1. While short-term price movements are mainly the result of existing momentum, the direction of this momentum will ultimately be determined by a preponderance of fundamental data. So during an uptrend in a currency, we often see the currency reacting well to good news and seemingly ignoring bad news. However I believe that if the bad news continues to accumulate, then the currency's momentum will eventually shift as a result. The trick is to keep track of the "accumulation" of fundamental pressure on one side or the other.

2. Surprising results have more impact than expected ones. It is not the actual economic numbers that are important in themselves, but the divergence between what the analysts expect and the actual number. I studied this phenomenon in detail in this research project:
February 4, 2007 "Economic Reports vs. FOREX Price Behavior"

Below is a description of the method. I have a spreadsheet set up that does all the calculations automatically, so although the method may look complex it's not very time consuming.

1. Using the calendar at Forex Factory, I look at a 4-week window of economic news releases in order to keep track of the accumulation of fundamental pressures. I use only the releases for which there are both a forecast and an actual number.

2. Most economic releases are expansionary and bullish for the currency if the economic number increases, i.e. CPI, GDP, Retail Sales, etc. However, some numbers like inventories and unemployment levels have an inverse relationship; the higher they are the more contractionary and bearish the indication. In order to maintain the convention that "higher means bullish," I simply insert negative signs into the data for these inverse type releases.

3. In order to calculate a "surprise index" value that is bounded by -1 and +1, I compare the difference between the forecasted number and the actual number to the sum of the absolute values of the forecasted and actual change from the previous month. The actual formula is:
(Actual - Forecast) / (|Actual - Previous| + |Forecast - Previous|)

4. For example, suppose the CPI for a country was 1.3% last month, the forecast for this month was 1.5% and the actual number is 1.4%.
Actual - Forecast = 1.4% - 1.5% = -0.1%
Absolute value of forecasted change = |1.5% - 1.3%| = 0.2%
Absolute value of actual change = |1.4% - 1.3%| = 0.1%
Surprise Index = -0.1% / 0.3% = -0.33

5. If the actual number came out drastically higher than the forecast, say 2%, then the index would be:
0.5% / (0.2% + 0.7%) = 5/9 = +0.55
This reflects my research findings that a higher actual change in the same direction as the forecasted change has the least amount of "surprise effect." It would take a very large increase in CPI to get the index anywhere close to a value of +1.

6. If on the other hand the CPI went even slightly in the opposite direction from the forecast, we would immediately have a very large surprise reflected in an index value of -1. For instance if actual CPI were to drop to 1.2% the surprise index would be:
(1.2% - 1.5%) / (0.2% + 0.1%) = -3/3 = -1

7. Once I have the surprise index value for each release, I multiply it by 1, 2, or 3 depending on the projected impact of the release as indicated by Forex Factory's color code system. So now each release has an index value that varies between +3 and -3.

8. I add up all the index values for each currency, and divide by the number of releases for that currency on that day. This way, if the NZD has one release and the USD has six releases, the total value of the USD news won't be six times as large as the NZD news. This gives me a final daily value for each currency, reflecting the effect of its news releases.

9. Once I have the final daily value for each currency, I add it to a cumulative index value for that currency to create the chart on the homepage.

Home

Trading Journal

Research Archives

Play HyperDonut!