The final macro component we will discuss is the prevailing attitude of traders and investors toward assuming risk. There are many methods for estimating this factor, and a wise approach is to consult several inputs and create a composite barometer or index.

When public risk-tolerance is high in the stock realm, you will likely see solid interest in IPOs, highly speculative stocks, etc. The same dynamic exists in currencies. For example, the NZD, CAD and Australian dollar (AUD) are considered “risk-seeking” currencies. Given they are “reflation plays” –– a trade that is banking on economic growth developing –– and sensitive to commodity prices, they are usually more sensitive to changes in the global economy.

Such currencies will typically have an inverse relationship to investor risk tolerance. It shows the Risk Aversion Thermometer, a trader risk gauge created by RBC Capital Markets, illustrates the rise and fall of risk tolerance. It shows the NZD/USD rate embarking on a multi-month uptrend around the same time the Risk Aversion Thermometer peaked. It shows the same pattern in the USD/CAD, where the downtrend actually reflects Canadian dollar gains vs. the U.S. dollar.

The relationships described here are by no means static. However, seeking out simple associations such as these can identify favorable trade conditions, as well as avoid trades that might look good from a technical perspective, but have potential weaknesses on the macro level.

Commodities are the driving force of growing economies. China has emerged as a major player in worldwide manufacturing and has become a large consumer of commodities such as oil and grains. Surprisingly, however, China does not produce much in the way of agriculture, so commodity based economies that stand to export heavily to China, such as Australia, New Zealand and Canada, will benefit. It shows how the U.S. dollar/Canadian dollar (USD/CAD) rate is inversely related to commodity prices, meaning the CAD gets stronger relative to the USD as commodity prices rise. Below the price chart is a list of several statistical correlations between the two series over different time periods (five days, 10 days, one month, etc.). Negative cor relation figures reflect data series that move in opposite directions, with 1.00 representing a perfectly inverse relationship (one goes up while the other goes down proportionally).

Positive correlation figures reflect the opposite, with +1.00 representing two series moving in tandem.

Another notable historical relationship is the impact of rising oil prices on Japan, represented by the Nikkei stock index and the Japanese yen (JPY).

Higher oil prices have a negative impact on global economic growth as a whole, and Japan specifically because it is highly dependent on imported oil.

With oil prices rising significantly over the past year, the impact on both Japan and the global economy has been noticeable. And when global economic growth is poised to slow, in this case because of higher oil prices, Japan typically suffers.

Given a large percentage of Japan’s capital inflows are the result of equity purchases, the link between higher oil prices and a weaker yen and Nikkei is clear. If global economic growth is waning, Japan will be negatively impacted because they are a large exporter. Hence, equity inflows into Japan will decrease when growth slows or is expected to slow — and rising oil prices have been an impediment to global growth. It shows Brent crude oil and the euro/Japanese yen (EUR/JPY) rate, along with various correlation figures.

An interest rate differential is the spread between the prevailing interest rates in two countries. All else being equal, capital flows to countries offering higher interest rates, which benefits those currencies.

In a time with historically low interest rate yields worldwide, traders will often accept greater risk to seek out those markets offering the highest yield. In the U.S., the Federal Funds rate is at 2 percent after the Federal Reserve raised the rate .25 points at its Nov. 9 meeting, while New Zealand has an equivalent rate of 6.25 percent (the next meeting is this month, with no hike expected), making the New Zealand dollar (NZD) a more attractive currency to hold than the U.S. dollar (USD) in terms of yield.

However, the NZD’s yield does not represent a risk-free rate. If the NZD depreciated vs. the USD during the time of your investment in New Zealand, the rate differential advantage could be wiped out. Hence, gaining an extra 4.25 percent (the difference between the two rates) per annum while also enjoying the potential capital appreciation on the long NZD position itself has been a logical move for many traders.

They show GBP/JPY rose toward the middle of the week, but they don’t indicate whether Wednesday’s unique rollover affects the pair’s price moves surrounding this event. To find out if this weekly occurrence has influenced GBP/JPY prices, we studied hourly price data from Jan. 1, 2003 to Oct. 31, 2004, or 475 trading days, and divided each day into six 4-hour periods, with the 5 p.m. rollover as the focal point: 1 a.m. to 5 a.m., 5 a.m. to 9 a.m., 9 a.m. to 1 p.m., 1 p.m. to 5 p.m., 5 p.m. to 9 p.m. and 9 p.m. to 1 a.m. the next day.

It shows the average gains or losses for each four-hour period during the entire week (30 total periods), which start and end at 1 a.m. ET on Monday morning. (The forex market is closed from Friday at 5 p.m. to Sunday at 2 p.m. Our data resumed on Sunday at 5 p.m., except between July 18 and Oct. 31, 2004, when it began on Sundays at 4 p.m.)

A distinct pattern appears in the two four hour periods surrounding each day’s 5 p.m. rollover, which is labeled “R”. GBP/JPY gained ground from 1 to 5 p.m. and fell from 5 to 9 p.m. each day except Thursday. Although GBP/JPY continued to rise immediately following Thursday’s rollover, the currency pair’s rally in the eight hours leading up to rollover was three times as large than its post rollover, eight-hour gain from 5 p.m. to 1 a.m. (0.12 percent and 0.04 percent, respectively).

On Tuesday and Friday, GBP/JPY fell throughout the day except in the four-hour period from 1 p.m. to 5 p.m. It also shows that the currency pair’s tendency to rise prior to this event and sell off afterwards has been strongest on Wednesday. The pound/yen increased 0.11 percent in the 12 hours preceding rollover — the figure’s second highest 12-hour gain — and then dropped -0.05 percent in the subsequent four hours. Despite this brief loss, it shows GBP/JPY’s most bullish period occurred from Wednesday at 5 a.m. to Thursday at 9 p.m.

It (see p. 25) compares various statistics for all 30 four-hour periods(average, median, maximum and minimum values and percentage of positive moves). It’s final six rows summarize the pound/yen’s intraday performance. It shows that each period’s average and median values are roughly in line with each other, which indicates GBP/JPY’s gains and losses are fairly reliable. As expected, all but one of the periods’ average gains also have a greater chance of gaining ground than losing it (i.e., the percentage of positive moves is greater than 50 percent).

Similarly, each of their’s average losses has a winning percentage of less than 50 percent.

It’s larger gains and losses also correspond to greater chances of moving in the desired direction. For example, Thursday’s 0.06-percent increases in the two four-hour segments from 9 a.m. to 5 p.m. have two of it’s three highest probabilities of gains (60 and 65.26 percent, respectively).

Wednesday’s shift from a 0.03-percent average gain from 1 p.m. and 5 p.m. to a -0.05-percent loss following the rollover is the second-largest drop off between time segments.

However, the currency pair’s probability of gains sank from 53.68 to 32.98 percent in that same period — it’s largest probability decline. It’s summary statistics confirm GBP/JPY tended to climb briefly prior to rollover and sink after the event. Overall, the currency pair rose 0.03 percent, on average, from 1 p.m. to the 5 p.m. rollover and then fell 0.02 percent in the following four hours.

Although this change is small, the prior period’s biggest individual loss widened from -0.51 to -0.90 percent, and its chance of gains dropped from 56.42 to 45.36 percent following the rollover — two further signs the decline is accurate.

It highlights GBP/JPY’s average four-hour gains and losses on Wednesday and Thursday and compares them to the currency pair’s overall performance. Though the pound/yen’s Wednesday performance adheres to the established pattern, its bullish behavior on Thursday stands out.

GBP/JPY posted average gains in five of it’s six periods on Thursday, which was the only day the currency pair gained ground following each rollover.

On March 19, 2001, the Japanese government lowered its overnight interest- rate target to 0 percent and has held it at that level for 44 months, which means that traders can borrow the yen and pay no short-term interest. In contrast, Great Britain’s short-term interest rate was 5.75 percent in early March 2001, and has been at least 3.50 percent over the same period.

Therefore, the British pound/Japanese yen has provided a daily rollover payment for buyers since March 2001.

It shows the average daily GBP/JPY performance on each day of the week from March 20, 2001 to October 29, 2004 (935 trading days), and compares each price move to its benchmark move, or typical daily behavior during the same period.

Because the forex market trades 24 hours a day, our Comstock data (via FXtrek) measured each day from 5 p.m. ET to 5 p.m. the next day.

The daily price moves are quite small, but they illustrate interesting patterns. GBP/JPY tended to trade sideways on Monday, lagging its 0.01- percent benchmark. On Tuesday, the currency pair dropped an average -0.03 percent, but Wednesday’s 0.03-percent gain began a modest three-day rally, which beat its benchmark each day and totaled 0.09 percent. It shows the pound/yen posted the largest average daily gain (0.04 percent) on Thursday, and inched 0.02 percent higher on Friday.

It shows each day’s average performance and compares it to its median, benchmark, maximum and minimum values. It also lists each day’s percentage of positive moves. A comparison of the table’s average and median values suggests GBP/JPY’s lackluster performance during the first two days of the week is accurate, but its climb on Wednesday and Thursday might be slightly higher.

For example, the pound/yen’s average is flat on Monday, but its median is -0.04 percent, which suggests a few extreme positive moves skewed the average higher than it should be. Similarly, Wednesday’s and Thursday’s medians are higher than their average values, which suggest the opposite condition.

GBP/JPY’s largest average and median gains (0.04 and 0.09 percent, respectively) as well as its highest probability of gains (57.75 percent) occurred on Thursday.

Getting a straightforward answer about daily rollover fees from your forex broker isn’t easy. Although rollover fees are based on the short term interest-rate differences of the two currencies you exchange for one another in the forex spot (cash) market, each FX dealer has its own rules, which can make the process confusing.

The spot market has a two-day settlement period, which means that if you buy one GBP/JPY standard lot (£100,000) on Tuesday, the transaction will settle on Thursday. If you hold this trade past 5 p.m. on Tuesday, your broker will roll the settlement forward to Friday and may add between $1 and $20 to your account as they calculate interest you earned.

As of Nov. 16, Great Britain’s short-term interest rate was 4.6 percent higher than Japan’s (4.75 percent and 0.15 percent, respectively, according to Forexnews.com). Therefore, the daily rollover credit on this trade is roughly $12.60, or ($100,000 * 4.6 percent)/365. (The actual rollover amount depends on your balance and the difference between GBP’s borrowing rate and JPY’s lending rate.)

Wednesday’s rollover is three times as large because it accounts for interest earned over the weekend. For example, if you enter and exit a trade on Wednesday prior to rollover, both trades settle on Friday. However, if you hold the trade “overnight” and close it after the rollover, your trade won’t settle until Monday, and you earn (or pay) three days of interest instead of one. Trades that should settle on holidays also earn (or owe) an additional day’s interest.

However, forex dealers are not obligated to pay interest. Some brokers, such as Forex Capital Markets (FXCM), only credit your account if you trade with at least 2 percent margin. Oanda calculates interest earned on currency pairs eachsecond, instead of once a day. Other brokers charge a daily rollover fee even if you’re long a higher-interest-rate currency.

Most FX dealers’ rollover time is 5 p.m. ET, but other brokers such as GFT Forex and MG Financial Group roll positions forward at 3 p.m. ET.

Many traders view the carry trade as a longer-term (i.e., days, weeks or months) approach since the interest you earn is based on the trade’s length. In theory, however, it’s possible to buy a currency pair such as the British pound/Japanese Yen (GBP/JPY), whose base currency’s interest rate (4.75 percent) is much higher than the quote currency’s rate (0 percent), just before the day’s rollover, earn the rollover credit and exit the trade following this event.

Because GBP/JPY’s rollover credit is three times as large on Wednesdays to account for the weekend (see sidebar), this study focuses on the currency pair’s behavior in the days and hours surrounding this weekly occurrence. We analyzed daily GBP/JPY performance since March 2001 to find out how Wednesday’s triple rollover costs influenced this pair’s behavior. We then studied hourly GBP/JPY price data surrounding each day’s 5 p.m. rollover time since January 2003 to find out how this event affected intraday price moves.

Overall, GBP/JPY slumped on Monday and Tuesday, but rose Wednesday through Friday. On an intraday basis, GBP/JPY climbed as the 5 p.m. rollover approached and sold off in the early evening — a pattern that was magnified on Wednesday. Both tendencies suggest traders bid up prices in anticipation of the daily rollover, especially Wednesday’s three day interest-rate payoff.

Rollover fees are often overlooked, yet they can seriously affect a forex trade’s profit or loss, depending on the currency pair, your forex dealer’s rules and each trade’s length. These charges are based on the interest-rate difference between the currency you buy and the one you sell (or vice versa) when you trade a currency pair and hold it past the daily rollover time — typically 5 p.m. ET. For example, if you go long GBP/USD and hold it “overnight,” or past 5 p.m., your forex broker may give you a small credit because you bought British pounds and sold U.S. dollars at the same time, and shortterm interest rates are currently higher in Great Britain than the U.S. Similarly, if you short this currency pair, your broker may charge a small fee as the trade “rolls over” into the next day. (For more information about rollovers, see “Rollover fees: Understanding the fine print.”)

While rollovers don’t apply to intraday traders, this feature is the crux of the carry trade, or buying a currency with a higher interest rate and simultaneously selling one with a lower rate and earning the difference between both rates (see “Getting a lift from the carry trade,” Currency Trader, Oct. 2004). Like dividends, any rollover fees you earn (or owe) are separate from the currency pair’s gains or losses, but they can help enhance a trade’s profit or mitigate its loss if the daily rollover is in your favor.

The unusual rally in the U.S. dollar after presidential elections implies going long on the sixth day after the election and exiting three months later. But two factors suggest caution: the dollar’s rise is based on only seven instances and the index hasn’t rallied following close elections. The good news is that over the past 31 years, the dollar has climbed modestly after election day, so buying the dollar in earlyNovember and holding it through the end of the month may be profitable, depending on current market conditions.

Although the Fed’s dollar index represents the longest set of available historical data (1973 to 2004), it is helpful to study past performance of the U.S. dollar index futures contract to see whether the trends discussed previously appear here. However, there are two problems with this comparison: The futures contract has only been trading since 1985, and the number is weighted differently than its Fed counterpart.

Despite these limitations, we analyzed both data sets over the past 18 years to see whether they have moved in line with each other. It compares the average gains and losses of both instruments and their benchmarks in eight periods surrounding election day.

Comparing reveals the U.S. dollar has lost more ground around elections over the past 18 years than since 1973. While the Fed’s index posted smaller average losses than the NYBOT’s futures contract, neither instrument consistently gained ground after election day.

However, the U.S. dollar’s futures contract moved in the same direction as the Fed’s index in each of the preelection periods; the relationship between both instruments was less clear after the election. Also, the Fed index and the DX contract followed the same pattern around presidential and mid-term congressional elections.

The dollar’s most interesting trends appear when mid-term congressional and presidential elections are measured as separate (and smaller) categories. There have been only eight non-presidential congressional elections since 1973, and It shows that the dollar sharply dropped in the 60 and 40 days both before and after election day. However, the index was more bullish in the three weeks before and (especially) four weeks after it.

The longer-term price moves appear less dramatic when compared to their benchmarks, or typical same-length losses during these eight years. For example, the -1.53-percent average decline in the 60 days prior to the election is roughly inline with its 1.40-percent benchmark loss in that period; the
dollar’s performance in the 60 days following mid-term congressional elections tells a similar story.

Also, the dollar’s average gains in the 15 days around these eight elections are unusual because its benchmarks lost ground in these periods. The dollar’s strongest gains occurred in the 10 days before and 15 days after congressional elections (0.65 and 0.46 percent, respectively) not only because they outperformed their benchmarks, but they were also in line with their median values.

It‘s final category shows the seven presidential elections had the opposite effect on the U.S. dollar. Here, the dollar posted its largest average gains in the longer-term (60- and 40- day) periods surrounding these elections. Also, the dollar’s presidentialelection- year benchmarks were quite bullish, which means not all the gains prior to and following this event are as noteworthy as they seem.

For example, the index’s 0.91-percent average gain in the 60 days before a presidential election was slightly below that period’s benchmark. Despite this caveat, the U.S. dollar rallied strongly starting in the third week after presidential elections, culminat ing in a 2.76-percent average gain by the 60th day following them.

They highlight the differences between the dollar’s average gains and losses around
presidential and mid-term congressional elections, respectively.

The dollar gained ground leading up to presidential elections, but actually lagged behind its benchmarks in four of the six periods. However, the index rose 1.07 percent, on average, in the 40 days preceding the event, beating its benchmark by 0.42 percent. The figure also clearly shows the dollar’s outstanding climb from the 10th day after election day. It reiterates the dollar’s fall in the 60-, 40- and 20-day periods before congressional elections.

While this downtrend reappeared between the 20th and 40th day after the event, the dollar posted gains and beat its benchmarks in the intervening seven periods (e.g., 15 days prior to and 20 days following it).

Although the dollar moved in a distinct pattern around the past 31 elections, its behavior varied based on the type of election. It shows the dollar’s average gains and losses in the 12 periods surrounding all election days and compares it to the dollar’s performance in even-numbered years — either presidential or mid-term congressional elections — and the benchmark moves for the periods of the same length over the past 31 years.

The dollar was more bullish surrounding presidential and midterm congressional elections than during less-significant ones. In even-numbered years, the index posted either smaller average losses or larger gains in 11 of the figure’s 12 periods.

The index’s performance in the first month after the election highlights this trend. In general, the dollar traded sideways in the second week following an election, but after presidential and congressional elections, its 0.17-percent, five-day gain doubled to 0.34 percent by the 10th day. This postelection rally during even-numbered years culminated in a 0.61- percent average gain by the end of the first month — nearly three times as large as the 0.21-percent climb in the same period overall.

It breaks down average gains and losses into different election types: presidential and midterm congressional; odd-numbered (non-national election) years; midterm congressional only (1974, 1978, 1982, etc.); and presidential only. Comparing the first category’s average and median values reinforces conclusions drawn. The dollar’s median value was higher than its average in eight of the 12 periods around elections in even-numbered years. Therefore, the dollar has likely performed better before and especially following presidential and mid-term congressional elections.

The table also shows the U.S. dollar sold off leading up to and after oddyear elections. While the index gained an average 0.18 percent in the 15 days before these events and 0.24 percent by the 60th day following them, it fell or traded sideways in the table’s 10 remaining periods.

To see how the Fed’s U.S. dollar index has fared around elections from August 1973 to January 2004, we measured its performance in six periods both before and after each election day. We tracked the gain or loss from the 60th and 40th days before election day to the day before it. We also measured
the dollar’s price moves from the preceding 20th, 15th, 10th and fifth day prior to each election to the day immediately before it. To gauge how the U.S. dollar performed after elections, we then studied its behavior from the day before the event to the fifth, 10th, 15th, 20th, 40th and 60th days after it.
It shows each election date over the past 31 years as well as the U.S. dollar’s individual gains and losses in the 12 periods surrounding it. For example, “Day -40 to Day -1” measures the dollar’s change from the 40th day preceding the election to the day before it; “Day -1 to Day 40” represents its gain or loss from the prior day to the 40th day following it. (Day 1 is election day.)

The table’s final rows show the dollar’s average, median and benchmark performance, or typical same-length performance since 1973. Each period’s probability of gains is also shown The dollar lost an average -0.95 percent in the 60 days before the election and nearly as much in the prior 40-day
period. These losses faded in the month preceding election day and it gained roughly 0.22 percent in the 15- and 10-day periods.

While the dollar lost an average -0.05 percent on election day (not shown), the index rose 0.10 percent by the end of the first week after it. Also, by the end of the 20th day, or a month after the election, the dollar climbed 0.21 percent. Although the dollar then fell -0.25 percent by the 40th day, it rebounded in
the third month.

Comparing the table’s average and median values reveals a more accurate picture of the dollar’s typical price moves. For example, the -0.95-percent average loss in the 60 days leading up
to the election was likely skewed lower by a few unusually large losses, such as the -7.59-percent sell-off in 1998, since its median value was flat. However, the -0.81-percent drop in the 40-day period is supported by the similar decline in the median value. Also, the aforementioned -0.25-percent average
move from Day -1 to Day 40 is countered by a 0.13-percent median increase for that period. (For more
information about discrepancies between average and median values, see “Average and Median,” p. 35.) It also shows the dollar has welcomed decisive presidential victories. The index has gained at least 1.62 percent by the 60th day after five of the table’s seven presidential elections. The years in which the dollar didn’t perform as well in this period represent close elections.

For example, the disputed 2000 election between Al Gore and George W. Bush led to a -2.32-percent decline, and Jimmy Carter’s fairly narrow victory over Gerald Ford in 1976 boosted the dollar only 0.76 percent by the end of the analysis window. Since neither John Kerry nor George W. Bush command a significant lead immediately before the 2004 election, the dollar may not rally as expected.

In the weeks leading up to the recent U.S. presidential election, the question on everyone’s mind (besides the outcome) has been how the event might affect the financial markets. Although most of the attention has been on the election’s impact on the stock market, analysts are also focused on the fate of the U.S. dollar, which is again approaching the eight-year lows hit earlier this year.

Discussions about how this election may influence the U.S. dollar are limited by the infrequency of presidential elections, making it difficult to draw solid conclusions. Also, fundamental analysis (e.g. the implications of economic policy) can tell you more about the dollar’s longer-term trends than its
daily (or weekly) movement surrounding an election.

To find out how elections — presidential and congressional — typically affect the U.S. dollar and whether the event could knock the dollar out of its slump and induce a rally in the next three months, we studied the moves of the Federal Reserve’s U.S. Dollar Major Currencies Index in the 60 trading
days before and after each election from 1973 to 2003.

While this period is relatively short for studying an annual event, it represents an important era in the U.S. dollar’s history. In March 1973, the United States and other major world governments decided to let their currencies trade freely against each other, a policy which scrapped nearly 30 years of fixed currency rates under the prior Bretton Woods Agreement. We also compared the behavior of the Fed’s U.S. dollar index to the New York Board of Trade’s U.S. dollar index futures contract (DX), which was
launched on Nov. 20, 1985, by the New York Cotton Exchange. While the Fed’s major-currency index acts as a proxy for the dollar’s strength, the NYBOT’s dollar index futures can be traded, so its historical performance adds perspective to this analysis.

Overall, the U.S. dollar tended to start sinking three months prior to an election and then rise modestly in the three weeks leading up to the event. This rally typically continued for a month following election day. The dollar posted its largest average gains after presidential elections compared to mid-term congressional elections.

This study reveals several pieces of information that could contribute to testable trade ideas. For example, short the EUR/USD following a higher-close wide-range bar, especially if it closed in the top 10 percent of its range and U.S. economic data was released that day. Similarly, go long after a lower close WRB.

Exit the trade one day later. For a longer-term trade strategy, buy the currency pair on the fourth day after a WRB that closed in the upper portion of its range on breaking economic news, and hold the trade until the 10th day.

Other ideas to explore: How does European economic news affect WRBs in the EUR/USD? Also, does the currency pair react to intraday (i.e., fiveminute, 30-minute, one-hour etc.) wide-range bars in a similar way?

These scenarios may not lead to tradable price moves, but additional insight into the currency pair’s historical behavior will only help you make smarter trading decisions.

Although many traders expect the market to continue moving in the direction of a wide-range bar’s close (whether it’s at the extreme upper or lower end of the bar), it shows EUR/USD tended to defy conventional wisdom and sink on the day after high closes and climb following low closes.

Economic news heightened this effect, which suggests traders initially overreacted to it. Also, the most extreme days (90th-percentile WRBs combined with the highest and lowest close locations) triggered the largest reversals. The table breaks down 80thand 90th- percentile WRBs into four main groups: closes within the upper or lower 20 percent of their range, and those in the upper or lower 10 percent. It also shows the WRBs with economic news in each category.

The first column, which represents the first day following extreme closes, contains the table’s most interesting price behavior. Nearly all (15 of 16) categories posted counter-intuitive moves that day, and economic news led to larger reversals in six of the table’s eight instances.

For example, 80th-percentile WRBs that closed in the upper 20 percent of their range fell an average 0.09 percent the next day; news-related WRBs with the same characteristics lost 0.19 percent. Similarly, WRBs that closed within the lower 20 percent rallied 0.16 percent the following day; however, wide-range bars that included economic news jumped a mere 0.06 percent (one of two exceptions to the rule).

The table’s last four rows, which show 90th-percentile WRBs that closed in the upper and lower 10 percent (overall and with news), are the best examples of EUR/USD’s tendency to veer in a different direction the day after economic news hit the Street.

The currency pair’s 90th-percentile wide-range bars weakened 0.08 percent the first day following closes in the top 10 percent and gained 0.11 percent

succeeding unusually low closes. Economic news magnified these next day price moves: The highest news related closes sank 0.33 percent while the lowest closes jumped 0.24 percent when combined with economic news. It’s remaining columns indicate the next-day reversals didn’t last, and all WRBs headed higher in the second week, led by those with extreme closes.

Although these trends are impressive, WRBs with the strictest criteria represent the smallest number of cases.

It reinforces the effect of economic reports and compares the average performance following all 80thpercentilewide-range bars to news related bars. The figure also shows EUR/USD’s benchmark price moves.

The figure highlights patterns detailed: No solid trends developed in the first five daysfollowing WRBs, regardless of whether they featured economic news or not. (Day 1’s drop after news-related WRBs is an exception to this rule.)

All wide-range bars (including those corresponding with economic news) began to outperform their benchmark price moves by the sixth day, and the news category beat overall WRBs starting on the seventh day. This trend intensified toward the end of the second week.

There were 319 daily bars that met this requirement from April 1, 1999, to Aug. 20, 2004. We measured how the EUR/USD reacted to wide-range bars by tracking its performance from the close of a WRB to the close on each of the following 10 days, as well as the closes of the 15th and 20th days afterward.

The market tended to climb higher, on average, in the two weeks following wide range bars, but the most compelling price moves followed WRBs with strong or weak closes — i.e., closing prices in the upper or lower 20 percent of their daily ranges. The EUR/USD tended to briefly reverse direction in the wake of such bars before outperforming other WRBs in subsequent weeks.

Because surprises in U.S. economic data such as Federal Reserve Bank announcements, gross domestic product (GDP), inflation (CPI and PPI), employment, trade balance and current account deficit impact EUR/USD price behavior, we also specifically studied WRBs that occurred on report release days. Roughly one-third of 80th percentile WRBs included some type of economic data released that day. The reports tended to enhance price patterns, producing larger gains or losses than either regular WRBs or those with extreme closes did alone.

Volatility is one aspect of market behavior that is always important to watch. It comes in a few different forms ( implied volatility and historical volatility) and can be analyzed from a variety of perspectives.

One of the basic things to know about volatility — in terms of the amount of price movement in a particular market — is the time period in question. It is one thing to know a market gained 20 percent, but it’s something else to know whether it made this move in two weeks or one year.

When it comes to price charts, the simplest unit of time to consider is one bar, whether that reflects one hour, one day or one week. As a result, traders often study a market’s range during a price bar, or the difference between its high and low. This measures the amount of intraday price movement; true range measures the degree of movement from bar to bar (period to period).

We studied exceptionally large daily bars in the euro/U.S. dollar currency pair (EUR/USD) over the past five and a half years (since the euro’s inception in 1999) to see whether recognizable price patterns developed after wide range bars (WRBs) occurred.

USD/CAD. After nearly taking out highs not seen since 1979, the recent technical breakdown in the Reuters-CRB Futures Index (CRB) spells trouble for “commodity-based” currencies such as the Canadian dollar (CAD), Australian dollar (AUD), and New Zealand dollar (NZD). Industrial production in Japan is decelerating, which is usually bearish for commodity prices. Also, China’s leading economic indicator hasn’t shown signs of rebounding, which could depress commodity prices in the intermediate term.

Falling crude oil prices should depress the Canadian dollar. Speculative positions in oil are very high. A more pronounced setback to the price of oil, which supports Canada’s energy-loaded equity market, is negative for the CAD. The negative correlation between the U.S. dollar/Canadian dollar (USD/CAD) and the CRB Energy Index is quite striking, and provides the foundation for a long USD/CAD position for longer-term traders if the CRB Index remains under pressure.

On the U.S. dollar side, the tug-of-war between the positive cyclical forces and negative structural backdrop for the U.S. dollar continues. Cyclical forces, such as stronger growth in the U.S. and Federal Reserve tightening, will push the dollar higher in the intermediate term, but the technical pattern emerging is complex and will likely continue to frustrate both dollar bears and bulls.

Medium-term players are still holding short positions in the U.S. dollar, which will accelerate an up move in the currency if they are forced to liquidate these trades. Although sentiment toward the dollar has turned up in recent weeks, it has not reached the same level as in previous corrections, which supports a view that medium term traders have not yet capitulated.

AUD/CAD. A bearish position in the Australian dollar/Canadian dollar (AUD/CAD) can be based on several macroeconomic factors. In March, consumer confidence in Australia made its largest ever monthly drop. Consumer spending is also weak. The recent setback to confidence and spending is troubling, given the unemployment rate in Australia is so low (just more than 5 percent) and the Australian stock market has been rising nicely for some time. With this backdrop, a rate hike at the April 28 Australian central bank meeting is not a sure thing.

Conversely, Canada, with its strong economic ties to the U.S., continues to show signs of economic expansion. Canada runs a current account surplus, while Australia has one of the worst current account deficits in the Group of Ten (G10).

New manufacturing orders and unfilled orders have been very strong. Canadian manufacturing shipments are highly correlated with retail sales in the U.S., which are rising at the present time.

The Bank of Canada (BOC) could begin to raise interest rates sooner and at a faster pace than most traders and money managers are expecting. However, the BOC is notoriously difficult to forecast in terms of their monetary policies.

David Dodge, a Governor of the BOC, has said a 200- to 300-point interest rate increase is possible because even at that level the country would still have debt service ratios well below the 20-year national average.

The divergence developing between these two economies/countries is clear. A contraction in their rate differential and weaker commodity prices will impact Australia more than Canada.

Technically, however, traders might need to wait a bit longer before deciding to commit capital. It would be beneficial to see a break of weekly support around .9425-.9450; at that point there will be more confidence that the marketplace shares in this view of the currency pair.