Historically, the most important source of information about Orlando weather, particularly temperature and rainfall, has come from the U.S. Weather Service reporting station in Orlando. When Roll first wrote about orange juice futures in 1984, this station issued two reports daily. These reports contained forecasts covering the next thirty-six hours. Since that time technological advances have extended the time horizon out to as long as fourteen days, although short-term forecasts continue to be the most reliable. Forecasts are made for maximum daily temperature, minimum daily temperature, and precipitation.2
Commodity Price Volatility
The size of the orange crop depends on a number of weather factors: temperature, sunshine, amount of rain, and frequency of storms. In fact, weather is the single most important factor impacting the supply of oranges to the market for concentrate. Figure 20-1 illustrates the time path of the spot contract from October 19, 1995, to January 20,
Figure 20-1 Spot Price for Orange Juice Concentrate
The volatile time path of the spot price for orange juice concentrate. Where’s the news driving orange juice futures? 1999. The figure makes clear that the spot price for orange juice concentrate is quite volatile. Is the weather around Orlando as variable?
Case Study 1: A December 1997 Cold Front
To gain some insight into the relationship between Florida weather and the market price for orange juice concentrate, consider some examples. On the afternoon of Tuesday, December 2, the following forecast was issued for the current evening and subsequent four days, as shown in table 20-1.
Notice that on Tuesday, December 2, rain had been forecast to occur in and around Orlando beginning as showers the next day and intensifying on Thursday. Note also that temperatures were predicted to drop over the course of the next few days, dipping below 40°F on the weekend.
On Thursday, December 4, a cold front passed through Florida. On the same day the price of concentrate jumped. Below are descriptions of two events, one concerning weather, and the other regarding the movement of orange juice futures on December 4.
Weather 12/5/973
Intense Thunderstorms Drench Southern Half of Florida
Widespread moderate to heavy rain fell over the southern half of Florida Wednesday night.
A cold front is pushing southeastward through the state, and a moist, subtropical jet has put itself in the front’s path. Intense thunderstorms developed Thursday afternoon over the lower Florida Keys and moved northeastward to bring another round of heavy rain to the region.
Key West had a heavy thunderstorm on top of already almost an inch of rain Wednesday night. Other rain totals included Miami with 1.10 inches, Sarasota with 1.76 inches and Tampa with 2.03 inches.
The threat of a freeze is enough to drive the price of concentrate up very quickly. Consider the behavior of prices for the March 1998 contract on Thursday, December 4, 1997. Trading began at 10:15 EST, with the opening price at 84.4¢. Prices hovered just below 84¢ for most of the day. However, at 13:55 prices jumped above 85¢, closing just below 86¢. Figure 20-2 below depicts the time path of these price changes.
The Dow Jones News Service offers a daily feature titled “World Commodities Summary.” This feature highlights the most active commodities markets every trading day and describes the events that
Table 20-1 Five-Day Forecast for Orlando, Florida, Tuesday, December 2, 1997
Figure 20-2 Price Path of March 1998 Contract, December 4, 1997
An illustration of intraday volatility: What happened between 13:00 and 14:00 on December 4, 1997? affected prices. What do we learn from the December 4 “World Commodities Summary” about the events that moved orange juice prices that day?4 Absolutely nothing: Not a single mention is made of the cold front sweeping through the state that day. In fact, this article makes no mention of orange juice prices at all. In contrast, consider the following description of the events affecting soybean prices the same day.
Nearby CBOT soybeans tested support levels just under $7 a bushel at midday Thursday, and finally toppled during the closing minutes. The session almost mirrored Tuesday’s, when the contract suffered more than half its daily losses just before the closing bell rang. Although technical factors were the main reason for Thursday’s dip, demand worries also played a role.
The U.S. Department of Agriculture said Thursday that U.S. soybean sales for the week ended Nov. 27 totaled 361,700 metric tons—down 50% from the previous week and down 39% from the four-week average. “The sales were half of what they were the previous week,” Henderson of Stewart-Peterson said. “That shows a slowdown in demand.”
The contrast between the paucity of news and the volatility of orange juice futures is striking. And it’s not an isolated event. One interesting aspect of news about prices of orange juice futures is its relative nonexistence. On a typical trading day, there are about five hundred stories about commodities that are available through Dow Jones News Service. Of These, there may be one or two about orange juice, and they usually concern nothing more than opening and closing prices.
This paucity of news led Roll to ask some obvious questions. To what are traders in orange juice futures reacting? Is it information not being picked up by the news wires? Or are they just reacting to themselves? Are they simply reacting to each other’s trades? In his 1985 presidential address to the American Finance Association, the late Fischer Black suggested that many traders cannot distinguish between relevant information and irrelevant information. He called irrelevant information “noise,” and labeled those who trade on “noise” instead of “information” noise traders. Are orange juice traders noise traders? Is noise-trader sentiment the driving force behind the price of orange juice futures?
Looking at December 4, it is clear that despite the absence of a report about Florida weather in the commodities news stories, there was an event. So, a skeptical reader may be reluctant to condemn orange juice traders as noise traders.
Case Study 2: A Typical July
But now let’s look at a summer month such as July, rather than a winter month such as December. Consider table 20-2, which shows the weather forecast for Orlando that was issued on the evening of July 1, 1997.
This is a typical forecast for July, with high temperatures in the 90s and low temperatures in the 70s. There was no indication of a possible weather event that might damage the orange crop in this report. Moreover, the only information about orange juice futures that appeared in the many Dow Jones News Service stories about commodities5 from June 30 through July 7, 1997, concerned the opening and closing prices of orange juice contracts. Yet, consider the September 1997 futures. As can be seen in figure 20-3, prices were quite volatile throughout, even into the calm, “uneventful” summer months. The low price for July 3 was 76.45¢ cents, the high price was 77.5¢, and the closing price was 77.45¢. The difference between high and low, relative to the close, was 1.36 percent. That change is typical for the S&P 500, where the information flows are quite substantial on a daily basis. But for the September orange juice contract, in July?! Not to put too fine a point on it, but was
Table 20-2 Five-Day Forecast for Orlando, Florida, Tuesday, July 1, 1997
Figure 20-3 Price Path of September 1997 Orange Juice Contract
The volatile time path of prices for the September 1997 contract. Is the market any less volatile in summer months, when there is less chance of a Florida freeze? there some unreported hint of a summer freeze before September? To paraphrase a fast food television commercial from the 1980s: Where’s the news?!
Case Study 3: Genuine News
In case you are thinking that the news services simply do not report news that is genuinely germane to orange juice contracts, here are two examples that they do. The first example is a news story that highlights the two major events that impact the market for orange juice concentrate—Florida weather and events in Brazil.6 Despite the emphasis on fundamentals, note the references to sentiment as you read the following excerpts.
Many traders are still bullish because of Brazil’s projection last Friday of a drop in its 1998–99 orange crop to 325 million boxes from 380 million this season, Scoville said. “We’ve changed fundamentals around 100% because we’re looking at sharply lower production from Brazil the coming year,” he said. Gains in futures have been limited this year because USDA last October projected this season’s Florida orange crop at a record 254 million boxes, he said.
Here is yet one more example of genuine news. On February 10, 1999, the United States Department of Agriculture (USDA) issued a report that stunned the orange juice futures market.7 In its World Agricultural Supply and Demand Estimates, the USDA predicted that the 1998 U.S. orange crop would be 194 million boxes, above the expected range of 185 to 190 boxes. In addition, the report predicted a record yield of juice per box: 1.60 gallons, up from the 1.57 yield of 1997.
The price of the March contract fell by 7.95 cents on the report, closing at 92.25 cents.
Commodities analyst Jack Scoville,8 quoted earlier, made an interesting statement about the reaction to the USDA report: “They got them coming and going.” Coming and going—the hallmark of volatility.
Case Study 4: The View from the Trading Floor
How about the view of those who actually trade orange juice futures? In June 1998 I interviewed Frank Tesoriero, an orange juice futures trader who at the time was also president of the Citrus Associates, a subsidiary owned by the NYCE.9 I asked him for his view on what drives the price of orange juice concentrate. We spoke about both sentiment and fundamentals, and began with a specific instance.
On June 24, 1998, the price for the July contract opened at 105 cents and dropped to 103 within the first ten minutes. In response to my question about what had precipitated the drop, Tesoriero stated: “It went down on absolutely nothing.” He went on to explain that it may have been “locals trying to pick each other’s pockets,” that when it comes to market price, “locals will try to whip it around.”
As for Brazil, Tesoriero described the difficulty of obtaining reliable and timely information, because those in possession of the information engage in strategic release. Moreover, those releases may come at times during the day that are difficult to anticipate. He noted that when it comes to providing information nonstrategically, Brazil has no counterpart to the USDA.
Recall that figure 20-1 illustrates the volatility of the spot price for orange juice. Notice that the second half of 1998 was even more volatile than the first half, with the price fluctuating between 98 cents and 123 cents.
Summary
The price of orange juice concentrate appears to be excessively volatile relative to the underlying fundamentals. The major fundamentals that affect the market for orange juice concentrate are the weather conditions in Florida and the supply of oranges from Brazil. Major news events about either one typically cause large changes in the prices of orange juice concentrate futures. However, the fact that volatility occurs in the absence of news about altered fundamentals is evidence of heuristic-driven bias and market inefficiency.
Chapter 21 Excessive Speculation in Foreign Exchange Markets
Foreign exchange traders engage in excessive speculation. So, why do traders speculate so much? Is it because they trade on strongly held, but erroneous, opinions? The evidence suggests that this is the case, and that psychological issues figure prominently.
This chapter discusses the following:
• a case based on the Asian crisis of 1997 and 1998, when the financial press reported that exchange rates were determined, not by economic fundamentals, but by psychology
• the general findings about excessive speculation in foreign exchange markets
Case Study: The Asian Crisis That “Began” In 1997
The statement about psychology mentioned above was part of a front-page story that appeared in the August 24, 1998, issue of the Wall Street Journal. Below is an excerpt from the article.
Frightened investors and quick-moving speculators in markets as far apart and different as Brazil and Hong Kong, Canada and Russia, Japan and Venezuela are scurrying to exchange local currencies for the U.S. dollar. …
Hopes that the crisis, ignited by the July 1997 devaluation of the Thai baht, would soon burn itself out have been dashed. …
In theory, a currency’s value mirrors the fundamental strength of its underlying economy, relative to other economies. But in the current white-knuckled climate, traders and lenders are trying to guess
what every other trader thinks. While traders use the most modern communications, they act by fight-or-flee instincts. If they expect others are about to sell the Brazilian real for U.S. dollars, they try to sell their real first. It isn’t just economics at work now; it’s a psychology that at times borders on panic.1
The statements about “psychology that at times borders on panic,” “fight-or-flee instincts,” and “expecting others to sell” recall an issue that was discussed in the first chapter. Remember the pick-a-number game? Its attendant lesson for practitioners was to factor in the importance of others’ errors.
Here is a thumbnail sketch of the events leading up to this near panic. In thirty years, the Asian tigers—Hong Kong, Singapore, South Korea, and Taiwan—achieved what the United States had taken a century to accomplish. These countries managed to increase fourfold the standard of living of their citizens. From 1994 though the first half of 1997, Asia accounted for one half of global economic growth. In order to integrate their economies into the global market, these countries established fixed exchange-rate policies against the U.S. dollar.
During the first quarter of 1997, the forecast for the tigers, as well as for the would-be tigers (Thailand, Indonesia, Malaysia, and the Philippines) appeared promising. Then, in the second quarter, the possibility of some problem loans in Thailand led investors to speculate that the Thais would be unable to maintain the value of their currency, the baht, against the dollar. However, investors did not anticipate that the second half of 1997 would bring major chaos to Asian financial markets.
Take the case of Indonesia. At the beginning of July 1997, the Indonesian rupiah was trading at 2,450 to the dollar. Prior to that, the rupiah had been very steady. But by early January 1998, it had moved to 10,225—an 80 percent decline in value!
Heuristic-Driven Bias by Borrowers
The following excerpts from the December 30, 1997, issue of the Wall Street Journal describe the problem facing Indonesian borrowers.2 As you read, keep in mind that most predictions are made by extrapolating recent trends—what Werner De Bondt (1993) calls “betting on trends.” In addition, people tend to be overconfident in their predictions. This leads them to be surprised more frequently than they had anticipated. The first paragraph of the article provides the background.
Counting on continued growth and a predictable currency, Indonesian nonfinancial companies more than doubled their foreign borrowing from the end of 1995 to mid-1997, according to Standard & Poor’s. As of Sept. 30, according to President Suharto’s new special adviser on debt, Radius Prawiro, Indonesia’s total foreign debt was $133.3 billion, of which $65.6 billion was owed by private companies.
The next portion of the article discusses the way in which Indonesian borrowers speculated.
Indonesian borrowers faced a simple choice. They could borrow rupiah at 18% or 20% and not worry about the exchang
e rate. Or they could borrow dollars at 9% or 10% a year and then convert the proceeds to rupiah. A year later, they figured, it would take 4% or 5% more rupiah to buy the dollars needed to repay the loan, but the cost still was less. …
The central bank, Bank Indonesia, had trouble keeping track of all the short-term foreign borrowing Indonesians were doing, but it knew this borrowing was heavy. “The currency was on a steady course,” Mr. Soedradjad notes. “And we had all these banks from all over the world landing here and handing out the money.”
Speculation can go either way. How did this one go? The article continues:
It is now clear that Indonesian companies made a huge bet that went bad: Few executives took seriously the possibility that Bank Indonesia would let the rupiah fall sharply. Even fewer were willing to pay the cost of hedging, insurance that banks and investments houses sell to protect against currency collapse. In late July and early August, Goldman, Sachs & Co. surveyed 34 Indonesian chief financial officers. Two-thirds had more than 40% of their debt in foreign currencies; of those, half were completely unhedged, and most of the rest had well under half of their debt hedged. That proved a costly mistake.
Beyond Greed and Fear Page 36