Sunday, November 8, 2009

Chapter 3 Price forecasting. USA, LLC

ArabicChinese (Simplified)Chinese (Traditional)DeutchEspanolFrenchItalianJapaneseKoreanPortugueseRussian

The truth is simply this: no one can accurately forecast any market. An old bit of wisdom says, "He who forecasts the future, lies. Even when he tells the truth." No one knows with certainty what the future holds, so any forecast is a lie; and even if a future event were correctly forecast, the forecaster didn't know positively that the event would occur, thus still a lie. Lie might seem a harsh word--no harm or any malice is intended. And yet, all forecasters know conditions change as soon as they make their forecasts. Thus, forecasters always tell you something they know will not happen exactly as they say it. A wise trader understands this and proceeds into the world of price forecasting as if it were a mine field. The object of getting across a minefield is to do so with out getting blown up. Not so coincidently, the expression used by traders to signify that they have lost more than they can stand is to say that they "had a blow up" or merely "blew up." In reality, traders can blow up for many reasons, but betting on a price direction that was wrong is a common mistake. What to do then? If a market price cannot be correctly forecasted, why do people constantly try? The short answer is that being on the right side of a price move has value. The quest for money causes people to forecast prices regardless of whether or not they are accurate.

This chapter is not designed to be a primer on how to forecast prices, but rather a discussion on the various ways that prices are forecasted so that risk managers can make more informed decisions on which risk management tools are more appropriate for their individual needs. Additionally, it is important for all risk managers to develop their own philosophy concerning how markets behave.

The Two Big Mine Fields

Traders believe that prices can be forecast or they cannot. Those who maintain that prices cannot be forecast believe in the Efficient Market Hypothesis (EMH) and those who believe otherwise are proponents of the Inefficient Market Hypothesis (IMH) (also called the Deterministic Hypothesis). The EMH says that the mines in the field have been randomly distributed and that out of all the traders who enter the field, a certain number on average will emerge and all of the others will blow up. No individual trader can possibly figure a way through. IMH declares that while an individual mine's location may be unknown, patterns and certain causes and effects can be known and a route through the field can be drawn with fewer blowups. Mines, for example, may have been placed in greater concentrations in the middle and thus the edges of the field are safer. Or left-handed people set mines and therefore the pattern can be skewed to the left. IMH maintains that with enough study and information, a better way can be found through the minefield.

Efficient Market Hypothesis

Louis Bachelier stated in 1900, "The mathematical expectation of the speculator is zero." (1) Bachelier's idea was termed the Random Walk Hypothesis (RWH) from an earlier discussion among scholars. The question was posed, "If you leave a drunk friend in a garden that is enclosed with a locked gate, when you arrive the next day to pick him up, where will the most likely spot be?" The answer early scholars arrived at was in the spot where you left him. If he was drunk then each step he took would be random and thus the expected result of his steps over some time would sum to zero and thus the most likely mathematical spot to find him is where he was left, ergo, the name Random Walk Hypothesis. RWH evolved in the 1960s to the EMH when Bachelier's earlier work was rediscovered and new research added. An efficient market has large numbers of traders who use all available information and all future expected information to formulate price. Since price has all known and knowable information embedded in it, including all random news as it occurs, it will be unrelated to any other price. Over the years, the EMH has been codified into three major forms:

* Weak Form--All past information is reflected in price discovery.

* Semi-Strong Form--All past information as well as all current known information is used to formulate prices.

* Strong Form--All past and current information plus all knowable information is considered in the pricing process.

The price forecasting literature has been filled with countless articles and research endeavors that support the EMH since the 1970s. The three forms have evolved to reflect the various beliefs among traders on the validity of the EMH. Weak Form believers think the markets are generally efficient, but not all the available information is fully incorporated into the pricing process. The Semi-Strong Form states that all useful information, past and present, has been used and only insider information could change the market price. The Strong Form advocates state that even insider information has somehow been embedded in the market price.

Inefficient Market Hypothesis

IMH is the theory that market prices are not determined with perfect information and instead are constantly evolving as more information becomes available and is used by traders. IMH champions believe that with knowledge and skill some measure of price forecasting can be valuable as market inefficiencies are discovered and acted upon. How inefficiencies are discovered fosters almost as many fierce arguments among believers as between the IMH devotees versus the EMH supporters. IMH adherents are broadly classified as either technical or fundamental analysts.

Technical analysis is a method of analysis based on the belief that where the market has been in the past is, in some way, the best indicator of where it will be going in the future. Technical analysts dismiss the Weak Form of the EMH. Technical analysis is divided into two categories: charting and mathematical modeling.

Fundamental analysis holds that price determination has a cause-and-effect relationship and once the cause is properly identified, the effects can be forecast with some degree of accuracy. Fundamental analysis uses economic data and relationships, knowledge about events and circumstances, and any other data or causal connections to ascertain price. Fundamental analysis rebuffs the Semi-Strong Form of the EMH.

Perhaps another form of IMH followers should be identified--insider trades. The Strong Form of the EMH states that not even insider information alters the random nature of price discovery because it has already been anticipated. IMH disciples dismiss the Strong Form of the EMH as rubbish. Insider information does move markets, the IMH folks argue, and is unknowable in advance so cannot possibly be incorporated in the price. EMH proponents counter that even though insider information actions may be unknowable, the effect on the market price over some time period is insignificant. Since insider information cannot be used legally to trade, this aspect of price forecasting will be left untouched.

Using the Efficient Market Hypothesis

If markets are efficient and therefore cannot be forecast, of what use is the theory? Certainly the EMH is not used to forecast prices, but if markets move efficiently then certain other aspects of the market price movement might have value to risk managers.

The broadest use of the EMH by traders occurs in the equity markets. EMH believers do not believe that active mutual fund managers are any better than a simple index of various stocks such as the Standard and Poor's Index of 500 stocks. An equivalent idea in agricultural markets is corn price in Iowa. The EMH model asserts that the price of corn in Des Moines is efficient and cannot be forecast with any accuracy. However, that does not mean that certain characteristics of the price cannot be helpful to a trader in the short run.

Next Day Pricing--What is the most likely price of corn tomorrow? The best guess is not a random number pulled from all possible numbers, but rather what the price is at the end of the previous day. Where the price was yesterday is a better guess for where it will be today than a simple random guess. We suspect today's prices will not be the same as yesterday's and yet we do not think they will be vastly different. If Des Moines corn price yesterday was $3 per bushel, the best guess for today is $3 per bushel rather than a guess that doesn't acknowledge or relate to the general level of prices in the recent past. On the other hand, a guess of today's price based on yesterday's price will be no more accurate according to the EMH than a guess of say $3.20 per bushel or $2.80 per bushel if the forecasts are based on the relative level of yesterday's price.

Short-Run Minimum/Maximum Prices--Where the market has made a new high or low in the short run is a better guide for the short-run maximum and minimum price forecasts than a simple random guess that ignores general price levels. For example, if, during the last three months, Des Moines corn price reached a high of $3.30 per bushel and a low of $2.75 per bushel, those two prices would be a better guide for potential highs and lows for the immediate futures than a simply random guess. The EMH says that this would, however, not be true for a longer period of time.

EMH believers will not try to second-guess where the market price is headed, but rather will use past short-run price movements only as a guide for general price level expectations. If, for example, a manager needed to value a bin of corn for inventory purposes that will be sold in two weeks, what today's market price is or last week's average is a better guide for valuing the corn now than trying to forecast what corn prices will be in two weeks, because accurate forecasting is impossible and thus the general level of prices now is the best price to use.

Using the Inefficient Market Hypothesis

Obviously the idea that prices cannot be forecast with any accuracy doesn't appeal to a lot of traders. The EMH appeals to a lot of academic researchers but has little appeal to the general trading population. Why else would they be trading? Traders are always looking for an edge and they use many techniques to forecast price direction.

Fundamental Price Forecasting

One of the major bodies of economic theory is concerned with supply and demand and the interaction between the two to determine price. Alfred Marshall first proposed the idea in the early 1900s that equilibrium price was the result of the scissor-like connection between a supply curve and a demand curve as shown in Figure 3-1.


Modern microeconomic theory says that an individual producer's supply response curve will be the upward sloping portion of his marginal cost curve and the market supply curve will be the horizontal formation of all individual marginal cost curves. Marginal cost is the cost of the next unit of production as shown in Figure 3-2. An individual producer will produce at a point where the price received is equal to the last unit produced, implying that all previous units had a cost that was lower than the price. To induce a producer to produce more, the price must increase, thus the marginal cost curve is also the individual's supply response curve. This wheat producer would produce 40,000 bushels at a price of $3 per bushel, but if the price were $3.50 per bushel he would produce 55,000 bushels.

Figure 3-3 shows the hypothetical supply curve for wheat, which would be the aggregation of the wheat producers in the United States. The market supply curve traces the quantity that will be supplied as the price of wheat changes. Thus price changes will cause a change in the quantity supplied.

How much the quantity supplied will change due to a change in price is called the price elasticity of supply. If the price changes a great deal and the quantity supplied doesn't change much, the supply curve is said to be inelastic. If the price changes a little and the quantity supplied changes a lot, the curve is said to be elastic. Price elasticity of supply is equal to the percent change in the quantity due to a percent change in the price or over some range from [Q.sub.1] to [Q.sub.2]. The formula is

E = Percent change in quantity supplied/Percent change in price = [Q.sub.2] - [Q.sub.1]/([Q.sub.1] + [Q.sub.2])/2 / [P.sub.2] - [P.sub.1]/(P.sub.1] + [P.sub.2])/2

If a price change of 1 percent causes more (less) than a 1 percent change in quantity supplied, then the curve is said to be elastic (inelastic). How responsive a quantity supplied change is due to a change in price is primarily a function of how fixed or inflexible the resources or production processes are relative to a time period. The biological process to produce most grain crops is fixed. If the price of wheat goes up a great deal in December, wheat producers cannot produce more wheat and get it to the market simply because the biological process of planting and growing wheat is determined by the season of growth. However, if the price of cheese goes up, cheese plants could respond more quickly by putting on extra shifts or adding more equipment. The wheat supply curve would be relatively inelastic (unresponsive to price change) relative to the cheese supply curve, as illustrated in Figure 3-4.

A market supply curve exists at a given point in time at which production technology is relatively fixed and the price of inputs used to produce the product is fixed. If either of these items changes, then a whole new supply curve will exist. It is said that supply will change when production technology or input prices change. If a new production technology becomes available that will be more productive then the supply curve will shift to the right. If a pesticide was banned that reduced the productivity of the wheat farmer, then the supply curve would shift left. Likewise for a change in the price of inputs. If fertilizer prices increase, then the supply curve will shift left. Figure 3-5 shows some hypothetical supply curve changes. In Figure 3-5, S1 represents the original curve while S0 shows a decrease and S2 an increase.

Major Points Concerning Supply

A change in the price of a product will cause a change in the quantity supplied, and how responsive that change is will determine the elasticity of supply. Changes in production technology or input prices will result in an entirely new supply curve that is either to the left or right of the old one. Also, since the market supply curve is a sum of all individual supply curves, changes in the number of producers will shift the market supply curve left or right. Fundamental price forecasters will concentrate on watching for changes in production technology, changes in prices of major inputs in the production process, and changes in the number of producers. They will also attempt to know the price elasticity of the supply curve so they can estimate the change in quantity supplied when a price change occurs.


Demand curves are derived from consumers' utility (value/use) of a product. The idea is called Diminishing Marginal Utility. One unit of bread has a certain value or use to a consumer. However, a second unit of bread will in most cases result in somewhat less value or use and a third even less. The theory is that to induce a consumer to take another unit of a product that clearly has less value than the previous one, the price of the product must be lowered. Consequently, an individual's demand curve slopes downward to the right. The market demand curve is the sum of all individual demand curves and slopes downward to the right just like individual demand curves as shown in Figure 3-6.

Price elasticity of demand describes how responsive quantity changes are to changes in price. Price elasticity of demand is equal to the percent change in quantity due to a present change in price, or over some range of [Q.sub.1] to [Q.sub.2]. The formula is

[E.sub.D] = Percent change in quantity/Percent change in price = [Q.sub.2] - [Q.sub.1]/([Q.sub.1] + [Q.sub.2])/2/[P.sub.2 - [P.sub.1]/([P.sub.1] + [P.sub.2])/2

If price changes 1 percent and quantity demanded changes by more than 1 percent, the product is said to be elastic and if the quantity demanded changes by less than 1 percent it is inelastic. The degree of elasticity of a product is roughly determined by the amount and availability of substitutes for the product. Items that have many substitutes tend to have an elastic demand because if the price changes, consumers can readily substitute other items. Products that have few, if any, substitutes have inelastic demand curves, whereas if a major price change occurs, consumers cannot adjust their consumption very much because of the lack of choices. Price elasticity of demand is fickle. What is a substitute to one consumer is not to another. To a consumer who is rich or simply brand loyal, a certain type of luxury car may have few if any substitutes and therefore have a fairly inelastic demand curve that otherwise would be elastic for another group of consumers.

The individual's demand curve is derived by holding income, tastes and preferences, and the price of other goods constant. When the sum of all individual demands creates the market demand, then the population of consumers is likewise fixed. If income, population, tastes and preferences, or the price of other goods change, then the market demands will shift--left for a decrease and right for an increase. If new health information reveals that red wine improves health, then peoples' tastes and preferences concerning red wine might change and shift the demand right, ergo an increase in the demand for red wine as illustrated in Figure 3-7.

Major Points Concerning Demand

Price changes cause a change in quantity demanded and the amount and availability of substitutes will determine how responsive the change in quantity demanded will be. A whole new market demand curve will exist if population, income, tastes and preferences, or the price of other goods change. Fundamental price forecasters will endeavor to know how price elastic the demand curve is for a product so they can more correctly estimate changes in quantity demanded when price changes. They will also estimate changes in income, population growth, tastes and preferences, and the price of other goods. For example, if the price of pork changes, then the demand for beef will be affected since beef and pork are substitutes for many (but not all) consumers.

The relationship between the price change of one commodity and what it does to the quantity demanded of another product is measured by cross price elasticity of demand. The formula is

[E.sub.x] = Percent change in quantity of A/Percent change in price of B = [Q.sub.2A] - [Q.sub.1A]/([Q.sub.1A] + [Q.sub.2A])/2 / [P.sub.2B] - [P.sub.1B]/([P.sub.1B] + [P.sub.2B])/2

If a 1 percent increase in the price of one product induces a positive change in the quantity demanded of another product, the two products are said to be substitutes (such as beef and pork). If, on the other hand, a 1 percent increase in the price of one product causes a negative change in the quantity demanded of another product, the two products are said to be complements (such as beer and pretzels).

Additionally, as consumers get older, their life-long consumption patterns don't change very much. That is, if someone grew up consuming bacon and eggs for breakfast, they will likely continue that trend over time. However, the consumption of bacon and eggs might slowly change due to health or weight concerns as the consumer ages. Demographic and cohort analysis (a similar group of people) are important fields of study for demand analysis, especially over time.

Putting Supply and Demand Together Perfect Market Model

A perfectly competitive market is defined as a marketplace with many buyers and sellers who are not large enough to have any undue influence, vying for a homogenous product. Not too many markets meet the criteria of a perfect market. However, many markets are said to be workably competitive. A market may have many buyers, for example, but only a handful of sellers, or vice versa. But if the small number of either buyers or sellers can't exert any type of monopoly power, then the results of the not-so-perfect market are similar to a perfect market. A perfect market is one where there is no excess economic profit. Prices reflect costs including a return that keeps the resources employed in that use (i.e., opportunity costs), but no more.

Consider, for example, that the price of wheat in Chicago is $4 per bushel but the price for the same kind of wheat in Dallas is $5 per bushel. The cost of transportation between Chicago and Dallas, or vice versa, for a bushel of wheat is 75 cents per bushel. The perfect market model says that the price difference between the two markets should reflect the cost of transportation, but no more. The current difference between the two markets is $1 per bushel, yet the cost of transportation is only $0.75 per bushel. A rational trader would buy the wheat in Chicago for $4 per bushel, ship it to Dallas and sell it for $5 per bushel, and earn an excess economic profit of 25 cents per bushel. Traders who do this are called arbitragers. Arbitrage is the process of capturing excess economic profits between two or more markets. The actions of arbitragers will cause the excess economic profits to vanish. In the example between the price of wheat in Chicago and Dallas, there will be several traders who will try to capitalize on the excess profits. Arbitragers will enter the market in Chicago (an increase in the number of market participants, thus a shift right of the demand curve for wheat) thus bidding up the price of wheat in Chicago. Likewise, arbitragers will be new market suppliers in Dallas at $4.75 per bushel or above (an increase in the population and thus a shift right in the supply curve) in Dallas, thus bidding the price down. Figure 3-8 shows these effects with supply and demand diagrams. Markets that are separated by space have spatial price differences. Perfect spatial markets differ by the cost of transportation, no more or less, as spatial arbitragers will bid away any excess economic profits.

Markets differ also by time and product form. Time different markets are temporal markets and should differ in price by the cost of storage (carry). Products that undergo processing from the raw form to a finished good should differ in price by the cost of processing and are said to be form markets. Thus the three types of market models are spatial, temporal, and form. Using the perfect (or workable) competition model, arbitrage traders will access what should be the price difference between each of these markets (transportation costs, storage costs, and processing costs) and determine if excess profits can be made. When they can exploit the differences, they will do so and drive the markets back to normal. Profit taking by arbitragers helps maintain competitive markets.

A futures contract that has a delivery date of December has a price relationship with the current cash price, say in October. If we take away the spatial and form differences and have the product only differ by time, what should be the relationship between the cash price of corn in Chicago in October and the futures price for delivery in December? The perfect market model says that the temporal price differences should be only the cost of storage from October to December (two months). If the price of cash corn in October is $3 per bushel and it costs $0.10 to store it until December, the December futures price should be $3.10 per bushel. If the futures price was $3.15 per bushel and the cash price was $3 per bushel, arbitragers would enter the market. They would sell the December futures contracts (promising to deliver in December) for $3.15 per bushel and simultaneously buy local cash corn for $3 per bushel. They would store the cash corn for two months incurring a cost of carry of 10 cents per bushel and earn 5 cents per bushel excess economic profit. Figure 3-9 illustrates how the action increases the demand for cash corn in October, thus increasing the price while the supply of December delivery corn increases, thus lowering the price in December. Actions of temporal arbitragers keep the futures and cash markets tightly linked via the cost of carry.

Soybean markets are linked to the finished markets for soybean oil and meal via the cost of processing the raw soybeans into the finished forms. The price difference between the raw product market and the finished product market should be the cost of processing. Form arbitragers will enter the market when there are profit opportunities.

The perfect market model provides a theoretical framework to analyze if markets are inefficient. If spatial, temporal, and form markets differ by the cost of transportation, storage or processing, no more, no less they are said to be perfect markets and no excess economic profit exists. If markets have potential profits in them because the price differences between the markets are larger than the costs of transportation, storage, or processing, then they are said to be imperfect markets and arbitragers will exploit the excess profits away. If market price differences are less than the cost of transportation, storage or processing whereby a potential loss exists, they are said to be not perfect markets. Not perfect markets function slightly differently than imperfect markets. In the previous temporal example in Figure 3-9, arbitragers simultaneously had positions in both the futures and cash markets to capture the excess profit. If, however, the December futures price was $3.07 and the cash price was $3 with cost of storage for two months at $0.10, an action of buying the cash and selling the futures would yield a loss of three cents per bushel. What might happen is traders who needed corn would buy the December futures delivery corn for $3.07 today. If they bought the cash corn today and stored it until December, the cost would be $3.10, thus it would be cheaper to buy the futures contract and wait for delivery in December. Figure 3-10 shows this action by increasing the demand for December corn futures, thus increasing the price and driving the price difference back to the cost of storage of 10 cents per bushel.

Arbitragers will keep spatial, temporal, and form markets closely tied together such that the movement in one market will also impact a related market. Their actions cause derivative market prices and cash market prices to "tend to trend together." Consider, for example, October cash corn prices at $3 per bushel and December corn futures prices at $3.10 per bushel and the cost of storage at 10 cents per bushel from October to December. The two markets are said to be perfect. A major winter storm hits all of the Midwest so severely that cattle have no grass to eat and need supplemental grain to survive. The demand for cash corn will expand, pushing local cash prices up. What would the December corn futures price do? It should move in tandem with the cash price and differ by the ten cents storage cost, that is, trend together. The December corn futures contract derives its value from the cash market and the difference should be the cost of storage.

The Law of One Price

The perfect market model provides the framework for the theory that price differences should be able to be explained by the cost of transportation, storage, or processing--thus only one price plus or minus expected costs. Some refer to the perfect market model as the law of one price. If the difference between two or more prices can be justified by cost, then the two prices are identical except for defensible costs and are said to be one price. The law of one price is also used to explain the differences between identical products that are separated by different currencies. The price of copper in the United States should be the same price as the same grade of copper in another country adjusted for currency differences. If not, currency arbitragers will enter the market. Acting separately in the same market, currency arbitragers and copper arbitragers will force price equilibrium between grades and between currencies. The Economist regularly publishes its Big Mac Index that is the cost of a Big Mac hamburger at McDonald's in several countries adjusted for currency differences. To be sure, the price differences are not always reflective of currency differences. But it does provide a reference point to look at other supply and demand factors that might cause a price difference other than currency.

The law of one price is simply another way of looking at the perfect market model. It provides a starting point for arbitragers as they attempt to extract any excess economic profits that may exist in imperfect markets. Arbitragers are the ones that keep markets separated by justified cost differences--transportation, storage, processing, or currency.

Artificial Price Floors

When market supply and demand curves are put together, the result is a market-determined price, or the classic scissor diagram. This model is useful for forecasters to use to look at the effects of circumstances or governmental action that would impose a floor or ceiling on price. In Figure 3-11, a governmental program decrees an artificial price floor. At the nonmarket-determined price, more producers are willing to produce the product than consumers are willing to buy at the artificial price and thus a surplus results. A surplus will cause stockpiles to increase and pressure for "black" markets to emerge where the product trades unofficially at a lower-than-decreed ceiling price. This regularly happens in many U.S. price-supported commodities such as sugar. The U.S. government supports a domestic price that is higher than the world price of sugar, thus there is a constant surplus of sugar on the U.S. market. Price forecasters have to be aware of various governmental price support programs and their likely impact on prices.

If an artificial price is set below market equilibrium, then a shortage will result as more consumers will want the product at the nonmarket lower price than suppliers are willing to provide to the market as shown in Figure 3-12. This occurs in cities that impose rent controls on apartments and in the last few years in the medical field when either governmental actions or actions of Health Management Organizations (HMOs) artificially fix a price that is below the market equilibrium price. It rarely occurs in agricultural commodities in the U.S. but does occur abroad such as in Mexico where corn prices were set very low due to pressure from consumers.

Price forecasters have to be vigilant in addressing the various nonmarket pricing schemes that exist in order to properly determine the likely impact on prices, not only for the domestic U.S. market but also for the international market as well if the commodity is traded globally.

Supply and Demand Driven Prices

Certain situations exist in a particular market that will result in most of the price movement being caused primarily by either changes in supply or demand, but rarely both. If most of a price movement is caused by changes in supply, the price movement is said to be supply driven. If the majority of the movement is from changes in demand, it is called demand driven price changes.

Supply Driven

Figure 3-13 shows a hypothetical situation for corn. Demand for corn is fairly stable as uses are well established and will change only by some small percentage as population changes or as new uses are developed. However, the supply curve will shift as growing conditions change causing the price of corn to be influenced more by the supply curve shifting than changes in demand. As the supply curve shifts from S to [S.sub.1] and [S.sub.2], the price of corn changes. Price forecasters would therefore concentrate on studying the factors that influence the supply of corn such as constant monitoring of weather and growing conditions as reasons for price changes.

Demand Driven

Fine wines are primarily demand driven in terms of price movements. People's tastes and preferences are influenced when they know that the vintage will be excellent. The supply of the wine is fixed at a certain volume and the entire price changes will occur on the demand side as the vintage's quality is assessed by wine consumers, as shown in Figure 3-14. The price of wine will change as the demand curve shifts from D to [D.sub.1]. Price forecasters will watch for reports from critics, restaurants, and other quality assessors to help determine how demand will change, as supply is fixed for a single vintage wine, and will gradually shift to the left as the vintage is consumed.

Seasonal and Cyclic Movements

Agricultural commodities have biological characteristics that impose a seasonal factor on the production process. Corn is planted in the spring and harvested in the fall in the Northern hemisphere and the opposite in the Southern hemisphere. Hard red winter wheat is planted in late summer or early fall and harvested in the late spring or early summer of the next year. The nature of seasons and biology also impact animal agriculture as well. Most ranchers calve in the late winter and early spring and then wean the calves in the fall. Each species of animals has different gestation periods and production characteristics and requirements. Therefore, each agricultural commodity will have a different seasonality associated with the production of the product. Seasonal price movements are price activities that occur within a calendar year or production period. Cyclic price movements are price tendencies that occur over several production periods or years.

It is important in price forecasting to understand whether or not a product has a seasonal price tendency or a price cycle. Almost all crops have a seasonal price pattern as shown in Figure 3-15. Some animals show cyclic price behavior as illustrated in Figure 3-16 for cattle.


Two major types of price forecasters use supply and demand analysis: "gut" analysts and econometric analysts. Gut analysts filter all of the various shifters of supply and demand through their brain and come up with a price estimate. It is impossible to teach how to be a gut forecaster. Each individual will learn a particular way and will process information differently. Gut forecasters use the basics of supply and demand and other standard economic models and then come up with a forecast based on experience, judgment, and intuition. Many gut forecasters are very good at what they do; no doubt many are very bad, but they don't last long. This is plainly one area of price forecasting that is difficult to measure and therefore is left to the reader. The other area, econometrics, while complicated and mathematically sophisticated, is easier to fathom than gut forecasting, if for no other reason than it can be explained in terms of tools to use that have cause-effect relationships.

Econometrics is the study of quantifying economic relationships. A combination of mathematics and statistics forms the core of the tools used. An economic relationship is expressed in a mathematical format and then a statistical tool is used to estimate the values of the mathematical model. The statistical tool most often used is multiple regression analysis.

The process involves the following:

1. Determine the economic relationship. Suppose the relationship between beef and pork is needed. The economic relationship is stated as: What does the demand for beef do when the price of pork changes? If beef and pork are substitutes, then as the price of pork goes up, the quantity demanded of pork will decrease (a movement along the demand curve for pork). Consumers will consume less pork, and because they consider beef to be a substitute for pork, they will shift some of their consumption to beef. This will cause the demand curve for beef to shift right, or increase. This is first-round effect. But, as the demand for beef increases, the price of beef will increase, assuming other factors are held constant. The rising price of beef might then cause a series of second-round effects with other substitutes (chicken or fish perhaps) and so on. At this point, forecasters must decide how complicated they want to make the economic model. Most forecasters stop with only the first-round effects. Other forecasters will simply ignore the potential effects of pork prices on beef prices because they consider them to be small and not worth the effort to quantify.

2. What is the mathematical expression of the economic model? If a forecaster wanted to determine the price of beef, she might start by creating a proposed mathematical expression as

[P.sub.b] = f ([Q.sub.b])


[P.sub.b] = price of beef

[Q.sub.b] = quantity of beef

f = Some as-yet undetermined mathematical functional form

Some as-yet undetermined mathematical functional form

This is a simple model that may or may not correctly specify the economic relationship between price of beef and the quantity of beef, but could be a good predictor of beef prices. Once the variables have been specified ([P.sub.b] and [Q.sub.b]), the mathematical equation must be specified, that is, what form does f take? Is it a simple linear relationship? Logarithmic? Quadratic? If it were assumed to be linear, then it would be expressed as

[P.sub.b] = a + [b.sub.1] [Q.sub.b]


a = intercept

[b.sub.1] = slope of [Q.sub.b] relative to [P.sub.b]

3. Determine what data to use and the time frame of analysis. How long of a time period is needed? Will the data used be daily, weekly, or monthly? Will the data be for a specific market or aggregated over several markets? If daily or weekly data are used, what allowance is made for incomplete data series (i.e., holidays when markets are not open)? If the data is aggregated, which markets are included and which excluded?

Econometrics, to be sure, is many times more complicated than this simple explanation. But no matter how complicated the process, three major areas are at the crux of each analysis: (1) What is the economic relationship? (2) How can it be expressed mathematically? and (3) What data are available and can be used? Some forecasters use very simple models while others use extremely complex ones. It is important to notice that even a simple model that expresses the price of beef as a function of the quantity of beef still requires a significant investment of time in development of the model, specifying the form, and collecting the data. And if the results are poor, refinements will have to be made in the model, mathematical form, or the data used. Needless to say, most econometric work is performed at universities or large consulting firms that have the expertise to put all of the elements together. Little wonder that most traders shun econometric analysis and concentrate on simple cause-effect relationships that they can more easily apply themselves without extensive analysis or outsourcing of the statistical work.

Technical Price Analysis

Technical analysis is based on the belief that where prices have been in the past can be used as a guide for the future direction. Technical analysis does not directly dispute fundamental analysis but plainly believes that all of the information is embedded in the price movement and that it is impossible to fully determine all of the factors that influence price. A technical analyst would say that fundamental analysis determines the reasons why market prices move and technical analysis studies the effects.

Two major types of technical analysis are used by the majority of technical traders--charting and mathematical modeling. Charting entails the physical recording of price in some visual form and then studying the past price profile. Mathematical modeling is sometimes called "form fitting" because it is the process of trying to find a mathematical formula that will mimic the past price movement.

Charting Analysis

The major idea with charting is to visualize price information. One of the first to use charts to determine price direction was Charles Dow in the late 1800s. Charles Dow used price charts to develop what was later dubbed the Dow Theory. The theory states that a major bull market (a rising market) will continue as long as the intermediate highs and lows continue higher as depicted in Figure 3-17. Likewise, a major bear market (a declining market) will continue as long as the intermediate highs and lows continue lower. Mr. Dow's theory, as well as many other tenets he proposed and used to analyze stock prices, is still widely used by technical analysts.

Chart Types

Single Price Charts A single price chart merely records a price for a designated time period. This price can be the average price for the time period, such as a day or week, or it can be any other price. Often the price used is the closing price for daily price charts, and these charts are referred to as closing price charts. There is no one right price to use in single price charts, as long as the same price is used consistently throughout the time period of the chart. Figure 3-18 demonstrates a single price chart.

Single price charts are the most common charts used when mathematical tools are used to analyze prices. They are also the most common charts used when mechanical trading systems are created using a computer.

Bar Charts Single price charts contain limited pricing data; a serious chartist prefers more information. Therefore, for serious chartists, the most common form of charting is a bar chart. A bar chart records the high, low, and settle (or close) as portrayed in Figure 3-19. The most widely used bar chart is a daily chart that records the high for the day, the low, and the settle (a weighted average of the last few minutes of the closing prices). Inter-day bar charts are also used by some traders whereby they record the high, low, and close for a designated time period such as from 10 A.M. to 11 A.M. Regardless of the time frame used, a bar chart will show the range of price movement on the vertical axis and time on the horizontal axis.

Point and Figure Charts The second most popular way to visualize prices is a point and figure chart. The point and figure chart records the price movement regardless of time. The vertical axis records the price level and the horizontal axis de facto records time, but does not do so for a specific time. Point and figure charts use "Os" to record falling prices and "Xs" to record rising prices as illustrated in Figure 3-20. Since the horizontal axis is not a specific date, point and figure charts reflect only price movements regardless of time while a bar chart reflects price movements as a function of time.

Point and figure charts must denote a box size to represent a price movement, such as 5 cents. If a price move occurs that is smaller than the box size, the movement is not recorded. Additionally, a reversal criterion (a change in direction) must be stated such as "two box sizes." For a price direction to be reversed and recorded, the reversal criteria must be met. Otherwise, the price movement will not be recorded. For a two-box reversal with each box representing 5 cents, a price reversal must be at least 10 cents in the opposite direction to be recorded as a directional change.

Candlesticks A minor charting tool called candlesticks has emerged during the last few years. A candlestick will record the basic information of a bar chart and also add the opening price and the closing price. A candlestick will show the close and open as a two-dimensional body and the high and low as lines called shadows, shown in Figure 3-21. If the close is higher than the open, the body will be white (green) and if the close is lower than the open the body will be black (red). Each candlestick represents a specific trading day, just like bar charts.

The purpose of each of the charting tools is merely to express visually what a price movement looks like in order to determine two major things: (1) how long a trend (direction) will continue, and (2) when a trend will reverse. Over the years an almost uncountable number of tools have emerged (enough to fill hundreds of books) to determine these two major price movements.

Trends and Turning Points

How long will a bull market continue? Is this bear market a short-term trend? These are questions that chartists attempt to answer. The tool most often used is a trend line as indicated in Figure 3-22. Trend lines have some of the concept of the Dow Theory in them. A bull trend will continue as long as the intermediate lows keep getting higher and thus do not break the trend line and may ultimately add another point to the trend line.

Other tools are also used to indicate that a trend will continue such as flags and pennant formations, triangles, and runaway gaps. All three of these concepts indicate to a chartist that a trend is approximately half way completed. Figure 3-23 shows a flag formation.

A reversal in the trend is beaconed when the trend line is broken by a price movement as displayed in Figure 3-24. Other popular reversing formations include exhaustion gaps and double tops and bottoms. Figure 3-25 shows an exhaustion gap.

The list of possible formations is almost limitless and no attempt is made in this chapter to cover more than a few of the most widely used tools so that the reader can get an idea of some of the tools and how they are used by a chartist.

Mathematical Modeling

Because charting analysts most often attempt to answer the questions of trend duration and turning points in price moves, another form of technical analyst has emerged, especially as computer speed and capacity has increased in the last few years. Mathematical modelers or mechanical analysts have become very popular as they attempt to forecast prices based on previous price information. Mechanical analysis falls into three major categories: curve fitting, moving averages, and oscillators.

Curve Fitting

Curve fitting is the general expression used to imply that for a given set of past price movements, an equation will be selected that fits the data the best. The selection of an equation can be done visually or by statistical estimates, usually with multiple regression. Figure 3-26 shows a simple linear model's forecast of the next price. The example is a simple linear regression model that estimates the best fit of the actual data to an estimated line.

More complex mathematical models have evolved over the years. In fact, in the last several years advances in chaos and complexity theory have created a following of curve fitters who use the relatively new mathematical tools to fit economic data in an attempt to be able to forecast market price.

Moving Averages

By far the most used mathematical tool for price forecasting is the rather simple concept known as a moving average. The idea is to calculate at least two averages of past prices, one a short-term average such as the last three days and a longer average such as ten days. When the two averages cross each other, a change in the trend is indicated. Three averages are quite popular whereby the short and intermediate average crossing implies a watch signal for a possible trend change; when the intermediate average crosses the longer average, a trend is deemed over. The very popular 3-, 9-, and 18-day averages are shown in Figure 3-27.

Moving averages are extremely popular because they are very easy to calculate and use with or without a computer. Some analysts use weights to change the averages according to some arbitrary criteria in an attempt to add accuracy.


An oscillator is an elementary arithmetic expression used to measure the rate of change of prices. The simplest one is called a momentum chart. The formula is

M = P - [P.sub.x]


M = momentum

P = current price

[P.sub.x] = price at time period x days ago

If M is negative then the current price is below the price at some previous time and if it is positive then the current price is above the previous price. These plus and minus differences are then plotted around a zero line, as illustrated in Figure 3-28.

Other popular oscillators are relative strength and stochastic indexes. However, the list of oscillators is long and varied. Oscillators are uncomplicated mathematical expressions used to measure some aspect of price change and thus, hopefully, be more useful to price forecasters.

A Final Word

At some point, all of us will make a price forecast. Sometimes it is as simple as whether or not to fill up the gas tank on the car now or wait a few days when we forecast (hope) that the price will be lower. It is critically important for businesspeople to seriously think about what they believe about price forecasting. If they believe it is impossible to forecast prices with any accuracy, then how they construct their marketing plans and capital outlay budgets will reflect their attitude toward the risk of price change. Conversely, if they believe that prices can be forecast with some accuracy, alternative price risk management strategies will need to be developed. How people feel about price forecasting is directly related to how they will handle the risk of price change.


1. If a trader believes in the Efficient Market Hypothesis, would he ever use Fundamental Price Analysis?

2. How can both technical and fundamental analysis be used by a trader?

3. What is the major difference between bar charts and point and figure charts?

4. What would be a good example of a demand driven product? Why?

5. What kind of product would have a very inelastic demand curve? Why?

(1) Bachelier, L. (1900). "Theorie de la Spiculabon" (The Theory of Speculation), Annales de l'Ecole Normale superiure.

Source Citation
"Chapter 3 Price forecasting." Risk Management for Agriculture. Delmar Learning, 2007. 29+. Agriculture Collection. Web. 8 Nov. 2009. .

Gale Document Number:A185431414

ArabicChinese (Simplified)Chinese (Traditional)DeutchEspanolFrenchItalianJapaneseKoreanPortugueseRussian

Personalized MY M&M'S® Candies

Obama on 60 Minutes DVD

Great Prices at


(Album / Profile)

Click here for the Best Buy Free Shipping Offers

Shop the Official Coca-Cola Store!

No comments: