Forex Trading System- The Mundo

OVERVIEW

In Currency Trader’s Companion: A Visual Approach to Technical Analysis of ForexMarkets(2004), we introduced the synthetic global currency, the Mundo, to which wegave an International Organization for Standardization (ISO) symbol of ICU for interna-tional currency unit. We arbitrarily defined the prevailing price of the Mundo as thearithmetic average price of the 10 most frequently traded major and minor USD cur-rency pairs: EURUSD, GBPUSD, USDJPY, USDCHF, USDCAD, AUDUSD, NZDUSD,USDSEK, USDNOK, and USDDKK.

All pairs are treated with equal weight, which means six of the pairs must be ad-justed so that the USD is the quote currency. The reciprocals (divide the exchange rateinto 1) of the USDJPY, USDCHF, USDCAD, USDSEK, USDNOK, and USDDKK must becalculated to create the JPYUSD, CHFUSD, CADUSD, SEKUSD, NOKUSD, and DKKUSDpairs. This ensures that any price change in any of the 10 pairs is measured in pips of thesame currency unit, the USD.
Summing the most recent values of these 10 pairs and dividing by 10 yields the cur-rent price of the Mundo.

FOREX BETAWe can now use an analogous coefficient to compare the volatility of a single currencypair to the volatility of the overall forex market as described in terms of the volatility ofthe Mundo.

Rather than using the ratio of the slope of a single currency pair and the slope ofthe ICUUSD pair, we will use the standard deviation of each. We can justify this change because of a major difference in the trading philosophies of stocks and spot currencies.Nearly all stock traders use a buy-and-hold strategy in which they hope that their invest-ment will more than better the current inflation rate over a long period of time. Thusstock traders hold a long position in their trades and, in nonleveraged positions, theyown the shares of stock outright.

Spot currency traders, by contrast, are not buying shares in a corporation or a mu-tual fund. They feel equally comfortable on either side of a currency trade, long or short.A forex trade is, in fact, the simultaneous buying of one currency while selling anothercurrency. Spot currency traders (particularly scalpers) may initiate a long trade, followa five-minute rally, liquidate the long position at its peak, and then initiate a short posi-tion in the same currency pair and follow that security’s decline to the next trough.

Therefore, spot currency traders are not particularly interested in the long-termslope of any currency pair. Instead they are more interested in the number of significantpeaks and troughs that occur during their trading sessions. The standard deviation istherefore employed in our model for forex beta (though many may agree that this is amisnomer).

To calculate the forex beta of, say, the EURUSD pair see Figure 6.1.

A running calculation of this statistic using streaming data informs the forex daytrader which currency pairs are showing the highest volatility relative to the whole spotmarket. This identifies the pairs with the highest risk/reward factor. The order of thesepairs may change throughout the day as central banks around the world open and close.

Table 6.1 illustrates the standard deviation and beta coefficient for each of the 10
pairs for the time range 1/1/2000 through 6/30/2003. During this period, the Mundo ex-hibited a standard deviation of 352.70 pips (or 3.53 U.S. cents).

Thus, when compared to the aggregate currency pair ICUUSD, the EURUSDshowed the highest beta coefficient and therefore carries the greatest risk/reward fac-tor. In other words, for the time period examined, the EURUSD was 2.28 times morevolatile than the average of all 10 currencies.

MUNDO LINE CHART

Figure 6.2 is the line chart for the Mundo during the same time frame.
The first aspect to note is that the Mundo has a lower parity rate than the USD,roughly 53 to 67 U..S cents. Second, slight changes in the same price direction in the 10 underlying currency pairs may cause an exaggerated price change in theMundo. This phenomenon is also partially due to the use of reciprocals in six of thecurrency pairs.

This study is in no way definitive or authoritative. It is simply a novel approach tocurrency pair selection using aggregate standard deviations.

WEIGHTING

The preceding example illustrates the calculation of the Mundo as the simple arithmeticmean of the 10 most frequently traded USD currency pairs. A more realistic approach isto weigh each component USD currency pair using its corresponding activity as theweighting factor. For this purpose, we have used the daily activity data for 1/1/2004through 3/31/2004. (See Table 6.2.)

MUNDO BAR CHART

The price of the Mundo ranged from 0.80 USD to 0.86 USD during the first quarter of2004 (See Figure 6.3.). Activity was calculated by summing the upticks and downticks ofeach component currency pair and dividing by 10. The standard deviation of the ICUUSDcurrency pair over this time period was 0.01147 USD. The beta coefficients for the tencomponent currency pairs using activity weighting are shown in Table 6.3.

Examples of usage for this table are: The GBPUSD is 2.5 times more volatile thanthe Mundo. The CADUSD pair is almost identical to the Mundo in volatility. TheJPYUSD pair occupies the bottom position probably because of its low parity exchangerate (110 JPY = 1 USD).
OBSERVATIONThe effect of weighting the currencies by their corresponding activity is illustrated inthe fact that the order of the currency pairs has changed, somewhat dramatically, fromthe order in Table 6.1 in which simple arithmetic averaging was employed.

The classical tools of technical analysis are well revered and time-tested. Our ap-proach here has been to capitalize on characteristics indigenous to a specific securitygroup (spot currency prices) and to develop unique tools to assist the trader in the un-covering of security properties that may prove enlightening and profitable. We are con-tinually tweaking these new methods and apologize for any infractions that may appearto violate any nuances of mathematical rigor.