Posts

Showing posts from October, 2010

The TIPS measure of CPI expectations

Image
Some government bonds (called TIPS) compensate the lender for changes in the CPI. For instance, if you lend $100 to the government, and the CPI goes up by 3% when it is time to repay, you will be paid $103 (a falling CPI has the reverse effects, within certain ranges). If a TIPS bond has a yield of 2% and inflation is expected to be 1% a year, we would expect people to want at least 3% (2% + 1%) yield on a regular (non-TIPS) government bond. In other words, the difference between the yield on TIPS and the yield on regular government bonds is the "implied" rate of inflation (i.e., the rate assumed by a decision-maker who considers both types of bonds and is open to buying either). Of course, market actors are simply making their best guesses; in retrospect, the assumption may prove to be wrong. This chart shows the yields for (5-year constant maturity) TIPS (red ) and Treasuries (blue) . The next chart shows the spread between the two, which is the "implicit" annua

Extrapolating the Unemployment rate

Image
This post is a bit of "extrapolation fun". What is the unemployment rate going to look like when the U.S. faces the next Presidential election in 2012? First, we have to answer another question. is this (largest and deepest post-war) recession qualitatively similar to other post 1940 recessions?... or, does this recession (and the Great Depression) have certain special characteristics that make them different? Ordinary recession, but huge scale: If it is just a question of scale, "dumb" extrapolation should give us a pretty good guess. Here's a chart I posted earlier , taken from from the excellent Calculated Risk blog , with my scribbles added. The blue rectangles to the right are the future. The red line shows the current recession. The time it took to (apparently) bottom has only been matched by the recent "dot.com"recession (the dark brown line ). It has taken about two years for unemployment to climb (downward in the chart) by 6%: i.e., from

US Unemployment Rate

Image
( Click here for the Slide-show version ) Here are two sample headlines: " Half a million more people employed " " Half a million more people unemployed " Can both are true? Yes, they can. Read on to understand how. To compute an unemployment rate, we ask: how many people want to work (how big is the "labor force"?) how many of them cannot find work (how many, of the labor force, are unemployed) This graph shows the total population of the US, broken down into four groups: first (between the top-most blue line and the red line), we have children (under 16 years) and people in the military or in prison, etc. next (between the red and green lines), we have civilian adults (16+) who do not want to work today. They could be going to school full-time, or retired, or stay at home parents, etc. finally, we're down to the "civilian workforce" (the green line) or civilians (16+) who want a job and are either employed or unemployed. Of these ,