The watchword for today is "Oversold."
Bonds were oversold, and snapped back yesterday. The market got oversold and REALLY snapped back yesterday. Even the dollar, which has taken a 30% haircut over the past few years, is oversold and bouncing (more on the greenback later today).
The terms oversold and overbought get tossed about quite often. Unfortunately, many people (especially glib pundits and talking heads) use the phrase incorrectly.
What is Oversold? It starts when an asset class gets sold aggressively. Then it moves on to one (or more) of these three measures based on Valuation, Technicals, or Psychology. (The same holds true for "Overbought").
Following the selling, I define the phrase "Oversold" in any of these measures:
• Oversold on a Valuation basis is simply when an asset gets sold to the point below its intrinsic value. That can happen when downside momentum runs away.
• On a Technical basis, an asset is oversold when it hits the lower boundary of an oscillator. Think Bollinger Bands, or a Stochastics. Many technicians will consider an asset oversold when it is two standard deviations away from a recent price or moving average.
• The Psychology measure is trickiest, as it is the least quantitative. I think of an asset as oversold simply when too many people are on that side of a trade. Beyond consensus, a unanimity of view often leads to an asset being oversdold from a sentiment perspective.
Let’s look at the recent Bond selloff/yield rally as an example. Barron’s Randall Forsyth, in Bond Yields Were Too Low, obserrved that "Since May 8, when bonds started their trip south, the T-bond ETF has
lost a stunning 6.7% of its value. That would be equivalent to a
900-point drop in the Dow Industrials."
We clearly have the aggressive selling part covered. And the title implies there was a valuation issue prior to the selling. But what of the other measures? Consider the following:
"Only eight times since 1990 have bonds been this oversold, according to Birinyi Associates’ Jeffrey Rubin. Yet that doesn’t hold out much hope for a rebound, his data show.
During these episodes, Rubin found that T-bond futures prices were off an average of 1.72% a month after being this oversold. Three months later, they recovered but by a mere 0.15%. Nor did stocks stage much of a comeback in those instances: the S&P 500 was off an average of 0.16% a month later, and up just 0.14% after three months. Summer rally? Rubin’s numbers aren’t encouraging."
Eight instances over the course of 17 years satisfies the technical aspect of this — as of Tuesday, Bonds were more than two standard deviations off of their recent prices and longer term moving averages. And from a Psychological perspective, before yesterday, it was hard to find anyone suggesting that Bonds were a buy. That leaves only the valuation aspect in applicable as a measure of being oversold.
We will look at the Dollar later today to see if it is oversold and bouncing . . .
Incidentally, the rest of the Barron’s article is well worth a read, as it goes into the details behind the bond selloff.
It’s for Real: Bond Yields Were Too Low
UP AND DOWN WALL STREET DAILY
Barron’s Wednesday, June 13, 2007