In response to my last post (see Watching the USD for clues to equity market direction), an alert reader pointed that the SPX had formed a bearish island reversal.
Wikipedia explained the island reversal formation this way:
In stock trading and technical analysis, an island reversal is a candlestick pattern with compact trading activity within a range of prices, separated from the move preceding it. This separation is said to be caused by an exhaustion gap and the subsequent move in the opposite direction occurs as a result of a breakaway gap.
I had grown up with trading aphorisms and folklore like this, so I decided to test out whether the island reversal formation had any trading information. The results were surprising, and it was another lesson in how asymmetric signals were at tops and bottoms (see The ways your trading model could be leading you astray).
Asymmetry strikes again
My study used the daily open, high, low and close data of the SPX going back to January 1, 1990. I looked for instances of island reversals and calculated the returns of each episode. There were, on average, 1.5 reversals per year. As the table below shows, the results were surprising.
Bullish reversals, where the island reversed upwards, did not perform very well. The index saw a one day bounce and went on to underperform going out about a week. Bearish reversals performed a lot better than expected. Absolute returns were positive whatever time horizon you chose, though the index underperformed on a 2-3 day time horizon.
This gave me another lesson in the asymmetry of trading models. In the current instance, we are ending the three day underperformance window. If history is any guide, then stock prices should see better returns in the days ahead.
Disclosure: Long SPXL
The philosophy of this site can be summarized by a variation of an old adage:
Give a man a fish, he’ll eat for a day.
Teach a man how to fish…he’ll want to get a boat.
I am not here to just give my readers a fish for the day, I would rather help them build their own boat.