Note - from June 24th 2009, this blog has migrated from Blogger to a self-hosted version. Click here to go straight there.
Just over a week ago, I gave an example of how a decent analysis ought to be done, presenting what I refer to as a Syncretic Analysis of the Crude Oil market. As I pointed out in that piece, analysis ought to integrate the three major 'modalities' of investment analysis - fundamental, technical, and sentiment. It could be argued that sentiment and technical are subsets of some broader mode of analysis, or that sentiment is a subset of technical analysis. I prefer to treat them as separate items.
You'll note that I seldom present some of the popular-but-esoteric forms of technical analysis - Elliott Wave theory and Fibonacci retracement levels - in my technical material. That's not because I don't think they're valid or useful (I use them both frequently), but because most of the time the bulk of the 'story' can be explained without requiring them. If there's a particularly interesting development that can be best shown using Elliott Wave Theory (like my 'enthusiastic amateur' wave count for Gold back on October 10th last year) for example, I will present it. As you can see from the Gold example, it's unlikely to win prizes for "Best in Show" in the prettiness stakes - but the information content is what matters. (Next time I will make sure to include Y-axis markers, too... ).
As far as 'Liberace Ratio' (also called the "Italian Rabbit-Fanciers Ratio") analysis is concerned, the principle problem in presentation is that it crowds everything - given the abominable job I did labelling the Gold chart on October 10th, imagine the visual damage I could wreak with a chart with a bunch of Fibonacci retracements (and projections) on it! I always compare my calculated target levels with 'key multiple' Fib projections: once I get the Commodities RantBoxes finished you will see what I mean.
All this is by way of preamble to today's 'semi-presentation' of a Syncretic Analysis of the major US markets. Today all I'm going to present is the Technical case. The Fundamentals will be in place later in the week, and the Sentiment stuff will wait until the weekend (Sentiment for an index requires calculation of implied option volatilities and put-call ratios, as well as futures Commitment of Traders analysis - a decent-sized job).
I'm going to concentrate on the Dow Industrials - the Nasdaq and the S&P move in very similar paths, and at present most of what goes for the Dow, goes also for the others (the Nasdaq indices are slightly more technically overcooked than the Dow; the S&P much more so.. and the Russell 2000 is off the charts).
Dow Jones Industrial Average - Technical Analysis
As with the Technical aspect of the Oil analysis, start with a weekly chart.
Note that I've highlighted on area in pink - where it appears that the %R/CCI combination got smacked consistently for a period of almost 8 months. This surge coincided with the flagrant upward push given to the markets after the invasion of Iraq in March 2003.
The scaling makes the chart a little deceptive - although they seem almost irrelevant, the pullbacks that were indicated by the %R/CCI analysis were often of the magnitude of 400-500 points in timeframes of under a fortnight - well worth trying to snaffle. Monthly charts are important for analysis of stock indices (usually I don't present them in the blog though - but I will this once). The Dow chart did not hit monthly overbought with a CCI divergence until December 2004 (see below)
As I'll show later in the week, the Sentiment analysis showed other signs that gave a heads-up that long-term shorts were not prudent during that time.
Anyhow - concentrate again on the Weekly chart above. Note the blue ellipse around the %R line - that's just to point out that the weekly chart is not yet overbought. The rally likely has some room to rise yet.
Now, the daily chart...
note the blue ellipse that gave a perfect long signal; oversold %R, CCI
divergence. likewise, the oversold %R, CCI divergence that developed
over late June - early July as indicated by the green lines. The more you see this setup, the more you will realise that it's the sensible thing to do to deploy it on all timeframes, as I said last week.
So. At present, the %R is not
quite overbought, and it remains an open question as to whether the CCI
will show a divergence when it does
hit overbought. It turns out that if you're good at "sums", you can
backsolve for the lowest price level that will result in an overbought
%R and a CCI divergence (i.e., that will result in the %R hitting
overbought, and the CCI hitting somewhere in the region covered by the
big blue question mark.
As it happens, I am good at sums - Odin gave me numeracy as compensation for a complete lack of design flair - so I've calculated the key number. I'm not telling you what it is though - but I will give you a hint. In another 2 days it will be 10730 (a little bit below that, actually).
It always pays to do a check of whether a number 'fits' with other calculation methods.
Looking at the ratio relationships
between the swings that connect recent swing highs and lows, you arrive
at an approximate swing high projection of 10720-10730, and a timeframe
that centres on July 14th-July 18th. That would also make the middle
swing the longest in both time and price (satisfying a very weak Elliott Wave idea, although the pattern does not conform to any "standard" wave that I can see).
To get all the technicals to line up that nicely - without some real finagling of the numbers - is reasonably rare, and less than one trader in a thousand would ever do it. I do point out that for the Monthly chart to get back to overbought would require a price level in the high-10800s, but since it's already been both overbought and CCI-divergent, I don't think that's too much of a concern.
The Technical Summary can be expressed thus: the
Dow is close - in both time and price - to a significant pullback. the
highest-probability area for this to occur is in the low-to-mid 10700s;
as such, hourly overbought levels above 10650 represent reasonably safe
shorting levels, particularly on extreme TICK readings.
