Up in Arms

By David Bukey
After dropping sharply in May, the S&P 500 index (SPX) also stumbled in early June, falling 10.2 percent from April 30 to June 4. At that point, the Arms index (TRIN), which moves inversely to stocks, jumped to 13.22 as selling pressure surged. It was the indicator’s fourth largest reading in 70 years — a potential (contrarian) buy signal, according to its developer Richard J. Arms. Over the next two weeks, the S&P 500 slid 2.1 percent at its lowest point, but it eventually bounced 5.3 percent higher (Figure 1).



 The TRIN measures the market’s momentum with a ratio of ratios, and its extreme readings — such as the June 4 spike  —  can also identify possible oversold or overbought levels. It divides the ratio of the number of advancing/declining stocks to the ratio advancing/declining volume (see “TRIN background” for more information).

A TRIN value of one suggests the market is neutral, as the number of rising and declining stocks is balanced with advancing/declining volume. The TRIN drops when advancing stocks outnumber declining ones and advancing volume climbs relative to declining volume; it rises when declining stocks outnumber advancing stocks and down volume is dominant.

In 2006, a Market Pulse article found the TRIN had been a fairly reliable countertrend indicator over the previous decade. After the TRIN rose to a 60-day high, the S&P 500 ETF (SPY) gained an average of 1.81 percent within 10 days. After the TRIN fell to a 60-day low, SPY declined 0.59 percent on average within 10 days.

A great deal has happened since 2006. Stocks cratered and volatility skyrocketed during the 2008 financial crisis. Stocks rebounded strongly in March 2009 before stalling approximately a year later. Has the TRIN continued to spot the stock market’s oversold and overbought levels since 2006 or has it generated false signals?

The following analysis compares SPY’s price action two weeks after TRIN extremes in two time periods — January 1996 to February 2006 (original study) vs. February 2006 to June 2010.

TRIN highs and lows
Daily TRIN values can be quite volatile, so analysts typically smooth the raw data with a moving average. This study uses a 10-day exponential moving average (EMA) and defines relative highs and lows as follows:

The 10-day TRIN EMA must make a 60-day high.
The 10-day TRIN EMA must make a 60-day low.

Figure 1 shows a daily chart of SPY with the 10-day TRIN EMA in 2010. Even when smoothed with a moving average, the TRIN was unusually volatile in June. After jumping in early June to its highest point in more than a decade, the TRIN EMA dropped down to normal levels within days. (Note: The raw daily TRIN value slid on June 10 to its lowest level since 1984.)

Figure 1’s TRIN EMA highs and lows didn’t point to clear oversold or overbought levels in 2010. For example, the TRIN EMA made two 60-day lows in the middle of an uptrend. And although the market rose briefly following the TRIN EMA’s 60-day highs in late April and late May, the market soon continued lower. SPY’s rally following the TRIN’s June 4 spike seems like the exception, not the rule.

At that point, it was tough to disagree with Jason Goepfert of SentimenTrader.com, who claimed “market breadth” indicators such as the TRIN, which measure internal data (as opposed to price), had become so volatile “that it’s hard to rely on extremes to carry the importance they once did.” According to Goepfert, the TRIN is especially vulnerable to distortions because it is a ratio of a ratio, unlike other breadth indicators such as the advance-decline line or put-call ratio.

Goepfert also believes the TRIN has been distorted, in part, by the massive volume in low-priced stocks such as Citigroup (C), which had an average daily volume of nearly 1 billion shares in May. It represented more than one-quarter of stock volume on the New York Stock Exchange (NYSE) that month

For the complete article, see the September 2010 issue of Active Trader magazine. Click here to subscribe.



|

|