In my most recent (5/27) Trader's Corner article, part of an ongoing basic series in technical analysis, I finished on chart gaps. Today, I'm moving on to another technical tool useful in spotting reversal points. This all, after a technical update.


I've been writing since late-May about my expectation that ..."it seemed likely that a double bottom would set up." Then, more specifically in my most recent Index Wrap that the type bottom would not be a 'V' shaped one, but more likely a 'W' formation. My chart highlights on the first two charts show the 'W' pattern on both the S&P 500 (SPX) and the Nasdaq Composite (COMP). SPX has also achieved an initial upside breakout above its down trendline. The SPX stocks led the market today, leapfrogging the tech-heavy COMP, but that's to be expected with the strong advance in the energy stocks today and they're all to mostly in the S&P.

What was encouraging for the bulls today, at least from my particular technical/psychological market perspective, was that my bullish sentiment model (CPRATIO at bottom of SPX chart), stayed more or less 'neutral' as to bullish or bearish extremes in traders' outlook. This is what tends to happen after extremes in bullishness lead to a correction leads to renewed caution. Traders were learning (?), had learned (?) that this market does have downside RISK.

The major resistance versus 'breakout point' remains the 1100-1107 area, the high end of the recent trading range and also where the 200-day moving average (2007) comes in as an important sign of regained (or still faltering) momentum for the S&P.

COMP has a bit further to go to pierce its current down trendline (Friday intersection = 2257) but the double bottom is a potent sign that sellers and buyers were finally in balance in the 2140 area which is what makes a bottom.

Like SPX, it will be an important indication, especially to fund managers, when/if COMP can get back above its 200-day moving average; and here, COMP has less far to go to regain its 200-day average. Trading above the 200-day average will be construed bullishly by many fund managers. They've been doing some trimming in the last few weeks, but have some cash for purchases as happened today for the recently out-of-favor big energy stocks.

When we look at the extent of this recent decline on the daily charts as seen above, it looks a bit like the sky has fallen, with the damage to our portfolios more severe than it is IF compared to where this now multiyear rally began in March of last year (2009) relative to how high the major indexes got to at the recent March (2010) high approximately a year later.

However, if viewed on a weekly chart basis, with level lines that represent the per cent retracements of the March '09 to March '10 advance. If prices drop back in the S&P 500 stock index (SPX), to the 33% line, it means that at the recent lows, SPX had only fallen one third of the amount that it had gained. What tends to freak many market participants (not the 'pros') is that the sell offs or market breaks cover that 1/3 (or 1/2, etc.) so FAST! That's just the nature of bullish versus bearish price swings. They go down a lot faster than they go up.

Generally speaking, retracements of just 1/3rd of a prior move is a (relatively) minor retracement as these things go. An 'average' retracement for a stock, after a rally or after a decline, is 50% of the prior move.

What makes me think, as seen above with the 'retracement tool' seen above, that the 33%, or 38% retracement level might tend to offer 'support' or where buying interest might resurface?

This brings me to the subject de jure: Retracements and why the heck would I care about whether the recent low that I just BOUGHT was this or that percentage of something else?!


We owe some debt to Charles Dow for his observations that an intermediate trend often will retrace (give back) anywhere from around 1/3 to 2/3rds of the distance covered by the primary trend, before the primary/major trend resumes. There were further refinements on retracements made by W.D. Gann, a famous stock and commodities speculator of the first half of the 1900's. Primarily, Gann found it significant to use charts that had retracements noted between a major low to high or high to low of 1/4, 1/2 and 3/4ths. Sometimes Gann used thirds as retracement levels to watch for clues to a support/resistance; i.e., 33% and 66%.

The origins of one of the most useful retracement theories for stocks and other markets came from someone who lived in the middle ages and was studying the population growth of rabbits. Leonardo Fibonacci was an Italian mathematician who was doing such work in the early 1200s.

The number sequence that is named after Mr. Fibonacci is where each successive number is the sum of the two previous numbers; i.e., 1, 2, 3, 5, 8, 13, 21, 34, 55, 144, etc. Any given number is 1.618 times the preceding number (approximately) and .618 times the following number.

There are technical indicators whose formulas rely on the Fibonacci number sequence, but the main application is to use the so-called 'fibonacci retracements' of .382 or 38%, .50 or 50% and .618 or 62%. The number 5 is in the Fibonacci sequence, and the others are ratios; .618 comes from the percent that each number is of the next higher number and .382 is the inverse of .618 (100 – 61.8 = 38.2). We'll stick to the shorthand and round off to 38 and 62 per cent.

What I find most useful in trading is to track what would constitute the 38, 50 and 62% retracement points, with the sometimes addition of checking where the 25 and 33% levels are on an initial decline. Sometimes 66% is key retracement that marks a trend reversal point.

There is a simple pragmatic reason for the popularity of keeping trace of retracement levels, is that buying or selling in these (retracement) areas often results in coming close to buying at the low and selling at the top. Maybe the saying of buy low, sell high owes something to the common retracements.

Most charting applications will calculate by the various level lines whatever retracements you tell it to, ranging from 25%, 33%, 38% (.38), 50%, 62% (.62) or 66% or even 75% by first pointing at the low, then the high (pullback retracements) or first at the high, then the low (for retracement rallies in a downtrend).

In a countertrend rally within a dominant downtrend, once a minor downside correction begins, I especially look to see if a recent high represented 38, 50, or 62% retracement of that high relative to the rally starting point. If the (up) swing high also failed at or under the key down trendline and was accompanied by an overbought condition, it often has been a tip off to short or to buy puts.


If the correction “resists” falling to 38% after a strong up move, I will add the 25% retracement line. If a deep correction falls below the 62% retracement, add the 75% retracement line.

If the correction of an uptrend falls BELOW a 3/4 or 75% retracement, I assume that the correction will be 100% or what I call a “round trip” back to the origin or starting point for the trend. Don’t forget to look at a “close-only” or line chart also – sometimes the retracement comes in at one of the 5 levels, but its on a closing basis as seen in the chart below.

All you need is some degree of assurance or assumption that a price swing has run its course and that a counter-trend/opposite move is developing. In a downtrend, the 38, 50 and 62 percentage figures are added to the most recent low if it looks like that was a bottom. Once it's apparent that a minor counter-trend rally is underway (prices and volume surge) the expectation is that in a 'normal' market, prices will rebound an amount that is equal to about half of the last decline, or a 'little bit more' (62%) or 'a a bit less' (38%). If prices climbed to the 62% retracement level, still within what you think is still an overall downtrend, this would suggest an area for a favorable exit if long, and favorable trade entry to short/buy puts. You can then place a stop just beyond the retracement level you got in at.


These guidelines were derived from the frequency at which market reversals occurred at or near the fibonacci 38, 50 and 62% retracement levels.

1.) A strong trend will usually see only a 'minimum' price retracements; e.g., around 33 to 38% (sometimes only 25%). If prices start to hold around this 'minimal' retracement area, trade entry could be indicated as there may NOT be a deeper correction.

2.) A 'normal' trend, not powered by something extraordinary, will often see a retracement develop of about half or 50% of the prior move. The most common level to buy or sell, especially in stocks, is this retracement amount, with an exit if it continues on much beyond 50%; e.g., 5% more.

3.) Within the range of 'normal', but not evidence of a particularly strong trend, will be a retracement of 62% or perhaps 2/3rds (66%); sometimes, 75%, especially in more volatile stocks. If prices hold this area, it's also a good target for initiating a buy or sell with an exit if the retracement exceeds 66%.

4.) If a retracement exceeds one level, look for it to go to the next; i.e., if a retracement goes beyond 38%, look for it to go on and approach 50%. If it exceeds 50%, look for 62%. If a retracement exceeds 62%, or a maximum of 66% (or 75%), then I look for a 'round trip' or a return all the way back to the prior low or high, an 'ultimate retracement' so to speak of 100%.

5.) Retracements are most commonly done from the lowest traded low to high and highest intraday high to low. We don't see retracments used so much based on highest Close to the lowest Close, etc. However I sometimes experiment with retracements based on closing levels as they also are worth exploring.

6.) The common retracement levels, ahead of reversals short-term or long, 'work' for all time frames or chart types; e.g., hourly (or less), daily, weekly and monthly charts.


In a strong trend, either up or down, the 'typical' retracement will be shallow. Don't wait too long in those situations!