When Leading Indicators Work—and When They Fail
When Leading Indicators Work—and When They Fail
A German industrial-production study compares economic indicators and identifies periods when every tested signal weakened. This is why validation matters.
A German industrial-production study compares economic indicators and identifies periods when every tested signal weakened. This is why validation matters.
Forecast Validation
Why a leading indicator needs proof
A series can sound economically relevant without consistently moving ahead of the outcome you care about. This paper examined that problem for German industrial production from 1991 to 2016. It compared survey-based ifo and ZEW indexes with order-flow and unemployment data.
How the authors checked the signals
The researchers used a leading–lagging method to assess which series tended to move before industrial production and how those relationships changed over time. This is not itself a sales-forecasting model. It is a way to judge candidate indicators and spot periods when their apparent lead becomes weak or unreliable.
A lead can weaken without disappearing
An indicator can still move broadly with industrial activity while offering less advance notice than before. For a planning team, that matters because a signal that arrives after orders have already slowed adds little warning. Reviewing its timing across several upturns and downturns can help distinguish a useful lead from a relationship driven by one period.
How a company can assess its own signals
Suppose a company uses an industry confidence index to review its demand assumptions. It can compare the index with its own sales cycle and examine the dates when their directions changed. If the lead was useful in some periods and inconsistent in others, the team can look for confirmation from indicators that capture different parts of its market.
What the study observed
In the authors’ analysis, the ifo and ZEW sentiment indexes had the strongest overall leading relationship with German industrial production. Recession windows were easier to anticipate than recovery windows. Around 1997 and 2005, all tested indexes performed weakly—an important reminder that an indicator is not permanently reliable.
What this means for business planning
One national industrial-production study does not prove that the same index predicts a particular company’s sales. Morrow’s planning principle is therefore to test candidate indicators against the company’s own history, consider lead time and economic logic, and keep monitoring performance. The study supports disciplined validation, not a guarantee that turning points will always be detected.
Want to know whether external indicators can improve your company’s forecast?
Want to know whether external indicators can improve your company’s forecast?
More to Discover
When Leading Indicators Work—and When They Fail
When Leading Indicators Work—and When They Fail
A German industrial-production study compares economic indicators and identifies periods when every tested signal weakened. This is why validation matters.
A German industrial-production study compares economic indicators and identifies periods when every tested signal weakened. This is why validation matters.
Forecast Validation
Why a leading indicator needs proof
A series can sound economically relevant without consistently moving ahead of the outcome you care about. This paper examined that problem for German industrial production from 1991 to 2016. It compared survey-based ifo and ZEW indexes with order-flow and unemployment data.
How the authors checked the signals
The researchers used a leading–lagging method to assess which series tended to move before industrial production and how those relationships changed over time. This is not itself a sales-forecasting model. It is a way to judge candidate indicators and spot periods when their apparent lead becomes weak or unreliable.
A lead can weaken without disappearing
An indicator can still move broadly with industrial activity while offering less advance notice than before. For a planning team, that matters because a signal that arrives after orders have already slowed adds little warning. Reviewing its timing across several upturns and downturns can help distinguish a useful lead from a relationship driven by one period.
How a company can assess its own signals
Suppose a company uses an industry confidence index to review its demand assumptions. It can compare the index with its own sales cycle and examine the dates when their directions changed. If the lead was useful in some periods and inconsistent in others, the team can look for confirmation from indicators that capture different parts of its market.
What the study observed
In the authors’ analysis, the ifo and ZEW sentiment indexes had the strongest overall leading relationship with German industrial production. Recession windows were easier to anticipate than recovery windows. Around 1997 and 2005, all tested indexes performed weakly—an important reminder that an indicator is not permanently reliable.
What this means for business planning
One national industrial-production study does not prove that the same index predicts a particular company’s sales. Morrow’s planning principle is therefore to test candidate indicators against the company’s own history, consider lead time and economic logic, and keep monitoring performance. The study supports disciplined validation, not a guarantee that turning points will always be detected.
Want to know whether external indicators can improve your company’s forecast?
Want to know whether external indicators can improve your company’s forecast?
More to Discover
When Leading Indicators Work—and When They Fail
When Leading Indicators Work—and When They Fail
A German industrial-production study compares economic indicators and identifies periods when every tested signal weakened. This is why validation matters.
A German industrial-production study compares economic indicators and identifies periods when every tested signal weakened. This is why validation matters.
Forecast Validation
Why a leading indicator needs proof
A series can sound economically relevant without consistently moving ahead of the outcome you care about. This paper examined that problem for German industrial production from 1991 to 2016. It compared survey-based ifo and ZEW indexes with order-flow and unemployment data.
How the authors checked the signals
The researchers used a leading–lagging method to assess which series tended to move before industrial production and how those relationships changed over time. This is not itself a sales-forecasting model. It is a way to judge candidate indicators and spot periods when their apparent lead becomes weak or unreliable.
A lead can weaken without disappearing
An indicator can still move broadly with industrial activity while offering less advance notice than before. For a planning team, that matters because a signal that arrives after orders have already slowed adds little warning. Reviewing its timing across several upturns and downturns can help distinguish a useful lead from a relationship driven by one period.
How a company can assess its own signals
Suppose a company uses an industry confidence index to review its demand assumptions. It can compare the index with its own sales cycle and examine the dates when their directions changed. If the lead was useful in some periods and inconsistent in others, the team can look for confirmation from indicators that capture different parts of its market.
What the study observed
In the authors’ analysis, the ifo and ZEW sentiment indexes had the strongest overall leading relationship with German industrial production. Recession windows were easier to anticipate than recovery windows. Around 1997 and 2005, all tested indexes performed weakly—an important reminder that an indicator is not permanently reliable.
What this means for business planning
One national industrial-production study does not prove that the same index predicts a particular company’s sales. Morrow’s planning principle is therefore to test candidate indicators against the company’s own history, consider lead time and economic logic, and keep monitoring performance. The study supports disciplined validation, not a guarantee that turning points will always be detected.