On-balance volume (OBV) is a proven-effective leading indicator that allows traders to spot turning points and valid signals across a wide variety of markets and time frames.
For the majority of technical analysts, volume plays an important role. Unfortunately, simply comparing one day’s volume to a three-month average will not tell you much about whether money is flowing in or out of a particular market or stock.
In the late 1970’s, my father gave me a book by Joseph Granville titled Granville's New Strategy of Daily Stock Market Timing for Maximum Profit. A few years later, I found Granville’s on-balance volume (OBV) on Compu Trac, one of the earliest technical analysis software programs. I was quickly hooked on the OBV. From the following examples, as well as my daily Charts in Play column, I think you will see why it is my favorite indicator for all markets.
Joe developed the OBV as a way to determine whether the smart money was buying or selling. It is calculated by keeping a running total of the volume figures and then adding in the volume if the close was higher than the previous period, or subtracting the volume if the closing price was lower.
If you are doing this in a spreadsheet, such as Excel, the starting volume can be arbitrary, as it is the pattern of the OBV—not the absolute number—that is important. When viewed on a monthly or weekly basis, this can be very useful in identifying major trends.
From the start, I analyzed the OBV in the same way that I would analyze a price chart. I used trend lines, moving averages, and support/resistance analysis to determine whether the OBV was positive or negative.
Divergence analysis was always quite important, though divergences are not always observed at every important turning point. Like my early work on Welles Wilder’s Relative Strength Index (RSI), it was critical to use divergence analysis on multiple time frames in order to generate valid signals.
In May 1985, my analysis of the OBV on the major currencies was instrumental in helping me identify major bottoms in currencies like the Deutsche mark (DMK) and Swiss franc (CHF), and therefore, the top in the US dollar (USD). At the time, most of the leading economists were expecting the dollar to remain strong for a few years. The dollar had bottomed in November 1980 with the election of Ronald Reagan.
In a May 21, 1985 appearance on the Financial News Network, a precursor to CNBC, I discussed a chart very similar to the one above of the Deutsche mark futures contract traded on the CME. In my early adaptation of the OBV, I had also added a 21-period weighted moving average (WMA) of the OBV, which is plotted in green.
The weekly chart shows that the DMK was in a well-established downtrend (line a) starting in 1982, as it was falling in value against the US dollar. In 1984, the decline accelerated, as it fell 25% to a low of 0.2881, or 3.47 DMK per USD. By comparison, in early 1984, it was 2.5 DMK per USD.
During this decline, the OBV rallied several times to its declining weighted moving average. Then in March 1985, the OBV moved above its weighted moving average and broke its downtrend, line c. The WMA flattened out over the next six weeks before starting to rise.
On several attempts, the OBV was unable break through resistance (line d). With the OBV now above its rising weighted moving average and with confirming bullish signals in the analysis of the CHF and British pound (GBP), it suggested these currencies had bottomed out. The daily technical studies had been positive on the currencies for several months, and this was another negative for the dollar.
The OBV overcame its resistance, line d, on July 5, 1985 (line a). One week later, the DMK also broke through its corresponding resistance, line d. This is one of the reasons I find the OBV to be such a valuable indicator, as it often leads prices by one or more periods. Obviously, this can make the risk/reward on new positions much more favorable.
The DMK tested its downtrend (line a) in August and had a sharp, three-week pullback. During this time, the OBV was much stronger, as it held well above its rising weighted moving average.
In late September, the G5 nations got together over a weekend at the Plaza Hotel in New York and agreed to devalue the dollar. As you can see on the above chart, the DMK futures gapped higher and accelerated to the upside, as there was concerted intervention to lower the dollar. By early 1988, the DMK had more than doubled.
NEXT: Applying OBV Analysis Across Other Markets
I have found that the OBV works on any market that has good volume, and I have long advised cash forex traders to keep an eye on the currency futures, where the volume data is very reliable.
One of the most simplistic ways to use the OBV is to see if it makes a new high with each price high in an uptrend, or makes a new low with prices in a downtrend. For the past five years or so, I have been reporting on the monthly OBV analysis of the gold futures. The arrows on the above monthly chart of gold reflect the new monthly OBV highs going back to 2003.
On the chart, you will see that each new high in the gold futures has been confirmed by a new high in the monthly OBV. The last closing monthly high in the gold futures was in April and was supported by a convincing new high in the OBV. It is also clear from the chart that the 21-period WMA has acted as a good level of support, as tests of the rising WMA have often marked correction lows.
As I referred to earlier, the OBV is the only indicator that routinely will break out ahead of prices. At the end of December 2008, the OBV closed above resistance at line c when gold closed at $884. It was not until the end of February that the gold futures overcame trend line resistance, line a, at $928.
More serious students of the OBV can also watch for when the indicator is rising or falling more sharply than prices. A good example occurred in the fall of 2010 when the OBV was rising much more sharply (see circle) than gold prices. The chart indicates that gold prices did catch up over the next few months.
If you can’t look at the monthly, weekly, and daily data, then at least look at the weekly and daily data. One of my favorite patterns to watch for is the weekly bottoming formation in terms of price and the OBV.
For 18 months, the OBV for corn was in a trading range, lines e and f. While the price chart was forming lower highs, line d, the OBV was forming higher highs (line e). This was a bullish sign.
In mid-August, corn prices closed above resistance at 412, and this was confirmed by the breakout in the OBV, point 2. For the next eight months, both corn and the OBV were moving sharply higher. The OBV tested its rising weighted moving average in the latter part of November, which presented a good buying opportunity.
Corn prices peaked in April and the new highs were confirmed by a new high in the OBV. But in June, when corn made a further new high, line g, the OBV formed a lower high, line i. The next week, the OBV dropped below its weighted moving average and corn subsequently broke support at line h.
In most cases, an eight-to ten-week divergence in the weekly OBV can lead to a multi-month correction. For corn, it will be important to see if the uptrend in the OBV (line j) does hold.
NEXT: Sector Analysis Using On-Balance Volume (OBV)
The OBV is also a valuable tool when analyzing the various market sectors. The Dow Transportation Average is one that I monitor closely. The weekly OBV on the Transports confirmed the March 2009 lows before rising sharply, and by that summer, it was in a clear uptrend.
The OBV peaked in August and then developed a trading range, as indicated by the resistance at line b. The Transports had a corresponding level of resistance at 4287, line a. In late February, the OBV moved above its weighted moving average and two weeks later overcame the resistance at line b. This breakout coincided (line 1) with the close in the Transports above the resistance at line a.
The Transports rallied almost 8% before peaking at the end of April 2010. The Dow Transports consolidated for the next four months, but in the first week of September (line 2), the OBV moved to new highs when the resistance at line d was overcome. The Transports did not overcome the corresponding resistance at line c until eight weeks later, in November.
The Transports made a short-term peak in mid-February 2011 and then declined 7.9% from the highs. The OBV held up much better than prices, and just three weeks later, it made new highs (line 3) and again lead prices higher.
Both the Transports and the OBV made new highs in early May, but as the Transports made a new high the week ending July 7 (line e), the OBV formed a lower high. This may be a significant divergence and does warrant close watching. A drop below its WMA would be an additional warning sign. A violation of support at line g would indicate that an interim top was in place.
NEXT: Trading Case Study Using Energy ETF XLE
As I stressed earlier, it is important to analyze the OBV in at least two time periods. The Select Spyder SPDR - Energy (XLE) is an ETF that I have covered regularly over the past six months.
XLE formed a series of lower lows (line 1) in the summer of 2010 and reached $48.56 in early July. In contrast, the OBV bottomed in June and formed a series of highs lows, line 2. The bullish divergence suggested that XLE was bottoming.
The initial downtrend in the OBV, line c, was overcome in early October, and when the intervening peak in the OBV was overcome, it confirmed the divergence. The longer-term OBV resistance at line b was overcome in the first week of November (line 3).
A month later, in December, XLE closed above key resistance, line a, which was a bullish sign.
The OBV surged higher for the next five months, confirming each high in XLE, and peaked in early-April 2010. As XLE was making a new high at the end of April, the OBV started to diverge, line d. It violated its weighted moving average in early May and formed lower highs, as noted in the May 12 article “Big Oil’s Big Top.”
The OBV is currently turning up from stronger support dating back to early 2010, but it is below its declining weighted moving average.
So what was the daily analysis telling us during this period? The daily OBV also surged in early November and confirmed each new price high through early March, line g. The daily OBV broke its uptrend, line f, on March 8 when XLE dropped back to first good support in the $73 area.
Because the weekly OBV was positive and had not formed any negative divergences, the action in the daily OBV was consistent with a correction. By March 28, the daily OBV had once again moved back above its weighted moving average. The daily OBV made a new high in April, but the weekly OBV did not confirm those highs.
This made the short-term uptrend in the daily OBV, line h, an important level to watch. Support was violated on May 3, and when combined with the weekly analysis, it suggested that a top was in place. The break in the daily OBV below the March lows provided further evidence.
XLE has not yet even retraced 38.2% of its major rally, but there are some early indications that a bottom could be formed before the end of the summer. To get the “all clear” signal, the weekly OBV needs to move above its weighted moving average, and the WMA must also start to flatten out, as it is currently declining. The daily OBV is still locked in a trading range, lines j and i, but traders should watch for a breakout.
Since there are many more market examples I would like to share with you, I may do a follow-up article if there is interest. Unfortunately, that’s all for this week. Remember, the OBV can also be useful on intraday charts if the market you are following is quite liquid. It’s essential to analyze at least two different time periods, however, and for long-term trend analysis, the monthly OBV is best. Most of the free, online charting services do include OBV analysis, so I encourage everyone to test it on their own.