The arrows highlight points where the indicator reverses from a reading at or above 3.2. Such action indicates danger, but unfortunately, does not tell us how much. Thus, the late-2007 sell signal was followed by a severe bear market, whereas most of the others were fairly limited in their bearish warnings.
CHART 29.4 S&P Composite, 1995–2012, and Bulls Less Bears at Tops
Chart 29.5 shows the same period, but this time from the aspect of the more reliable buy signals, triggered when the indicator reverses from the 0.8 level. There are two observations worth making. First, the indicator reaches this reading more often in a primary bear market and generates relatively weak signals. Second, when it does slip below 0.8 and starts to rise, this represents a high-probability rally signal.
CHART 29.5 S&P Composite, 1995–2012, and Bulls Less Bears at Bottoms
Chart 29.6 takes a slightly different tack. First, it just represents the number of bearish advisory services that has been inversely plotted to correspond with stock market price movements. Second, the actual plotted series is a 9-week RSI that has been smoothed with an 8-week MA. Buy signals have been flagged with oversold upside reversals. The dashed lines, on the other hand, show that sell signals typically come after some form of negative divergence has materialized. If you look at the previous chart, you will also discover that bull market peaks are usually formed not at the point of maximum bullishness, but at some lower peak that develops afterward.
CHART 29.6 DJIA, 1995–2012, and Inverted Bearish Momentum
Major Technical Principle An extremely high sentiment reading is not usually that bearish, as it is often a leading indicator. Bull market highs have a greater tendency to form when sentiment diverges in a negative way.
Finally Chart 29.7 shows that sentiment data, just like any other oscillator, occasionally lend itself to trendline analysis. The chart also reminds us that the principles of divergence can also be applied to the interpretation of this and similar data. For example, the market low of 1982 and high of 1987 were both preceded by a divergence.
CHART 29.7 S&P Composite, 1976–2001, and Bears
Quite often, the trend of sentiment can be as important as the level in identifying important market reversals. The chart shows many examples where it is possible to construct a trendline for the sentiment series. When the line is violated, a trend-reversal signal is given. The dashed vertical lines indicate sells, and the solid lines buy signals.
Market Vane and Bond Market Sentiment
Sentiment indicators are also published for the futures market. One of the most widely followed contains data issued by Market Vane. Each week, the firm polls a sample of market participants. The results are published as the percentage of participants that are bullish. The theory is that when a significant number of traders are bullish on a particular market, they are already positioned on the long side. That means that there is very little potential buying power left. The implication is that the price has only one way to go, and that is down. In a similar manner, if most participants are bearish, they have already sold or gone short. Since selling pressure has likely reached an extreme, prices can only move in one direction and that is north. An example in Chart 29.8 using reversals from the extreme levels of 85 percent and 15 percent indicates that such extended reversals offer good signals. Unfortunately, they are few and far between. An alternative method of interpretation is to construct trendlines for the sentiment series and observe when the price is reacting with a similar break of its own. Once again, we do not see a lot of signals, but when they are available, the result is quite effective.
CHART 29.8 20-Year Government Bond Yield, 1994–2012, and Market Vane Bond Bulls
One problem with these statistics is that they are based on the opinion of short-term traders, which makes them somewhat erratic, with implications limited to near-term price movements. A technique for surmounting this drawback is to calculate an MA of the raw data, say, with a 4-week span, which smoothes out week-to-week fluctuations.
An alternative is to plot a 10-week MA of a 14-week RSI as shown in Chart 29.9. In this instance, the overbought/oversold zones are constructed at 40 percent and 60 percent. The arrows show reversals from beyond those levels that offer fairly consistent and reliable signals of changes in the trend of the 20-year yield series.
CHART 29.9 20-Year Government Bond Yield, 2000–2012, and Market Vane Bond Bull Momentum
Major Technical Principle Quite often, the trend of sentiment can be as important as the level in identifying important market reversals. The same observation can be made for fundamental indicators, such as price earnings (P/E) ratios and so forth.
Combining Sentiment and Momentum
One useful approach for identifying early reversals in trend is to combine sentiment and momentum into one series, the combo. The bottom panel of Chart 29.10 shows an indicator that combines the Market Vane bulls with a 14-week RSI of the Barclays 20-year government exchange-traded fund (ETF). Both series have been subtracted from 50 to allow for plus and minus numbers, with the total being divided by 2.
CHART 29.10 Barclays 20-Year Trust, 2003–2012, and Two Momentum/Sentiment Indicators
Buy and sell alerts occur when the 10-week MA of the combo, shown in the center window, crosses above and below its 6-week MA from below and above zero, respectively. In other words, a sell signal can only be generated from a reading above zero and a buy from below. Whipsaw signals have been indicated in the ellipses. When it is considered that the bulk of this period included choppy ranging action rather than persistent trends, it seems that this approach worked quite well.
Mutual Funds
Data on mutual funds are published monthly by the Investment Company Institute (ICI.org). The statistics are useful because they monitor the actions of both the public and the institutions. In recent years, money-market and tax-exempt mutual funds have become widespread; therefore, the data used here have been modified to include only equity funds. Technical analysts usually calculate mutual fund cash as a percentage of assets. In a sense, this data should be treated as a flow-of-funds indicator, but it is discussed here as a measure of sentiment.
Mutual Fund Cash/Assets Ratio
Mutual funds consistently hold a certain amount of their portfolios in the form of liquid assets in order to accommodate investors wishing to cash in or redeem their investments. A useful indicator is derived when this cash position is expressed as a percentage of the total value of mutual funds’ portfolios, a figure known as total asset value (see Chart 29.11).
CHART 29.11 S&P Composite, 1970–2013, and Mutual Fund Cash/Asset Ratio
The index moves in the direction opposite to the stock market because the proportion of cash held by mutual funds rises as prices fall, and vice versa. There are three reasons for this characteristic. First, as the value of a fund’s portfolio falls in a declining market, the proportion of cash held will automatically rise even though no new cash is raised. Second, as prices decline, the funds become more cautious in their buying policy since they see fewer opportunities for capital gains. Third, the decision is made to hold more cash reserves as insurance against a rush of redemptions by the public. In a rising market, the opposite effect is felt as advancing prices automatically reduce the proportion of cash, sales increase, and fund managers are under tremendous pressure to capitalize on the bull market by being fully invested.
One of the drawbacks of this approach is that mutual fund cash data did, by and large, remain above the 9.5 percent level between 1978 and 1990 and lost a lot of validity as a timing device during this period. It is true that the market was in a rising trend, but one of the functions of an indicator of this nature is to warn of setbacks such as the 1980 and 1981–1982 bear markets, not to mention the 1987 crash.
One way around this problem, originally devised by Norman Fosback of Market Logic, is to subtract the prevailing level of short-term interest rates from the cash percentage levels themselves. In this way, the incentive for portfolio managers
to hold cash due to high interest rates is neutralized. This adjustment to the cash/assets ratio is shown in Chart 29.12. It is a definite improvement on the raw data, but unfortunately, it, too, fails to explain the 1987 crash.
CHART 29.12 DJIA, 1965–2013, and Cash/Asset Ratios
A final alternative, devised by Ned Davis Research, compares switch fund cash and mutual fund managers’ cash to total mutual fund assets. This series is also adjusted for interest rates and appears to offer the best results of all. The labeled buy (B) and sell (S) signals in Chart 29.13 are generated when this series crosses below the lower dashed line; they remain in force until it crosses above the upper dashed (selling) line.
CHART 29.13 S&P Composite, 1965–2013, and a Switch Fund Cash/Asset Ratio
Margin Debt
Trends in margin debt are probably better classified as flow-of-funds indicators, but since the trend and level of margin debt are also good indications of investor confidence (or lack thereof), they are discussed in this section.
Margin debt is money borrowed from brokers and bankers using securities as collateral. The credit is normally used for the purchase of equities. At the beginning of a typical stock market cycle, margin debt is relatively low; it begins to rise very shortly after the final bottom in equity prices. As prices rise, margin traders as a group become more confident, taking on additional debt in order to leverage larger stock positions.
During a primary uptrend, margin debt is a valuable source of new funds for the stock market. The importance of this factor can be appreciated when it is noted that margin debt increased almost tenfold between 1974 and 1987. The difference between stock purchased for cash and stock bought on margin is that margined stock must, at some point, be sold in order to pay off the debt. On the other hand, stock purchased outright can theoretically be held indefinitely. During stock market declines, margin debt reverses its positive role and becomes an important source of stock supply.
This occurs for four reasons. First, the sophistication of margin-oriented investors is relatively superior to that of other market participants. When this group realizes that the potential for capital gains has greatly diminished, a trend of margin liquidation begins. Margin debt has flattened or declined within 3 months of the vast majority of the 14 stock market peaks since 1932.
Second, primary stock market peaks are invariably preceded by rising interest rates, which in turn increase the carrying cost of margin debt, therefore making it less attractive to maintain.
Third, since 1934 the Federal Reserve Board (the “Fed”) has been empowered to set and vary margin requirements, which specify the amount that can be lent by a broker or bank to customers for the purpose of holding securities. This measure was considered necessary in view of the substantial expansion of margin debt that occurred in the late 1920s. The liquidation of this debt pyramid contributed to the severity of the 1929–1932 bear market. When stock prices have been rising strongly for a period of time, speculation develops, often resulting in a sharp rise in margin debt. Sensing that things could get out of control at this stage, the Fed raises the margin requirement, which has the effect of reducing the buying power of the general public from what it might otherwise have been. Normally, it takes several margin-requirement changes to significantly reduce the buying power of these speculators. This is because the substantial advance in the price of stocks—which was responsible for the requirements being raised in the first place—normally creates additional collateral at a rate that is initially sufficient to offset the rise in reserve requirements.
Fourth, the collateral value of the securities used as a basis for the margin debt falls as stock prices decline. The margin speculator is faced with the option of putting up more money or selling stock in order to pay off the debt. At first, the margin call process is reasonably orderly, as most traders have a sufficient cushion of collateral to protect them from the initial drop in prices. Alternatively, those who are undermargined often choose to put up additional collateral or cash. Toward the end of a bear market, prices fall more rapidly, and this unnerving process, combined with the unwillingness or inability of margin customers to come up with additional collateral, triggers a rush of margin calls. This adds substantially to the supply of stock that must be sold, regardless of price. The self-feeding downward spiral of forced liquidation continues until margin debt has contracted to a more manageable level.
Most people think that the level of margin debt is the most important way to interpret this data. It is true that the higher the level, the greater the market’s vulnerability when the numbers begin to contract. Perhaps a better way to express this statistic is to express the level of debt as a percentage of outstanding market capitalization. That way, the true vulnerability of the market would be represented in a more proportionate way. However, I believe that it is the trend of margin debt that is all important because trend reversals signal whether traders are confident, i.e., willing to take on more debt, or pessimistic, i.e., liquidating it. For this reason, margin debt is a useful indicator when expressed in relation to its 12-month MA, as shown in Chart 29.14.
CHART 29.14 S&P Composite, 1980–2012, and NYSE Margin Debt
Crossovers offer confirmation of major trend reversals. Most of the time, this relationship is reliable, but it does encounter whipsaws from time to time, as you can see from the ellipses.
An alternative, shown in Chart 29.15, is to plot a Know Sure Thing (KST) for margin debt (a moving-average convergence divergence [MACD] or stochastic could also be substituted) and use the positive 9-month MA crossovers as buy signals. As you can see, this technique has been pretty successful, but the two dashed arrows remind us that in technical analysis the probabilities are with us but never reach that perfect holy grail level of 100 percent.
CHART 29.15 S&P Composite, 1962–2012, and NYSE Margin Debt Momentum
Sentiment Using Option Data
Sentiment indicators based on short-selling data appear to have been distorted in recent years, in part because of the introduction of listed options and futures. The other side of the coin is that options can themselves be used as a basis for the construction of sentiment indicators. Their performance is far from perfect, but definitely worth consideration.
Put/Call Ratio
Perhaps the most widely followed option-derived indicator is the one that measures the ratio of the volume of puts to the volume of calls. A put gives an investor or trader the option to sell a specific stock index or commodity at a predetermined price over a specified period. In effect, the purchaser of a put is betting that the price of the underlying asset will go down. This is a form of short sale in which the trader’s risk is limited to the cost of the put. (The risk on a short sale is theoretically unlimited.)
A call, on the other hand, is a bet that the underlying asset will rise in price. It gives a purchaser the option to buy a security at a predetermined price over a specified period.
It is normal for call volume to outstrip that of puts, and so the put/call ratio invariably trades below the 1.0 or (100) level. This indicator measures the swings in sentiment between the bulls and the bears. In theory, the lower the ratio, the more bullish the crowd and the more likely the market is to decline, and vice versa. A low ratio means that very few people are buying puts relative to calls, whereas a high one indicates that a larger number of traders than normal are betting that the market will go down.
For stock market data, a good source is the Chicago Board Options Exchange (CBOE) web site (www.CBOE.com), where it’s possible to download historical data. My preference is for the Total Exchange Volume put/call data, as it is all encompassing. It is also possible to obtain other breakdowns such as indexes, equity volume, and so forth.
A 5-day MA of the ratio is shown in Chart 29.16.
CHART 29.16 NYSE Composite, 2007–2010, and a CBOE Put/Call Ratio
Readings in excess of 125 seem to offer good indications of when pessimism has reached an extreme. Sell signals are not as prescient,
and Chart 29.17 shows that they tend to end up with negative divergences rather than reversals from an extreme level.
CHART 29.17 NYSE Composite, 2007–2010, and a CBOE Put/Call Ratio Showing Divergences
Negative divergences appear at B, C, and D. A positive one can be observed at A. We also see a positive divergence at the 2009 bottom in Chart 29.18.
CHART 29.18 NYSE Composite, 2004–2012, and a CBOE Put/Call Ratio Showing Divergences
This one shows a different approach, with a 35-day smoothing of a 25-day MA of the raw data. The arrows flag reversals from overextended levels, which have had limited success. The two dotted arrows between 2004 and 2008 show how the bull market high was preceded by a series of weaker and weaker peaks in the ratio. Conversely, the two solid arrows a couple of years later flag a positive setup.
The VIX
The Market Volatility Indicator, or VIX, is a trademarked ticker symbol for the Chicago Board Options Exchange Market Volatility Index, a popular measure of the implied volatility of S&P 500 index options. It is often referred to as the fear index and represents one measure of the market’s expectation of stock market volatility over the next 30-day period. The VIX is quoted as a percentage estimating the implied volatility of the market, which is the expected annualized movement of the S&P 500 over the next 30 days.
When prices are trending steadily upwards, there is generally a declining level of volatility as complacency sets in. Conversely, when a market is falling, a growing level of fear results in an expanding trend of volatility. As a contrarian indicator, the higher the VIX, the more bullish the market is and conversely, the lower the VIX, the more bearish the market is. Consequently, Charts 29.19 and 29.20 plot this indicator inversely so that its fluctuations generally match that of the market in terms of direction. Chart 29.19 indicates one use of the VIX with the aid of positive and negative divergences as the human emotions of greed and fear appear to lead prices.
Technical Analysis Explained Page 43