Option Investor
Trader's Corner

Completing a Correction: Ways to Measure

Printer friendly version

This article is a natural closing article in a series I've written recently, mostly about the ins and outs of fibonacci retracements. Now that our big downside correction appears to have run it's course, for now, I'll take a final look at what retracements occurred at the recent top, and the last bottom. After that, I'll move on to one way of viewing the common chart formation for bull market corrections or interruptions, that of a Down-Up-Down pattern called an A-B-C pattern in Wave terms.

As discussed last time I wrote this column, its common to see 38, 50 and 62 percent retracements of a prior move, part of the so-called Fibonacci number series. WD Gann, a legendary stock speculator (mid 1900s), found it significant for retracements between a major low to high or, high to low, of 1/4, 1/2 and 3/4ths; i.e., 25, 50 and 75%. Gann also found significance in the 1/3 and 2/3rd retracement levels; I have learned to look closely for rally or downswing failures in the area of the 66% retracement point when a move has slipped past the fibonacci 62% retracement.

In so-called 'A-B-C' corrections, having a big decline, a short rebound and another long down segment. The two down 'legs' in a DOWN-side correction will of course be the longest. What's more, the two down-legs tend to be either EQUAL in length (a 'measured move') OR downswing 'C' is 1.25 to 1.62 times longer than 'A'. It's common in stocks to see a second downswing that is a fibonacci 1.62 times the first downswing.

SPX, as seen in the daily bar chart below, had a first decline of 129 points from peak (1556) to trough (1427), comprising price swing 'A'. The second decline from peak (1504) to trough (1370) carried 134 points, making price downswings 'A' and 'C' nearly equal. If you assume equal length price swings as a factor in covering puts that is not a bad call for the lead index in this market cycle. As long as you accept that a equal 'measured' move tends to be a 'minimum' for how far the second downswing will carry; and, the second leg down (C) might carry significantly lower still, as was seen in the Dow just recently.

The above is not to suggest that there will not be a lower low than already seen at 1370 at some point. But the close above the 21-day average is a bullish positive. I attempt to identify trading swings that are both substantial and predictable. How to get predictable? If you get a big downside correction, especially after a long rally, DON'T assume that the FIRST rebound is anything more than a pause in the two-part downswing comprising an A-B-C correction.

Rebound 'B' couldn't exceed more than 62% of the first decline, making it all the more likely that another good-sized decline of at least equal length was coming. What a trading useful thing to know from pattern recognition!

In terms of this thing I talk about all the time as 'pattern recognition', it's interesting to place the recent 1370 SPX low in context of its broader weekly chart. Suppose I was wondering (I was!) if the second leg down in the S&P was going to EQUAL the first decline OR was maybe was going to carry farther; e.g., another 50% further. However, if a low was ALSO coming down to support implied by the low end of a well-defined long-term weekly chart uptrend channel, this weekly chart factor would suggest the 'equal' moves outcome provided the trendline held.

I had this weekly chart pattern outlines drawn for some time; only the price fill is unpredictable. You can project straight lines as in my weekly chart markings below, but it doesn't mean that a decline will STOP right at support implied by its long-term up trendline. WRONG! The recent upside reversal came from a low, at 1370, right at technical support implied by this long-established trendline. I've seen it unfold this way in market cycles so much as for it to no longer seem kind of 'uncanny'.

The Dow, consisting of just 30 stocks and not being capitalization weighted to boot, can get more volatile in its price swings, particularly in a mini-panic environment. While the second part ('C') of its two-part decline was 1.25 times the first, the shape of the A-B-C is the same basically, is the exact fibonacci retracement of 62% of the first down leg before a sharp part-2 decline kicks in.

The A-B-C decline seen in the Nasdaq 100 (NDX) was very similar to INDU in that the second downswing of 189 points (C) was 1.27 times the 148 point decline (A) from 2060 to 1912. If you were 'keying' off the S&P 500 dominant index, NDX was a buy when it got down near 1800.

With the Nas 100, the retracement at the end of rebound 'B' was just 50%. The index couldn't get more traction than that. The 'natural' resistance levels are 38%, 50%, and 62-66%. That another downside correction was 'due' so to speak was being 'announced'.

Technical indicators are great when they can 'confirm' or not confirm what is being suggested by other market measures; e.g., the bottoming-reversal suggested by:
1.) the equal leg decline in SPX and
2.) the rebound from a major SPX up trendline.

My important 'sentiment' indicator can be yours too. Whenever total CBOE equities put volume equals or exceeds total daily call volume (a 1:1 ratio, or less) my rule of thumb is to look for a major trend turnaround to the upside within 1-5 trading days. [Conversely, when call volume is twice equities put volume (CBOE), that's often a precursor to a top.]

In the market activity there were two days registering such a put to call 'imbalance' relative what happens most of the time and is seen below. Within 2-days where my call-put ratio dipped into 'oversold-extreme bearishness' territory I also label as 'BULLISH'. Contrary as it is, a certain level of bullishness is associated with tops and a certain level of bearishness, with bottoms and these can be pretty precisely measured. Sure worked here, so far. Stay tuned!


Please send any technical and Index-related questions for my answer or feedback to Click here to email Leigh Stevens support@optioninvestor.com with my name ('Leigh Stevens') in the Subject line.

Trader's Corner Archives