2006
4 min read

In financial planning, there are many assumptions advisors must make simply because it is impossible to predict the future with any kind of precision, and in general it pays to err on the side of gloom.

Portfolio returns of 8% may be used as a conservative estimate by some, for example, even if they actually think the portfolio can earn an average of 11%. On the other side of the ledger, 3% inflation is a common assumption, given the central bankers’ expressed target of about 2% — and for the most part, the consumer price index has indicated they are on track.

But what if the CPI data is wrong? What if the real inflation rate is not 2% but 4%, 5%, or even higher? According to one economic researcher from the U.S., it is higher — much higher — at least south of the border.

It’s not so much that the CPI data is wrong, but its usefulness in planning is limited, says David Ranson, president and head of research at H.C. Wainwright & Co. Economics. Although his research is based on U.S. data, Ranson points out that most developed nations use roughly comparable models when calculating CPI.

“An increasing cost of living is certainly a symptom of inflation,” he says. “The problem with an increasing cost of living is that for each one of us, it is different. Each city has a different inflation rate; therefore, each community, each family, each age group. That’s not a very good way to define inflation if it’s different for every single person.”

The same problem applies to using labour costs as a measure of inflation — the rate of inflation among bank presidents’ wages tends to be somewhat higher than among those of tellers. Like the CPI, labour costs are a lagging indicator of inflation.

Similar problems arise with many other traditional definitions of inflation: rising interest rates; rapid growth of the money supply; subdued equity market performance — all tend to be symptoms of inflation, rather than defining characteristics.

And an accurate definition of inflation is key to identifying periods of inflation and protecting a portfolio against it.

“We’re getting a little bit closer to a workable definition when we go into other assets, such as real estate and commodities,” Ranson says. “High returns on hedge assets tend to occur when inflation is present. This is actually a reasonable definition as to whether inflation is or is not present.”

To test this definition, Ranson looks to the market, finding a higher correlation between hard assets and the type of market behaviour that indicates inflation.

“If we look at real estate and commodities, we have market prices, which are an unimpeachable source of information, and they do behave systematically differently when other symptoms of inflation are present, and the opposite way when those symptoms of inflation are absent,” he says.

To further test this, Ranson re-examined the CPI model. He argues that CPI data is flawed because of the heavy weighting it gives to “historical” data, the most egregious being the 23% allocation to “owners equivalent rent” on primary residences.

While this measure accurately reflects the average cost being paid for housing in the economy, it does not reflect month-to-month changes in inflation. For example, the cost of carrying a 30-year mortgage taken out 25 years ago should remain relatively static, aside from any change in interest rates at the time of renewal.

Assuming that the homeowners conservatively locked in their rate for five years, there would be no change in this measure between 2001 and 2006. Clearly, according to this measure, the housing market has remained steady.

Ranson suggests a more accurate reflection of inflation in the housing market would be to capture the month-to-month changes in the cost of buying the same home. This would take into account not only the changes in house prices but also the shifts in mortgage rates.

Taking these changes into account and substituting them for the lagging data in the CPI, Ranson recalculated the inflation rate, finding it to be not in the 2% to 3% range so often quoted but 8.4%.

The bond market is one of the best indicators of inflation, he says, although it tends to lag actual changes. Using the bond market as “jury,” Ranson tested various commodities as indicators of inflation.

It should be mentioned that Ranson’s seminar was sponsored by Bullion Marketing Service, which offers the Millennium BullionFund, a mutual fund that exclusively holds physical platinum, gold and silver.

There are numerous other products now available designed to provide exposure to the precious metals market, ranging from ETFs specializing in a single metal, to shares in gold mining stocks and shares in commodity pools.

Ranson finds the three primary precious metals to be the best predictor of the bond market’s reaction to inflation, with all three having a correlation coefficient above 0.7, and a lead time of at least 12 months over the reaction of the bond market. In other words, if precious metals are heading higher, Ranson’s data suggests bonds will head lower in about a year’s time.

To negate the effects of inflation altogether, a fixed income portfolio would need a precious metals allocation of 18%. For equities, this allocation rises to a whopping 40%. Using a 40–60 split, that would suggest an overall allocation of 31%.

He admits, however, that such absolute inflation hedging is a ridiculously high bar to set and that such a large allocation would not be suitable for many investors. While the portfolio would be “bulletproof” against inflation, it would be subject to the sometimes violent fluctuations of the precious metals market.

Filed by Steven Lamb, Advisor.ca, steven.lamb@advisor.rogers.com

(12/12/06)

Steven Lamb