Why "Consumer Confidence" Is The Most Manipulated Economic Indicator

The propaganda machine's favorite subversive mechanism to instill confidence in the prevalent population, is by nudging none other than the Confidence Index itself. And indeed, over the past few months, the Conference Board and UMichigan indices have been on a straight upward slope. What better way to get Joe and Jane Sixpack out to the mall to spend, than to give them the impression that all their neighbors are out, doing just that (and buying 5 iPads in the process). Don't forget the US economy is based on reckless spending and credit, and the Sixpacks love that, especially if they don't have to pay their mortgage ever again (as Obama told them to do). Yet the Consumer Index is nothing but just another massive scam. To wit: observe the ABC Consumer Comfort index, a representation of consumer "confidence" which according to many is much more detailed, due to its larger sampling size, and more up to date, due to its weekly updates, yet one which, refusing to drink the Kool Aid has never once been mentioned on the biggest propaganda channel of all, CNBC. Yesterday, the latest ABC numbers came out, and they were a stunner: at -49, the index is barely a few points away from all time record lows, and well into deep recession territory. In fact, the only confidence consumers can have, is that there is absolutely nothing predictive about the CONsumer indices out there: the spread between the ABC and the administration darling's Conference Board just hit a record high, as see on the chart below. In fact, it has gotten so difficult to sway popular opinions with these simplistic methods of subliminal control, that even Goldman Sachs is concerned about the utility of such indices going forward. Goldman's Jan Hatzius, in discussing Consumer Confidence, notes: "Since the monthly spending figures are noisy estimates of the true underlying trend, it does make sense to put some weight on confidence alongside actual spending when trying to gauge the outlook for consumption.  This still points to a U-shaped rather than V-shaped recovery in spending." If after reading this, you still think that consumer indices are indicative of anything at all you probably were one of those who did buy 5 iPads over the past month.

First, we present the record divergence between the Conference Board and the ABC Consumer Comfort index:

As we said: mere propaganda. But don't take our word for it: here is Goldman's very own Jan Hatzius:

Which Way for Consumer Confidence? (Hatzius)
 
•         The April consumer confidence report from the Conference Board came in ahead of expectations, rising to the highest level since September 2008.  However, other, alternative measures of confidence—the University of Michigan index and the ACS/Washington Post index—have fallen this month.  So is confidence rising or falling?
 
•         A statistical “principal components” analysis of these three confidence indexes finds a small decline in confidence in April relative to March.  Over a longer period, confidence has been essentially unchanged since last September, though it is up substantially from the lows of late 2008 and early 2009.
 
•         There are at least two possible explanations for the gap between healthy spending growth and still-sluggish confidence.  First, the weight of higher-income consumers—who seem to be doing considerably better recently—is larger in the spending figures than in the confidence indexes.  Second, and nevertheless, the confidence data suggest that the rebound in spending is still likely to be U-shaped rather than V-shaped.
 
Today’s April consumer confidence report from the Conference Board came in ahead of expectations.  The overall index rose 5.4 points to 57.9, the highest level since September 2008, with present conditions up 3.4 points and expectations up 7 points.  Does this mean that the consumer has turned the corner?
 
There are certainly a number of encouraging signs.  Most importantly, the consumer spending data themselves have significantly beaten expectations in recent months.  In the first quarter as a whole, we estimate that real consumption rose 3¾% (annualized), and this follows better-than-expected data in both the third and fourth quarter of 2010.  Similarly, more bottom-up indications from retail companies and consumer-exposed financials have also been looking significantly firmer.
 
However, the verdict on the consumer confidence data themselves is still considerably more equivocal.  This is not just because the Conference Board data are noisy—after all, two months ago the question was how much weight we should put on the sharp 10.1-point decline in the index from January to February.  And it is not just because confidence is still far below the long-term average of 96.  It is also because the pickup in the Conference Board measure is at odds with other measures of the consumer’s mood.  The (preliminary) consumer sentiment index from the University of Michigan dropped 4.1 points in April from March’s final reading, and the monthly average of the weekly ABC/Washington Post index edged down 2.0 points as well.
 
So which indicator should we believe?  Since most economic indicators are a noisy estimate of the underlying reality, it typically makes sense to put some weight on each measure.  In order to do this, we have performed a “principal components analysis” of the three indicators.  The “first principal component” is the data series that best represents the variation over time in all three of our indicators over time, and we can therefore think of it as an optimal combination of the information contained in all three series.
 
The table below shows all three series plus the first principal component from our analysis.  The principal component series currently stands at -2.7, above the lows seen in late 2008 and early 2009 but basically unchanged since last September and still at a very low level in absolute terms (similar to the trough of the recession of the early 1990s).

Alternative Measures of Consumer Confidence

 

 

Conference Board

U Michigan

ABC/WaPo

1st Principal Component

2008 - Jan

87.3

78.4

-24

-1.0

2008 - Feb

76.4

70.8

-36

-2.0

2008 - Mar

65.9

69.5

-31

-2.1

2008 - Apr

62.8

62.6

-39

-2.7

2008 - May

58.1

59.8

-48

-3.2

2008 - Jun

51.0

56.4

-44

-3.4

2008 - Jul

51.9

61.2

-43

-3.1

2008 - Aug

58.5

63.0

-49

-3.1

2008 - Sep

61.4

70.3

-43

-2.5

2008 - Oct

38.8

57.6

-48

-3.7

2008 - Nov

44.7

55.3

-52

-3.8

2008 - Dec

38.6

60.1

-50

-3.7

2009 - Jan

37.4

61.2

-51

-3.7

2009 - Feb

25.3

56.3

-50

-4.2

2009 - Mar

26.9

57.3

-48

-4.1

2009 - Apr

40.8

65.1

-48

-3.4

2009 - May

54.8

68.7

-45

-2.8

2009 - Jun

49.3

70.8

-50

-2.9

2009 - Jul

47.4

66.0

-50

-3.2

2009 - Aug

54.5

65.7

-46

-3.0

2009 - Sep

53.4

73.5

-47

-2.6

2009 - Oct

48.7

70.6

-49

-2.9

2009 - Nov

50.6

67.4

-46

-3.0

2009 - Dec

53.6

72.5

-45

-2.6

2010 - Jan

56.5

74.4

-47

-2.5

2010 - Feb

46.4

73.6

-49

-2.8

2010 - Mar

52.3

73.6

-45

-2.6

2010 - Apr

57.9

69.5

-47

-2.7

 

 Source: Haver Analytics.

So which indicator should we believe?  Since most economic indicators are a noisy estimate of the underlying reality, it typically makes sense to put some weight on each measure.  In order to do this, we have performed a “principal components analysis” of the three indicators.  The “first principal component” is the data series that best represents the variation over time in all three of our indicators over time, and we can therefore think of it as an optimal combination of the information contained in all three series.
 
The table below shows all three series plus the first principal component from our analysis.  The principal component series currently stands at -2.7, above the lows seen in late 2008 and early 2009 but basically unchanged since last September and still at a very low level in absolute terms (similar to the trough of the recession of the early 1990s).