While being able to chart how any number of your time series have performed over a defined period is certainly useful, you may also want to see the correlation between two of these time series. For example, besides seeing the performance of the 1-year IR swap rate and the 10-year IR swap rate you may also want to see the correlation between the two market data parameters over the same period.
In the Historical Analysis Page, once you have created two or more time series in Section 1 (see Working With Time Series Definitions), in Section 2 you can then create correlation definitions. Each correlation definition lets you plot the historical correlation between any two time series predefined in Section 1. Once you have created the correlation definitions, you can then create charts for them in any combination (see Displaying Charts Of Historical Market Data).
To define a correlation definition, in addition to specifying the two time series for which you want to see the correlation, you must also specify the following parameters:
Frequency
The Frequency dropdown list specifies the period (in terms of business days) used to calculate the returns data for each of the selected time series. The returns data is then used in the correlation calculation.
You can specify frequencies of 1d, 2d, 3d, 4d or 5d. For example, if you specify a frequency of 1 day, then for each time series the returns data for each day is calculated by taking the results on that day / results on the previous business day. If you instead set the frequency to 2 days, then for each time series the returns data for each day is calculated by taking the results on that day / results on the day that is two business days prior to it.
Term
Defines the number of samples taken when calculating the correlation.
You define the duration using a number and one of the SD shortcuts—d or w or m or y. For example, if you specify 3m, each point plotted on the chart will indicate how correlated the two sets of market data were over the three months up to that date; if you specify 1y, each point plotted on the chart will indicate how correlated the two sets of defined market data were over the year up to that date.
You can also delete existing correlation definitions.
To define a new correlation definition:
| 1. | In the Historical Analysis page in Section 2 click the Add New Line button. |
| 2. | In the Series 1 and the Series 2 columns define the two time series between which you want to see the correlation. |
To delete an existing correlation definition:
| 1. | In the Historical Analysis page in Section 2 check the correlation definition you want to delete. |
| 2. | Click the Remove Line button. All checked correlation definitions in Section 2 are deleted. |