14 Behavioral Techniques

There are some approaches to trading that are directly dependent on human behavior and caimot be represented by pure mathematics. When viewed in the short term most sjstems will be more representative of behavior than economic factors because, over a few hours of one day, the longterm influence of the trend is very small. The concepts presented in this chapter deal specifically with human reactions. Measuring the news covers an area that has always been important to a floor trader, yet been greatly overlooked by technicians, it offers great opportunity for research. Event trading, which studies maiket reaction to price shocks, has become increasingly important for both proactive and reactive traders; contrary opinion takes the form of a poll or consensus of opinions of traders and maiket publications It may help answer the question,-Whatis everyone else doing?", or at least "What are they thinking of doing?"

The principal woiks of Elliott and Gann are included with a review of the Fibonacci series and ratios, which form a significant part of their technique. The way in which traders respond to maiket moves, and the remarkable similarity that can be found in Nature, give serious underlying substance to these methods. Because not all of the assumptions upon which these sjstems are based can be quantified, they are substantiated by the performance of the sjstems themselves. Both are fascinating and open areas of creativity essential to broad sjstem development. They are grouped together with discussions of natural phenomena and the rapidly growing area of financial astrologj, all of which should leave your grey matter stimulated.

MEASURING THE NEWS

If you can keep your head when all about you are losing theirs, maybe you havent beard the news

Rudjard Kipling

The news is one of the greatest single elements affecting free trade. As a medium, it carries both fundamental and technical information about all markets, directly or indirectly, and is indispensable. If not objective, the news services could materially alter any opinion by the inclusion or omission of relevant information. The impact of news is so great that a speculator holding a maiket position according to a purely technical sjstem would do best not reading, listening to, watching, or in any sense being exposed to news that might be cause for deviating from the sjstem. In a study commissioned by the Wall Sfreet joumal, it was shown that 99°o of the financial analysts polled read the papei regularly, and 92" considered it the "most valuable" publication they read.

As an element of a trading program or as an indicator of its own. the news is invaluable, if we could measure the impact of unexperted news, the importance of the Wall Sfreet.Joumal or the wire service articles, the USDA crop reports and the CFTC positions

reports, and the anticipation of news, we would have a better view of the direction of the maikets. But first we must be aware of the complications of analyzing the news. There is the problem of objectively selecting which items are relevant and which sources are most important. The most difficult problem is quantifjing the newshow do you rank each item for measurement, and on what scale do you determine cumulative importance? Some news items are known to have a greater impact on the maiket-weather disasters, major trade agreements, key crop reports-but these must be ranked against themselves and other issues to produce a numerical sjstem of analysis. On a single day, an address bj-the President about foreign trade may be ranked as a +6, a meetmg of the Federal Reserve a +7, unemployment a +4, the NAFTA Agreement a +3, a continued lack of rain in the west as a +2, large grain stocks as a 4, and a key article in the Wall Sfreet journal on the improved Russian crude oil production as a 5, giving a net score of +4 to the news-interpreted as moderately bullish. Such a ranking scheme would be considered analytic in nature, because it fries to weigh relevance on a predd;ermined scale without knowledge of individual effects.

Klein and Prestbo attempted such a study by assigning values of 3, 2, and I to articles in the Wall Sfreet joumal of decreasing importance. Their interest was the stock maiket, and their work was straightforward and some of the conclusions general in nature. They showed a direct correlation between the relevant positive and negative news articles and the direction of the stock maiket. As it was scored over 6-week intervals before and after major tuming points in the Dowjones Industrial Averages, the news would stay about 70°o favoring the current maiket direction. Having eliminated the possibility of the maiket influencing the news, they could conclude that, in refrospect. the maiket reflerted the nature of the news.

The focus of news that influences commodity prices is different from that which moves the stock maiket.

because it is limited in scope and often specific to tlie particular physical maitet. News releases and economic reports are an implicit part of commodity pricing, newspaper releases become very important in the last phase of a long move, which is easily seen when weather is a primary factor, but also when trade agreements are being negotiated. For example, traders will watch the weather report beginning in June when there has been a shortage of rain. They will stay long while no rain is in sight and quickly sell if a thunderstorm develops over the Chicago Board of Trade building. They will also protect their positions going into the weekend, in the event of unexperted rain, then reset their net longs on Monday. The news items that affect prices the most are:

Release of economic statistics ("inflation, sales, balance of trade, unemployment)

Action by the Federal Reserve or other central bank

News relating to the availability of funds

Government reports on production and stocks

Unexperted news (a potential price shock such as an assassination)

Trade negotiations, agreements, and legislation

Weather

In-depth studies by the Wall Sfteet Joumal

Front-page news articles (or lead television news) on high prices, strikes, and other topics with a dirert economic impact

Maitet letters, research reports, and comments from accepted authorities and influential organizations

3 Atiiofueh an- -sis -f the Cjuniitment -f lefmtl can beEmudin XTter 16-It A-uldbe n-led tiiat -ther goverunient refits released ou the

Ranking and Measuring

The problem of ranking and assigning values to news items requires a knowledge of how others see the news Klein and Prestbo shidied this problem for the stock maitet and concluded that about 90°o of the Wall Sfreet Joumal readers perceived news in the same way (bullish, bearish, or neufral). The same relationship can be assumed for futures maitets. A reason for the uniform interpretation of the news articles is the publicized analysis. Within minutes of the release of an economic report or Fed adion, the wire services begin to quote independent and poll opinions of the meaning of the report, then transmit those interpretations over their news media to he relayed to most traders. The professional analysis is taten as corrert, and later discussions based on that interpretation serve to solidifi the opinion.

News can also be measured empirically, by shidjing the immediate impact of an experted or surprise news item. For statistical reports, such as the Producer Price Index, care must be taken to use the corred figures. Maitet readion is a combination of expedations compared with adual figures, and corredions in the previous months data, in some months it is difficult to know whether the jump in prices was due to a downward revision of the previous month, or current values that are lower than expectations. When testing these fadors, the historic data must include the adual numbers released in the reports, and a separate value for the revision of the previous month. Indusfrj expedations are also important, but they can be determined empirically from the maitets readion. Therefore, following the release of an economic report, we can exped the price change to he expressed as:

Price change =fl a( current value - expedations) + b(revision -previous value))

where a and b are weighting fadors, a - b implying that the current value is more important than the previous one. It is necessarj to make the assumption for this tjpe of measurement that the effects of a news release are most important in the short term, and that their influence on the maitet is diluted daily. A starting point for representing the way in which the news decreases in value is to modify the standard physics formula:

where I is the net influence and T is the elapsed time since the release of the news. In science this relationship, which represents the physical impact declining with the square of the distance (or time), applies directly to sound and light.

Measured empirically, the impact of economic reports is worth studjing. For futures markets, the CFFC releases its Commitments of Traders report each week (upgraded from monthly reports), it tells the distribution of holdings among laie and small speculators and hedgers as a percent of total open interest. This report is watched with the idea that the small speculators are usually wrong and the large hedgers right (this is covered in a later section of this chapter).

It is important to understand the difference between the analytic and empirical approach to news. In the analytic method, the value of specific events is determined in advance, and when they occur their preassigned value is compared with the effects. Using the enpirical method, the historic effects of each event are tabulated and measured, then applied to subsequent news items.

The analytic spproach has the advantage of woridng in an environment where multiple events are occurring simultaneously, and only the sum of all news items can be calculated. When testing the empirical spproach, the event of interest may not be clearly distinguished from other news of importance occurring at the same time. Finding situations that isolate a

single tjpe of event, from which you can measure its impact, may be impossible. The primary disadvantage of the analytic approach is that it does not account for the discounting or anticipation of the news. An event of modest importance may become neufral or more significant relative to other concurrent or anticipated events; the empirical method would not be subject to that problem since it measures reactions and not expectations.

Trading on the News

Even without a sophisticated method of measurement, there are many professional speculators who trade on the news. When a bullish news item or report is introduced and the maiket fails to respond upward, the experienced trader looks for a place to sell, it shows that expectations exceeded reality and prices had alreadj anticipated the bullish interpretation of the news; there may be a large number of sellers above the maiket. Similarly, opening calls, available for most maikets, are transmitted via wire services beginning about a half hour before the opening bell. Regardless of the means for dd;ermining the opening direction, an experienced trader may take advantage of a higher opening call to place a sell order. There are frequent cases of so many traders wanting to sell a higher opening that the infiux of orders after a call, before the opening, has changed the direction from higher to sha ly lower on the open.

Agricultural weather markets function purely on news. Traders with long positions wait for the 5day forecast hoping for no rain., they anticipate a loss of a specific number of bushels per acre for every drj day once the rainfall is below a given level. Weather maikets are nervous and are characterized by evening-up on weekends., they rely heavily on anticipation and emotion. It is said that a farmer loses his crop three times each year: once for drought, once for disease, and once for frost. In 1976, the news carried numerous articles on the desperate wheat crop in the westem states, showing films of virtually barren fields, and yet the United States harvested one of their largest wheat crops on record. Weather maikets have a history of volatile anticipation but are a riskj way of making long-term price predictions.

The discounting of news is as important as the news itself. An old saying in the maiket, "Buy the rumor, sell the fact," implies that anticipation drives the price past the point where it would realistically adjust to news. When the actual figures are released, there is invariably an adjustment bad; to their proper level. The pattem of anticipation foi each economic report or news event should be watched closely.

Maricet Selectivity

The maiket seems to focus on one news item at a time. Although the same factors are always there to affect prices, they must reach a point of newsworthiness before they become the primarj driving force. For heating oil, the combination of unexpected, sustained cold weather compared with available inventories will activate a weather maiket Interest rates, the U.S. dollar, and tension in the Middle East will either be magnified out of proportion or completely ignored, while the anticipation of demand rises sha ly. A maiket analysis of shortages not sipported by news may as