Bullish Percent Index: Overbought Or Bull Market?
5 Comments Published August 6th, 2009 in Technical AnalysisOne of my favorite breadth charts is the Bullish Percent Index. It isn’t as common as advance decline measures because it is based on point and figure charting.
Point and figure charting itself is based on pure price action and ignores both time and even small price movements. In P&F charts the X represents demand (or bid or buying) and the O supply (or selling). Unlike candlestick or bar charts, it is perfectly normal for a point and figure chart to not need to be updated (when price doesn’t move beyond a threshold either up or down). This is the great advantage of point and figure charts.
Other than that, the basics of technical analysis such as support and resistance apply to point and figure charts. Also, point and figure charts can provide systematic buy and sell signals - something that ‘normal’ charts leave open to the traders discretion. The simplest buy signal is the chart to the left - when price breaks above a previous high (without the column of O’s breaking below their previous low).
So to calculate a Bullish Percent Indexes, we simply take a look at each and every stock compromising an index and track how many of them out of the total constituents are in a point and figure buy signal. If, 120 out of the 500 stocks in the S&P 500 index are in a buy signal for example, then the Bullish Percent Index for the S&P 500 for that day would be 24%.
Doing this by hand would be extremely cumbersome, but thankfully we have computers that can do the calculations in a fraction of a second. Here is the chart of the Bullish Percent Index for the Nasdaq Composite:

According to the traditional interpretation, a Bullish Percent Index of 70% and higher is considered overbought. And if there it experiences a 6% (or more) decline, it will offer a sell signal. Personally, I prefer to not wait for the sell signal. Once you know that the market has weak legs, you can use other indicators to give you more short term guidance.
Right now we are seeing almost every single measure of the market provide extreme breadth levels from the Bullish Percent Index. Normally seeing the bullish percent indexes for so many markets and sectors reaching this high in synchronicity would be a red flag. However, there is an argument for such strong momentum to be a signal of a protracted rally.
This was the same concept that I shared earlier this week about incredibly powerful thrusts measured from the advance decline breadth. These short busts of powerful buying are usually precursors to lasting uptrends. Think of it as a turbo booster on a rocket which lifts it through the heavy atmosphere before it can glide easier through the thin air of space.
Consider that the last time the Nasdaq composite BPI was this high was back in 2003-2004 during a powerful bull market. It is the same case for the NYSE, and the Nasdaq 100 index. This is how oversold can become meaningless.
Bullish Percent for Major Indexes:
- NYSE Index — 77%
- Standard & Poor’s 500 — 83%
- Nasdaq Composite — 70.55%
- Nasdaq 100 Index — 90%
- Dow Jones Industrial — 77%
Bullish Percent for Sectors:
- Financial — 86%
- Consumer Discretionary — 79%
- Consumer Staples — 73%
- Energy — 60%
- Healthcare — 79%
- Technology — 87%
- Industrial — 71%
- Materials — 85%
- Telecom — 56%
- Utilities — 62%
- Transportation — 75%
The recent price action in Accredited Home Lenders (LEND) reminds me of “the Great New Pattern” invented by praetorian2 (aka Harris B. Kupperman) over on elitetrader.com

The pattern is massive capitulation by the longs marked by sustained selling. The key factors are not only back to back price delines but seeing a decline accelerate (wider range candlesticks to the downside). Volume is also an important factor since it tells you how more and more longs are getting sucked into the selling by having their stop losses triggered or just panicking.
At the maximum point of anxiety, you watch for possible reversal. And at that first glimmer of light, you start scaling in. Kupperman actually scaled in as price was falling, anticipating the V bounce. This is very risky but he had made this pattern his own and knew instinctively how to lower his basis point without getting his head chopped off. The exit strategy is also piecemeal, if I remember correctly.
Here’s an example from the archives (way back in 2001):

For more information, on this set up check out the original thread on the GNP by praetorian2 and this follow up one by Robert Tharp.
We hear a lot that success in trading comes from finding your own way. This is a great example of that. Since Harris first ‘found’ this pattern it has continued to be repeated over and over again in the markets. And by consistently playing this setup he went on to great success as a trader.
Last I heard, Kupperman is still running his hedge fund, Praetorian Capital Management (with an offshore clone), and has assets of ~$6 million under management.
I noticed something really strange with the chart of Ascent Solar (ASTI) served up by QuoteTracker yesterday. I didn’t want to mention it before having Jerry Medved take a look and see if I was making a silly mistake (as I usually do). As always Jerry is extremely helpful and lightning fast to reply - does he take Jolt Cola on IV?
First I checked other charts to confirm what the correct chart looked like. Then I narrowed it between IB and QT. I checked IB’s own chart (yes they have charts but they suck). Their chart of ASTI matched other non-QT charts. So IB’s data and charting was fine. It was definitely how QT was getting the data or drawing the chart.
We still haven’t been able to figure out what is causing the problem. I’ve tried clearing the data from QT and re-backfilling. But the chart looks exactly the same (messed up). Then I double checked that I didn’t have Heikin-Ashi technique applied to the chart. Nope. Just candlesticks.
It looks like only the first few hours are effected. I’ve included a screengrab which shows the two charts together so you can see what I mean. Click to see full size image:
![]()
Anyone else have this problem? If you’ve got QT installed, pull up a chart of ASTI and drop me a comment.


Recent Comments