![]() | ||
|
| ||

|
Hello Fellow Chart Watchers! Hopefully you were on the sidelines for last week's continuation of the big slide. As John Murphy indicates in this week's column (below), "Cash is King" and most Chart Watchers (who aren't short sellers) should have exited stocks weeks ago. Technical sell signals have been in place for months. Based on our Sector Rotation Model, we are continuing to watch for confirmed signs of life in the Cyclical/Transportation sector before we even think about opening a long position in this market. While we're waiting, we have been studying up on some longer term charts and looking for the signals that were given in the past when a stock came out of a prolonged slump. SharpCharts' new "User-Defined Dates" feature makes this a snap! This process reminded me of an article I've been meaning to write for a while now titled: With lots of Chart Watchers studying longer term charts now, I thought that this week would be a good time to review the benefits of Percentage Scaling when it comes to looking at the past and comparing the past to the present. The biggest problem one has when comparing past and present charts is that as prices climb (or fall), the relative impact of a similar change in price changes (often dramatically). This is just common sense: a move from $2.00 to $3.00 is much more significant (+50%) than a move from $102.00 to $103.00 (+0.98%). The same effect appears on longer duration price charts when the standard "Linear" vertical scale is used. High prices near the right edge of the chart cause older prices (which are typically much lower), to be compressed so much that price patterns and support/resistance lines cannot be seen. ![]() The dominant feature in this chart of IBM is the current volatility on the right-hand side. In contrast, the left side of the chart appears much more sedate. The trendline that connects the 1993 low with the dip in 1995 is still well below the current price level and in no danger of being tested anytime soon, right? Wrong! The huge irony here is that, on a percentage basis, those older price movements are often just as dramatic as the recent moves. Fortunately, there is a different way to scale the vertical axis so that large percentage moves always appear large regardless of the price level they occured at. Of course, I'm talking about using "Logarithmic" scaling. The difference is dramatic: ![]() Now we see that the price movements back in 1991 and 1992 were much more painful than those currently underway. Remember that "safe" 1993-1995 trendline? It is actually undergoing a critical test as we speak! By scaling prices so that price changes reflect the impact they had on a percentage basis, the Logarithmic scale provides a more accurate long-term picture. "Yeah, yeah - I know all about using log scale on long-term charts. What does it have to do with the MACD indicator?" OK, I'm getting to that. When you are looking at charts with large price movements on them, you need to make sure that you use indicators that also incorporate Percentage Scaling. Just as linear price plots can be misleading on long-duration charts, linearly scaled indicators can also hinder accurate analysis. Unfortunately, one of the most popular indicators, the MACD, is in that category. Look at the MACD line in the chart below: ![]() The red circle shows that the MACD recently set an all-time low for IBM, implying that the stock is was in worse shape at that time then it had been in the past. On the left side of the chart, the MACD was pretty tame during the early 1990's. In places, it is so flat we can barely see any change at all. Sound familiar? ;-) Just as the Log scale brings out the percentage action in price plots, the Percentage Price Oscillator (aka "the PPO") brings out the percentage action in the MACD line. Look at the PPO line in the chart above. It shows that IBM's stock was much weaker in 1992 (green circle) than it is right now. We also see many more ups and down in the PPO line than the MACD line on the left half of the chart. The PPO is calculated exactly like the MACD except it includes an extra step that scales its readings on a percentage basis. In addition to the advantage illustrated above, that extra step also allows you to directly compare the PPO reading from one stock with the PPO reading from another stock - something else that the MACD cannot provide. Check out the "PPO" setting on our Market Carpets to see that feature in action. Just be sure that the carpet is in "Absolute" mode to compare actual PPO values (see the instructions page for details). Fortunately, most other indicators do not have the same problem as the MACD in this respect. Any "bounded" oscillator such as RSI, ADX, and CCI are percentage-scaled by definition. Of the indicators that we support on StockCharts, only MACD and Average True Range (ATR) are not percentage scaled in one way or another. So as we sit on the sidelines waiting for the market to come to us, make sure you are studying those historical charts the right way - with percentage scaled price plots and percentage scaled indicators.
NEED FOR PERSPECTIVE. . . There are times when we have to stand back to get a better perspective on major trend direction. This is one of those times. Chart 1 shows this week's breakdown in the Dow under chart support near 10,300. [That's now overhead resistance]. We think the Dow will have to test the lows of the past year around 9700. The Dow will also be testing the rising trendline to the bottom right. ![]() CASH IS KING. . . In every economic downturn, all stocks eventually fall - even defensive ones. We've been seeing that this week. At such times, cash is king. That doesn't mean everyone should go out and sell all their stocks. But it argues for higher cash positions at such times. Bond funds also continue to outperform stock funds. If you like John Murphy's commentary, visit his website to learn how to receive it on a daily basis. Here are some links that should help you get started:
Questions? Comments? Concerns? Problems? Suggestions? Simply 'r'eply to this email message and I'll see what I can do. Take care, ABOUT YOUR SUBSCRIPTION: You are receiving this mailing because your email address was entered into the subscription form on our web site. To unsubscribe, or to switch between our HTML version and the Plain Text version, use the form on our Subscriber page. If you have any problems, simply 'r'eply to this email and I'll try to fix things by hand. |