Aug

3

Zimbabwe inflationThis is a shout-out to all the futures exchanges: How about a contract on the Consumer Price Index? There is a wide dispersion of opinion on what CPI will be in the months and years to come. There are plenty of pundits predicting deflation and roughly an equal number predicting Zimbabwe-style hyperinflation. Many people feel a need to hedge against inflation. They would be natural buyers, and they'd probably lose money most of the time in the hopes of making a killing someday. As opposed to S&P futures, where just about all the trading is in the front month, for CPI futures the volume would likely be distributed over several years forward. There are already successful contracts on Fed Funds that have that feature (not to mention natural gas and some other commodities).

CPI is announced once a month, at 8:30AM. Futures contract settlement dates could be set on those announcement dates. A reasonable contract size would be $10000 for full CPI percentage point. Maybe the January 2013 contract would be trading at 4.5% now. If someone went long and then the January 2013 CPI measurement turned out to be 5.5%, then he'd make $10000 per contract. That's pretty simple. How about it? Maybe one of the more nimble exchanges, such as the CBOE Futures ( http://cfe.cboe.com/ ), will pick up on this.

Rocky Humbert responds:

As I've written previously, the current 10-Year TIP may not be a good measure of inflation over a short time horizon — because it's structured to capture the CPI over a ten-year holding period and is path dependent. Using a *constant* maturity 10-year TIP may exaggerate this effect. If you use a 2-year or 5-year TIP, you may find that it produces more accurate inflation forecasts for short-term horizons.

There's tons of academic literature on predicting CPI — if you Google-Scholar the subject, you can spend weeks reading the papers and building your models. Monthly changes in CPI have been nominally small over the past decade (although big in percentage terms), and it should be fairly easy to predict the CPI out a month or two — but becomes incrementally more difficult for each subsequent month. I've dubious of the investment value of knowing next month's CPI — but I would love to predict with confidence whether the CPI is going to be either consistently negative or consistently over 3.5% next year this time. That could be extremely useful — whereas a CPI between those two bounds shouldn't matter much.

As of this writing, the 5-year tips "breakeven" CPI is 1.38% and the 10-year tips "breakeven" CPI is 1.77%.


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search