I’ve been having a discussion with Charles about using candlestick analysis on the chart of EWJ, which is an exchange traded fund (ETF) that tracks the MSCI Japan Index. Charles pointed out the large number of open windows (gaps) on the chart which should serve as support/resistance. I brought up the question of whether those gaps were legit because of the fact that in many cases the underlying index, MSCI, actually has long real bodied candles where the gaps appear on EWJ. But those gaps may still be support/resistance because the long candles on MSCI should also be support/resistance. So in that case candlestick analysis may still work.

I think things are less clear when you start trying to use candlestick reversal signals on EWJ, or any ADR of an equity which has its main trading activity overseas while the U.S. markets are closed. Some of the most important reversal candles, like hammers and shooting stars, are made when you have an intraday reversal. Those intraday moves will be completely missing on the ADR, or in this case on EWJ. I tried to get a candlestick chart of MSCI to compare to EWJ but all I could get was a line chart. So instead I pulled up a chart of the Nikkei index. I know this isn’t an exact comparison, but EWJ does generally track the Nikkei. I’ve pointed out some of the differences on the charts below:

Even when you do get a reversal candle on EWJ it’s usually useless as an entry/exit signal because they’re often followed by a huge gap. I see the same thing on charts of many ADRs and whenever one of them pops up in one of my scans I just skip them because of all the gaps.

There’s another situation in which I think you’d be wise to question using candlestick reversal analysis. The charts of the Dow, S&P 500 and other indices which contain NYSE stocks often have a ‘fake’ opening value. I say it’s fake because NYSE stocks don’t all open at 9:30. It may take an extra five minutes for some NYSE stocks to open. So the index may be close to unchanged and then move once the stocks start opening. In the case of a morning when the futures are strong up or down and many stocks are gapping accordingly the Dow chart won’t show a gap at all. So even if all 30 Dow stock have big gaps they won’t be reflected in the index because of the NYSE stocks’ stuttered openings. DIA, the Dow tracking stock, would be a better choice to use for doing candlestick analysis on the Dow. (And I’d use SPY for the S&P 500.) Here are charts of the Dow and DIA:

(that last balloon says ‘not a doji’)

Those differences are much more subtle than the EWJ case, but there are certain days when it makes a big difference. The days after 9/11 are a good example of that.

I won’t go so far as to say that candlestick analysis doesn’t work in the cases above, but I do think that you get better information using the data that gives you all of the intraday activity. But I guess whatever you do just be consistent.