As far as market timing signals go, it’s time to put the transportation sector out to pasture.
In recent years, the original intent of Dow Theory — which uses new highs and new lows in the Dow Jones Industrial Average (DJIA) and the Dow Jones Transportation Average (DJTA) to signal bull or bear trends — has been altered so that short-term divergences between the two averages are being used to predict market direction. Unfortunately for practitioners of this version of Dow Theory, the relative performance of the transports is actually a poor indicator.
The most recent example occurred late summer of last year, when transportation stocks underperformed the rest of the market by a wide margin in the third quarter. During that time, the iShares Dow Jones Transportation Average Index Fund (NYSEARCA:IYT) shed over 6% even as the SPDR Dow Jones Industrial Average ETF Trust (NYSEARCA:DIA) gained more than 4%.
This huge shortfall received quite a bit of attention at the time, including this blog in The Wall Street Journal that contained the following quote from Peter Boockvar, managing director at Miller Tabak: “This is a major divergence that should not be ignored. It tells me the risks of being in the market at these levels are growing.”
Boockvar certainly wasn’t alone in his assessment. However, the subsequent results show the limitations of this performance relationship as a timing signal. While the market indeed lost ground early in the fourth quarter due to the uncertainty surrounding future tax policy, stocks have bounced right back and are up over 2% since the end of the third quarter.
This isn’t the first time the relative performance of the two indices has provided a false signal regarding broader market direction. Since late 2009, there have in fact been seven occasions of severe under-performance for the transports … and the subsequent performance of the DJIA has been very positive on balance, based on data from Yahoo Finance:
|Interval||IYT||DIA||DIA, + 3 mos.||DIA, + 6 mos.|
|1/31/12 – 2/22/12||-3.6%||2.4%||-3.3%||1.8%|
|7/1/11 – 9/30/11||-24.9%||-13.3%||11.9%||19.9%|
|1/13/11 – 2/23/11||-4.5%||3.1%||2.2%||-7.8%|
|4/30/10 – 7/6/10||-16.4%||-11.5%||12.6%||19.8%|
|12/16/09 – 2/8/10||-9.1%||-5.2%||4.6%||7.5%|
|9/15/09 – 10/30/09||-11%||0.2%||3.6%||15.1%|
The Dow-Transports relationship, while not an effective predictor of the market itself, does serve as a signal regarding the relative performance of the transports themselves. A look back over the past 10 years shows a nearly perfect correlation of 0.91 between IYT and DIA.*
As a result, large divergences tend to correct themselves in time. During the mid-October to mid-November sell-off, for instance, IYT lost only 4.4% even as DIA fell 7.2% — making up for some of the under-performance that had occurred in the third quarter. In this sense, large performance gaps can signal that a mean reversion is in the offing.
In fact, this may be happening right now. A look at the chart below shows that the out-performance of transportation shares has now reached extreme levels, indicating that the group may be on the verge of lagging once again:
As the chart indicates, the transports are on a tear: Since Nov. 15, IYT has tacked on over 15% behind the strong showing of airline stocks such as Delta Airlines (NYSE:DAL, +43%), United Continental Holdings (NYSE:UAL, 26%), and Southwest Airlines (LUV, +28%) — well above the 8.1% gain registered by the Dow.
Predictably, this development is bringing with it plenty of bullish comments. But out-performance for transports can be just as misleading as under-performance. Most notably, IYT gained 18% in the first five months of 2008, trouncing the -4% return for DIA. Later that year, of course, the market crashed … and IYT lost nearly 60% of its value peak-to-trough.
The moral of the story? Transports behave much as any other sector does, with a cycle of divergences and mean reversions over time. Use this as a trading opportunity, but not as an indicator of future market performance.
As of this writing, Daniel Putnam did not own a position in any of the aforementioned securities.
*Correlation runs on a scale from -1.0 to 1.0, with -1.0 being perfectly negative correlation and 1.0 signaling perfectly positive correlation. The higher the number, the greater the correlation between the two securities.