# Composite indexes of leading, coincident, and lagging indicators.

Composite Indexes of Leading, Coincident, and Lagging Indicators

BEA's composite index of leading indicators increased 0.2 percent in April; the contributions of the individual components to the change in the index are shown in table 1. This estimate is preliminary and is based on data for 9 of the index's 11 components. (The table on the facing page shows the etimates as both levels and as percent changes; in this discussion, "estimates" refers to percent changes.)

In March the leading index also increased 0.2 percent. This first revised estimate for March is based on data for all 11 of the index's components. The preliminary estimate for March, released last month, was a 0.8-percent increase.

For any month, the difference between the first revised estimate and the preliminary estimate is the first revision. The size of the first revision for March, 0.6 percentage point, is larger than usual. First revised estimates are subject to further revision for as many as 10 additional months, as source data for the components undergo revision. The leading index is also subject to occasional comprehensive revisions to introduce changes in the component series and in statistical procedures.

During the 10-year period 1977-86, the first revision accounted for about two-thirds of the total difference between the preliminary and final estimates. During the past 12 months the average absolute size of the first revision, 0.37 percentage point, was virtually identical to its 1977-86 average of 0.35 percentage point. (However, the average absolute value of the preliminary estimate during the past 12 months, 0.64 percent, was smaller than that during 1977-86, 0.88 percent. Thus, relative to the preliminary estimates, first, revisions during the past 12 months have been larger than usual.)

The first revision for March is typical in that the inclusion of the credit and inventory components, which are not available when the preliminary estimates are prepared, accounted for the biggest part of the revision. March's first revision also reflects (as do first revisions in general) revisions in source data for some of the components.

The inclusion in the first revised estimate of two components that are not included in the preliminary estimate has both a direct and an indirect effect. The direct effect is simply the sum of the contributions of the two components. The indirect effect is the change in the sum of the contributions of the other nine components that results from the reduction in their effective weights when additional components are included. These direct and indirect effects offset each other if the two components omitted from the preliminary estimate behave in the same way as the nine included components.

In table 2, the first revisions in the leading index are separated into two parts--one attributable to revisions in the source data for the original nine components and the other attributable to the inclusion of these two components; this latter part is further separated into its direct and indirect effects. In March, for example revisions in source data for the nine components contributed 0.06 percentage point ot the first revision; including the creidt and inventory components made a direct contribution of --0.52 percentage point and an indirect contribution of --0.14 percentage point.

Clearly, the inclusion of these two components was the more important factor in the first revisions during the past 12 months; the average absolute value of this effect was almost three times as large as that of revisions in source data for the other nine components. The direct effect of including two additional components usually far outweighed the indirect effect, but on occasion (e.g., November 1987) the indirect effect dominated.

Eliminating the part of the first revision that results from including additional components in the first revised estimate is one goal of research currently underway at BEA. One possibility is to replace these two components with others that are available in time for inclusion in the preliminary estimate; another is simply to remove these two components from the index. The results shown in table 2 suggest that either alternative amy reduce the first revisions to the leading index.

BEA's composite index of leading indicators increased 0.2 percent in April; the contributions of the individual components to the change in the index are shown in table 1. This estimate is preliminary and is based on data for 9 of the index's 11 components. (The table on the facing page shows the etimates as both levels and as percent changes; in this discussion, "estimates" refers to percent changes.)

In March the leading index also increased 0.2 percent. This first revised estimate for March is based on data for all 11 of the index's components. The preliminary estimate for March, released last month, was a 0.8-percent increase.

For any month, the difference between the first revised estimate and the preliminary estimate is the first revision. The size of the first revision for March, 0.6 percentage point, is larger than usual. First revised estimates are subject to further revision for as many as 10 additional months, as source data for the components undergo revision. The leading index is also subject to occasional comprehensive revisions to introduce changes in the component series and in statistical procedures.

During the 10-year period 1977-86, the first revision accounted for about two-thirds of the total difference between the preliminary and final estimates. During the past 12 months the average absolute size of the first revision, 0.37 percentage point, was virtually identical to its 1977-86 average of 0.35 percentage point. (However, the average absolute value of the preliminary estimate during the past 12 months, 0.64 percent, was smaller than that during 1977-86, 0.88 percent. Thus, relative to the preliminary estimates, first, revisions during the past 12 months have been larger than usual.)

The first revision for March is typical in that the inclusion of the credit and inventory components, which are not available when the preliminary estimates are prepared, accounted for the biggest part of the revision. March's first revision also reflects (as do first revisions in general) revisions in source data for some of the components.

The inclusion in the first revised estimate of two components that are not included in the preliminary estimate has both a direct and an indirect effect. The direct effect is simply the sum of the contributions of the two components. The indirect effect is the change in the sum of the contributions of the other nine components that results from the reduction in their effective weights when additional components are included. These direct and indirect effects offset each other if the two components omitted from the preliminary estimate behave in the same way as the nine included components.

In table 2, the first revisions in the leading index are separated into two parts--one attributable to revisions in the source data for the original nine components and the other attributable to the inclusion of these two components; this latter part is further separated into its direct and indirect effects. In March, for example revisions in source data for the nine components contributed 0.06 percentage point ot the first revision; including the creidt and inventory components made a direct contribution of --0.52 percentage point and an indirect contribution of --0.14 percentage point.

Clearly, the inclusion of these two components was the more important factor in the first revisions during the past 12 months; the average absolute value of this effect was almost three times as large as that of revisions in source data for the other nine components. The direct effect of including two additional components usually far outweighed the indirect effect, but on occasion (e.g., November 1987) the indirect effect dominated.

Eliminating the part of the first revision that results from including additional components in the first revised estimate is one goal of research currently underway at BEA. One possibility is to replace these two components with others that are available in time for inclusion in the preliminary estimate; another is simply to remove these two components from the index. The results shown in table 2 suggest that either alternative amy reduce the first revisions to the leading index.

Printer friendly Cite/link Email Feedback | |

Publication: | Survey of Current Business |
---|---|

Date: | May 1, 1988 |

Words: | 687 |

Previous Article: | National income and product accounts tables; selected NIPA tables and reconciliation and other special tables. |

Next Article: | Pollution abatement and control expenditures, 1983-86. |

Topics: |