Market-Timing in Another FormLike their domestic brethren, international managers are slaves to indexation. And like domestic markets, there is no shortage of index benchmarks from which to choose. Here we will focus on the Morgan Stanley Capital International EAFE (Europe, Australasia, Far East) and Emerging Markets Free indices. They underlie two ETFs that trade during U.S. hours: the EFA and EEM, respectively. The EFA was launched in August 2001, and the EEM in April 2003. These ETFs reflect the action in markets whose trading hours on any given day precede, follow and overlap American trading hours.
|Related, but Not Closely Correlated |
Beyond CorrelationNearly a decade ago, I set out on a massive data-mining study designed to extend Japanese candlestick analysis. The final products of this study were incorporated into daily quantitative futures-market newsletters published in a previous life. As useful as candlesticks are, they are not a mutually exclusive and collectively exhaustive classification system capable of providing a unique label for all days. The key to developing such a classifier is normalization of the key identifiers of a day's structure -- its open, high, low, close and midpoint -- by locating them on a stochastic distribution of the day's range, a variation of the Steidlmayer distribution familiar to aficionados of Market Profile analysis. As no specific notation for this concept exists, a standard relational notation will be used. For example, if the range between a day's open and close -- corresponding to the body on a candlestick -- exceeds the mode of the day's stochastic distribution, it will be designated as "O>=C." In the table below, the first classification would be for a day where the open/close range, the open/midpoint range and the midpoint/close range all exceeded the stochastic criterion.
|Source: Howard Simons|
A Time of Brief HistoryWhile this classification scheme can produce detailed forecasts of various kinds given enough data, the short histories of the EFA, and especially of the EEM, preclude a complete and statistically significant data-mining approach. Instead, we can ask some simple questions for starters. Others will follow at a later date.
- Does a white candle (close greater than open) with a positive interday change in the SPDR lead to higher values the next day in the EFA and EEM?
- Conversely, does a black candle (open greater than close) with a negative interday change in the SPDR lead to lower values the next day in the EFA and EEM?
- Do white candles in the EFA and EEM lead to higher values the next day in the SPDR?
- Finally, do black candles in the EFA and EEM lead to lower values the next day in the SPDR?
|Results of the Four Questions |
|SPDR White Candle + Positive Change|
|Question 1||Total Conditions Met||Next Day Up||Correct|
|SPDR Black Candle + Negative Change|
|Question 2||Total Conditions Met||Next Day Down||Correct|
|ETF White Candles + Positive SPDR Change|
|Question 3||Total Conditions Met||Next Day Up||Correct|
|ETF Black Candles + Negative SPDR Change|
|Question 4||Total Conditions Met||Next Day Down||Correct|
|Source: Howard Simons|