Volume Analysis

October 30, 2004

Mr Joseph Granville is generally acknowledged to be the father of the concept of On-Balance-Volume. He is one of three giants in the field of Technical Analysis of share price movements who made their mark in the 1960s. The other two were the incredibly talented and unbelievably hard working Mr Richard Russell (of Dow Theory Letters Fame), and the now legendary Mr Harry Schultz (who was one of the pioneers of cyclical analysis of share price movements in general, and of the application of the work of Nikolai Kondrat'eff to this field of endeavour in particular).

Of course, there were others, such as Mr Ian McAvity and the Aden Sisters. And we should not forget the Bank Credit Analyst which - amongst other things - resurrected the work of Mr Elliott of Elliott Wave fame. This work became the foundational knowledge base of important analysts such as Mr Bob Prechter and Mr Robert Beckman.

But credit where credit is due: The aforementioned three gentlemen towered above everyone else; and the Bank Credit Analyst remains the most credible publication of its kind in the world today.

Although he was not the first to understand the importance of "volume", Mr Granville became famous for his argument that price movement often masks what is really happening behind the scenes. He understood that prices can be manipulated in the short term by organizations with "deep pockets". The following is an example of such manipulation:

An Institution wishes to accumulate a significant line of shares in a particular company. It starts by buying a modest line of these shares (say 100,000), and it thereby pushes the price up as it mops up immediately available volume. It then withdraws, and the act of withdrawing causes weak holders - who have been watching the price move up - to fear that they may have missed the boat. These weak holders place sell orders which causes the share price to react downwards. As the price is in the process of reacting downwards, the same Institution applies (say) 10% of its original 100,000 shares to "force" the price down even further - possibly to a lower level than when it first started buying. This, in turn, causes more weak holders to panic out, and the Institution picks up another line of (say) 100,000 shares. Net result: The share price remains virtually constant, and the Institution has picked up 190,000 shares. "On Balance" there has been accumulation even though the share price has landed up where it started - or may even have fallen.

Mr Granville developed a method of charting the above activity. If a share price closed higher at the end of any day, the volume that changed hands during that day was regarded as having been buying volume, and it was regarded as selling volume if the share price ended up lower at the end of the day. (As an aside, it is for this reason that weekly and monthly OBV statistics are less reliable, because an entire week's volume could be classified as buying or selling pressure if the price is manipulated sufficiently on the Friday's close).

Over the past couple of decades the Financial World has witnessed an explosion in the number of technical analysts in the markets, but it has been noteworthy how few of these analysts seem to take volume into account when they arrive at their conclusions. For example, a "Head-and Shoulders" pattern on the share price charts is relatively common, but very few of these turn out to be true "reversal" patterns because the underlying volume does not validate the pattern. Nevertheless, this widespread failure has been exceptionally valuable to those Technical Analysts who do understand the importance of volume. Why? Because the less wide spread an analytical technique is amongst practitioners, the more reliable it becomes. It is precisely for this reason that Mr Russell's Primary Trend Index is so valuable - there is only one analyst (himself) who knows what the eight constituents of the Index are. (Although buried in his prodigious work spanning the decades he has let most of them slip out in throw away comments J)

Getting back to Mr Granville's work, if we take the concept of OBV and we apply it to a few key charts below, a picture (which is probably pretty reliable because of the limited usage of this technique) starts to emerge.

First, lets look at the Dow Jones Industrial Index (All charts are courtesy of Bigcharts.com)

Focussing on the weekly OBV (less reliable, but probably reasonably accurate over very long periods of time) we note that the OBV chart has been above the zero line since 1995 - notwithstanding the market having peaked in price at the end of 1999. Indeed, in mid 2003, the OBV line even broke to new highs.

In reality it is probable that this "buy" signal more accurately represented a flight to safety by the big end of town, but this is just armchair speculation. The fact is that there has been stubborn buying pressure for over a decade - right up to January 2004.

However, since January 2004, the OBV line has been trending sideways to slightly down and, when you blow the chart up on the dailies, the following picture emerges:

Since mid February 2004, the OBV chart has displayed a series of falling tops and bottoms - which, in terms of Mr Granville's original argument - is indicative of "distribution".

Now let's look at a chart of the gold price - which does not have volume (chart courtesy http://tfc-charts.w2d.com/chart/GD/M

This chart clearly shows a price bottom in early 2001.

There are two bell weather charts in the gold shares that do reflect volume and which could probably be used as proxies for the market as a whole. These are Newmont Mining and Anglo Gold. So let's look at their OBV charts:

The above chart is particularly interesting because it shows that whilst the OBV chart did indeed bottom out in early 2001 - and rise strongly in 2002 - disbelief (or perhaps something else?) started to manifest in mid 2002, and large volumes of shares were subsequently distributed by holders who appeared to have believed that the gold price had seen its peak. Note how the selling pressure became so exaggerated that the OBV line even went into negative territory in early 2004 and then again in mid 2004.

Using Mr Granville's logic, the volume of NEM shares that changed hands from 2002 onwards was primarily selling pressure as shares were distributed.

One would expect to see a similar pattern on the chart of the world's second largest gold mining house, namely Anglo Gold, which is shown below:

And this is where things start to get very interesting.

Yes, Anglo's OBV chart also shows selling pressure coming in during mid 2002 but, unlike NEM's chart AU's OBV did not get into negative territory in 2003. Indeed, On Balance, it has been rising since early 2003.

Importantly, if one draws some trend lines on both of the OBV charts above, one comes to the following conclusions:

  1. NEM's OBV chart has recently given a buy signal as it broke up through a falling trendline
  2. AU's OBV chart is approaching the apex of a triangle from which - based on NEM's chart behaviour - it might reasonably be expected to also break up.

There is in all likelyhood a very interesting story that underlies the differences in the OBV charts of these two counters, but it is probable that we will have to wait a couple of years to see which one is the "real" powerhouse in the gold industry. My bet - based on the above - is that the Oppenheimer family controlled Anglo American group will probably land up as the world leader, but we will have to wait and see.

Now let's turn our attention to a couple of important bell weather stocks in the US economy.

First, let's look at General Motors:

This chart reflects significant indecision amongst investors. Note how the OBV dipped into negative territory in late 2002 and again in early 2003, and then rose to new highs (even as price peaked at a lower high), only to plummet back to its old level. On Balance, there has been neither net accumulation nor significant net distribution in over a decade - but there has been a lot of activity in the last couple of years.

Because Walmart is the largest retailer in the USA and, as Consumer Purchasing represents 66% of the US GDP, Investor perceptions of the fortunes of this particular business are particularly significant.

On Balance, there has clearly been net accumulation of this company's shares since around 1996 with only a brief dip in mid 2000 before investors once again became sufficiently confident to stay in the market.

Notwithstanding this, the daily chart of Walmart ALSO shows net distribution since around February 2004.

On Balance, around 200 million shares of this company have been distributed in the last eight months.

Now let's look at the OBV of one of the world's largest and most sophisticated Financial Institutions, Citigroup:

Well, there's not much ambiguity here. On Balance, there's been net distribution since the market as a whole peaked in 1999/2000.

What about JP Morgan - the world's largest marketer of derivative instruments?

This one is very interesting. Despite all the hullabaloo regarding the "risks" of its vast exposure to derivatives, the market appears to be relatively sanguine. On Balance, there has been a strong recovery since the OBV chart bottomed (in negative territory) in mid 2002.

But hold on a second, look how the OBV chart fell from +0.5 billion to -0.5 billion from mid 2001 to mid 2002. And look at how the OBV chart is now also coming to the apex of a triangle of indecision - within the context of the most recent high having been lower than the peak of 2001. Text Book theory says that in the context of falling tops, the most likely direction of the breakout will be "down", but one cannot anticipate such an outcome. Conceivably, the entire market could rise to enter a new upward phase.

In this context, the direction of the break from this apex is likely to be hugely significant - particularly given Citi's OBV chart.

What about GE?

This one has been relatively strong, but with indecision clearly manifest since mid 2002.

Conclusions

So where does all this leave us? What conclusions can we draw based on On-Balance -Volume?

  • First, it is clear from all of the charts that Bullishness has been absent in the recent past. Indeed, On Balance, there has been distribution since around mid February 2004.
  • Next, if we look past the individual counters to the Industrial Index as a whole, the market is clearly showing that the big boys have been quietly exiting since mid February.

Now I want to hone in (for the nth time) on the phenomenon of falling volume whilst prices were rising on the Dow Jones Industrial Index during 2003.

The reader' attention is drawn to the point on the price and volume chart represented by December 2002 - ie about 1 cm to the left of the number 2003. That is the point at which volume last peaked

If you take pen and a ruler, and you join the tops of the volume bars from around February 2003 to February 2004, you will note that the trendline points downwards. Yes, the OBV line trended upwards during that period, because On Balance, share prices were rising during that period. But it is precisely because prices rose whilst volume was contracting that the volume oriented Technical Analyst was alerted to a structural deterioration that was taking place in the market as a whole.

All of which brings us back to the ultimate (and over riding) power of Dow Theory, and the following is a direct quote from the book entitled "Technical Analysis of Stock Trends" by Edwards and Magee, 5th Edition, 1977 at page 19 - in the chapter that discusses Dow Theory:

".. in a Bull Market, volume increases when prices rise and dwindles as prices decline; in a Bear Market turnover increases when prices drop and dries up as they recover"

In the end analysis, history is likely to show that Mr Richard Russell was indeed "king" of all the Technical Analysts. He, virtually alone amongst Stock Market analysts, called a commencement of the Primary Bear market as both the Dow Jones Transport Index and the Dow Jones Industrial Index fell to lower lows at a time when prices had lost connection with the reality of underlying value.

But the bottom line is this: The ultimate reason why Dow Theory is so powerful is that it has - as one of its base tenets - the concept that "Volume Goes with the Trend"; and it is a matter of quiet amazement to me that so few of the vast number of Technical Analysts today even pay lip service to this fundamental and over riding concept.

It was this falling volume that provided absolutely incontrovertible evidence that what we were witnessing during 2003 was indeed a technical upward reaction within the confines of a Primary Bear Market.

Finally, it was On-Balance-Volume analysis that provided absolutely incontrovertible evidence that there has been net distribution of shares since mid February 2004.

We have volume analysis to thank for confirmation that the next significant technical signals are likely to be "SELL" in the Industrial Equities, and "BUY" in the Gold and Silver equities.

Mr Richard Russell and Mr Joseph Granville, I salute you.

 

Brian Bloom

October 30th, 2004

U.S. ranks third in world gold production with 240 tons per year