A Note about the Cumulative Volume Index (CVI)

The Cumulative Volume Index is one of my favorite market strength indicators. This momentum indicator works by subtracting all of the volume of the declining equities from all of the equities that are gaining. The resulting positive or negative number is then added to the previous results. This indicator is very useful in determining how much strength a bullish or bearish move has behind it.

In addition, if there is a divergence in the direction of the CVI in contrast to the S&P, it often serves as a warning sign that there will be a direction change in the market. If the CVI begins to trend down while the averages are moving upwards, it is a bearish sign. If it begins to trend up while the averages are moving downwards, it is a bullish sign.

I'd like to show how useful the CVI can be by giving three real examples of it predicting a change in market direction.

This first chart shows a weekly chart of the SPY ETF (basically the S&P 500) in yellow and the CVI in red and green. The time period that is important here is June 2007 to October 2007.  You can see that the CVI made a high in June 2007 and then failed to meet that high again in October. In contrast, the S&P 500 made a new high in October. The divergence here was a strong indicator that something was wrong with the S&P's last rally. After that high, the S&P fairly rapidly fell from 1576.09 to 666.79. 

This is a daily chart of the CVI versus the SPY in April 2010. If you remember, we had a pretty nasty 15% correction in April. Once again, the CVI was warning people that were looking for the signal. On April 26th the S&P hit a new recent high of 1219.80. However, the CVI failed to make a new high. What followed was a market correction down to about 1040; completing a 100% retracement of the previous market uptrend.  Don't think however that all signals from the CVI will warn you of that big of a drop. When the divergence happens it can be for market adjustments of all sizes and it can also give you bullish signals.

Here is a daily chart for the relevant period from June 3rd to June 10th, 2010. Here the CVI made a new relative high on June 10th in comparison to June 3rd. In contrast, the S&P was lower on June 10th than it was on June 3rd. This was a bullish divergence signal. After this indicator made the signal the market moved from 1087.85 to 1131.23 (about a 4% upward move).

I hope these examples give you a good idea of how to spot divergence between the S&P 500 and the CVI. The signals don't always work, but they work often enough that if you combine them with some other indicators or fundamental data, you can have a leg up on the market. 

Please remember to always do your own research before investing. If you would prefer for an investment advisor to handle your investments decisions please feel free to contact us as panopticmc@gmail.com.

Fibonacci Numbers and Moving Averages during the Market Madness

Despite Thursday's record breaking market "anomaly" nearly giving Asif and I coronaries, I thought it created an opportune time to introduce a technical analysis technique that played an important role in the selloff. I'll also use three S&P 500  Charts to go over some of the more interesting moving average relationships with the S&P's recent price movements that occurred around the flash crash.

After an 8% market correction that ended in early February, the S&P started on an impressive rally that ended on April 26th. During that period, the S&P advanced approximately 14.5%. On February 25th, the S&P began using the 20-day moving average as support. I have indicated where the S&P used the 20-day as support with green arrows on the chart below. The S&P finally dropped below its 20-day moving average on On 4/27. This is considered by to be a bearish indicator. On 5/4, the 10-day moving average fell below the 20-day moving average. Many traders will switch from going long to short when they see that type of crossover. However, remember that moving averages are lagging indicators and can be prone to whipsaws (multiple consecutive crossovers resulting in false signals). If you look at 2/22, you will see that there was a bullish crossover of the 10-day over the 20-day. That was thirteen days after the rally had already begun.

 Once it becomes evident that a stock may be in for a correction, people begin to wonder how far it is will drop. In order to arrive at some estimates, you need to look for likely support levels. As stated in a previous article, common moving averages such as the 50, 100, and 200-day moving averages are likely targets. On 5/4 (the light blue arrow), the market clearly bounced off the 50-day moving average. The subsequent day it opened below the 50-day. Other common price levels for a stock to find support are areas of demand. Where there was previous resistance that has been broke, the general rule of thumb is this price now serves as support. 1150 in this case, because of the strong resistance at that level in mid January. Instead of the S&P falling to any of those areas of support, the market closed at a Fibonacci retracement level.

In technical analysis, common retracements of a stock occur at 38.2%, 50%, 61.8%, and 100%. These percentages are all Fibonacci numbers. Why would a stock retrace to these levels? Some people argue it's another case of Fibonacci numbers occurring in nature. Personally, I think it's another case of a self-fulfilling prophecy, but I don't think it is worth arguing over why it works just as long as it actually does work. In order to determine the Fibonacci retracement levels, I took the lowest and highest prices achieved from the rally bottom on February 5th to the rally top on April 26th. The charting software then automatically determines where the Fibonacci levels are located. If you look at the chart above, you can see that on 5/6 the S&P closed at exactly a 50% retracement of its previous rally.* Why is predicting that information useful? Well, if you were short the market that day, you would have an idea of where to cover your short. Or, if you were long, you would have some idea of how much more pain to expect. Since this level serves as some support, the market may rally off of it. Or, it may just serve as a breathing point before the market continues lower (likely to another major level of support.)

*At the time I captured the picture of the chart, there had not yet been a cancellation of a large number of trades that had occurred on 5/6/2010. Once those trades were invalidated, the S&P's official closing price was changed to 1128.15. Given that before these changes were made, the closing price of the S&P was at the 50% retracement level, I believe that the data is still valid.

Support levels do not always work, but they work often enough that they are extremely useful. PMC often tries to buy stocks at or near major support levels. It provides a little insurance that the stock will not go significantly lower. Ideally, you can purchase a stock with multiple support levels directly under where you are purchasing it. Think of them as potential safety nets to break the stock’s fall. It is all too common that if a stock breaks one support level it will drop until it reaches the next one. There are real reasons for why this happens. People have stop loss orders right below support levels that will sell their stock automatically if that level is breached. Traders will get short a stock that has broken a support level and will not cover their short until they reach the next support level. If you know what price levels of a stock or index where the demand is likely to increase, then you can gain a predictive edge against people that completely ignore this information.

 If I haven’t yet convinced you that Fibonacci's and moving averages can affect the intraday fluctuations and closing prices of stocks, let's look at what happened to the S&P on Friday, 5/7. In this first chart you can see that the S&P had its intraday low occur just below the 200-day moving average, here pointed out with a green arrow.

Now, look at where the S&P closed for the day. Though it looks like it closed slightly below 61.8%, on smaller time scale charts, like a 5-minute chart, or hourly chart, you can see that the market closed only very narrowly above 61.8%; reaffirming 61.8% retracement as a support level.

The previous two charts were both daily charts, but I want to demonstrate how moving averages also work on other time frames. Below is a weekly chart of the S&P 500. Thursday was one of the craziest days the stock market has ever experienced. Many members of the media were asking what caused the market to turn around so suddenly. Well, the chart below shows you exactly why that happened. Take a look at the green arrow. The market bottomed at the 50-week moving average. Once this occurred, people, and more importantly, automated trading algorithms switched from being short to being long.  Even in times of market madness, it can follow predictable patterns. Interesting stuff, eh? 

I hope you can see from these examples how powerful moving averages and Fibonacci levels are over market fluctuations. As previously stated, they do not always work. Even when they do work it is often difficult to know exactly which levels are the ones to watch for. Having a long history of following a stock and its price movements can increase the accuracy of your predictions.

If you like the information provided in the article but would prefer for an investment advisory to handle your investment decisions, please feel free to contact us at panopticmc@gmail.com.

A Supplemental Note on MACD

This note is a supplement to the one below it. I felt it may be useful to show how MACD can be an effective tool on more than just a daily chart of a stock. MACD can be used on any time scale chart and works with indexes in addition to individual stocks. Below is a monthly chart of the S&P 500. The chart covers the period from late 1996 until the present. During that period, the S&P has crashed twice. One was the .com bubble burst that lasted from 2000 until 2003 and the other was the mortgage crisis crash that took place mostly during 2008.

Remember, usually the MACD histogram bar chart and the stock price chart move in the same direction. That is, if the stock price is going up the MACD histogram is going up and vice-versa. If they are moving in opposite directions, this is a signal that the a directional change may occur in the stock so that it will again be moving in the same direction as the MACD histogram. In those situations, the MACD histogram is a leading indicator. It gives you an idea of what the price of the stock or the index may do in the future.

Sadly, the MACD histogram did not call the bottom of the 2008 market crash. There was no divergence in the MACD histogram and the S&P price to give a signal. The MACD line crossing the signal line did serve as a useful indicator, but it was a lagging indicator. There was already a substantial recovery occurring in the S&P before MACD gave any signal of it. However, MACD divergence did make an accurate and timely prediction of the market top in 2000, the turnaround in 2003, and the market top in 2007. I'm going to discuss the 2003 turnaround and the market top in 2007 because they are the easiest to see.

The first market top occurred in 2000. From that point you can see that the S&P fell from about 1550 to roughly 765 (a downtrend) between the years 2000 and 2003. During that period, the MACD histogram had a pretty deliberate trend moving from the negative to the positive (an uptrend). In early 2003, the MACD line crossed over the gray signal line and the histogram turned positive. From that point, the market continued its trend upward until late 2007. However, the MACD histogram had been trending down towards the negative and finally went negative with the MACD line crossing below the gray signal line in late 2007.

In both situations, people paying close attention to the MACD on the monthly S&P 500 chart would have had a pretty good chance of catching both the bottoms and the tops of the market. Technical investors like to see confirmation of a signal. In the case of the 2007 market top, not only was the MACD signaling that things were going to make a turn for the worse, there was also a "double top". Do you see how the market basically topped out in 2007 at the same point it did in 2000? This is a very common chart pattern that technicians look for, but I'll save further discussion on that topic for another time.