Sentiment data can be valuable when assessing the current conditions of an economy. Such kind of data are often surveys that capture the behavior and outlook of a particular group (like consumers) or a sector (like manufacturing or small businesses) in an economy. Intuitively one can see value in sentiment data – it allows for economists and investors to gauge what future hard data will be. After the Presidential election, sentiment data began to rise, largely due to the new administration’s “America First” pledge, which signaled tax cuts for the middle class and businesses, fighting unfair global trade, and rebuilding American roads and bridges.
After the election, several survey-based data began to rise, like the ISM Manufacturing PMI, the Small Business Optimism Survey, and the University of Michigan’s Consumer Sentiment and Expectations indices. All of these surveys jumped from October to November, sending a signal to the market.
As the market responded to the Trump administration’s policy goals, i.e. fiscal stimulus, actual hard data remained slow to react. For example, the ISM Manufacturing PMI jumped from an index of 52 in October 2016 to 57.2 in March 2017, while the NFIB Small Business Optimism Survey increased from 94.9 to 104.7 during the same period. The hard data, on the other hand, indicated a mixture of results. New orders, which is a signal for business demand, recovered from -4.0 percent year-on-year growth in October 2016 to 2.9 percent annual growth in February of this year. However, overall production, which should theoretically increase along with new orders if demand is high, remained stagnant. October’s industrial production index actually fell -0.76 percent, just prior to the Presidential election. By February 2017, industrial production only increased 0.45 percent year-on-year.
The same can also be said with the University of Michigan’s Consumer Sentiment and Expectations indices. Consumer sentiment was 87.2 in October 2016 and by March reached 96.9. Consumer expectations saw a similar jump from 76.8 in October to 86.7 in March. With such a rise in both indicators, one would expect that consumers are willing to spend. However, that has not entirely been the case.
Exhibit 1. University of Michigan’s Consumer Sentiment Index Source: Reuters
The US economy is largely centered on consumer’s ability to spend, more importantly to spend using credit. That means that large investments like homes and automobiles are two important sectors for economic growth. As many may have expected home and auto sales to increase during this recent uptrend in sentiment, hard data shows mixed results. Existing home sales actually fell when comparing data from October 2016 to February of this year – home sales in October were 5.53 million versus 5.48 million in February. New home sales actually increased from October’s 568 thousand to 592 thousand in February, but new home sales are still far below pre-2008 levels and are still short of 2016’s high of 622 thousand in July. Auto sales also disappointed from October 2016 to March 2017, with sales declining from 17.9 million to 16.53 million, respectively.
Exhibit 2. Economic indicator chart of selected soft and hard data Source: Reuters
If sentiment among consumers and businesses is high, why are Americans not increasing spending on more expensive, long term assets, and why is production not growing at a faster rate than before? There is no certain answer, but there are a few assumptions that we can make about survey data. First, surveys can be unreliable, and, therefore, not accurately reflecting what we should expect when hard economic data is released. Many people simply follow the obvious signs of economic trends, which may lead to overly positive responses in expansionary periods or overly pessimistic responses in contractionary periods. Also, sentiment surveys are often behavioral-based in their questions, rather than data specific. For example, the University of Michigan’s Index of Consumer Sentiment is derived from five questions ranging from how financially sound the survey participant is to how much better or worse off they expect to be in the future to whether it is a good time for people (not the respondents) to buy major household items. The ISM manufacturing PMI does a slightly better job with their survey because there are questions about new orders from customer and inventories, but the responses are not exact, with managers only able to response with “better”, “same”, or “worse”.
Second, expectations of significant policy changes in the near future could have also skewed the responses, translating to more positive answers in the surveys. As with any election season, candidates make promises to their constituents in order to receive votes. President Trump promised tax relief for the middle class and the business sector, as well as promising to boost infrastructure spending. This could have increased future expectations in the manufacturing sector and consumer goods space. If manufacturers believe that policy makers will add protections to their business and consumers believe that there will be lower taxes in the future, there is incentive for the participant to respond favorably.
Another possible source for the divergence is the political divide. The University of Michigan also collects political affiliation data along with the responses from the participants, and the conclusions are fascinating . Republican expectations for the economy surged from an index of 70.9 in June 2016 to 121.4 in December 2016. Democrats were far more pessimistic, with the index falling from 91.9 in June to 67.8 in January. (Respondents who identified as independents had a favorable outlook on the economy, with the index rising from 74.6 to 88.7 in the same period.) Altogether, consumer expectations diverged by 74.6 points – the widest divide in the history of the index. Therefore, part of the reason for the rise in the soft data could be due to political excitement from the Republican base.
Exhibit 3. A Comparison of the University of Michigan’s Index of Consumer Expectations and Index of Consumer Sentiment Source: Survey of Consumers: The University of Michigan
The markets have also been very optimistic that more friendly business policies will take effect. The top three US stock market indices (S&P 500, Dow Jones, and Nasdaq) all set all-time record highs in mid-February, rising over 10 percent since the election. However, as the Trump team settled into the White House and began having trouble fulfilling campaign promises , expectations have fallen.
The S&P 500 fell over 1.7 percent from the end of February to April 13 (the time of writing), and the Dow Jones fell over 1.8 percent during the same period. The yield curve has also shown signs of flattening, with the 10-year yield falling from 2.36 at the end of February to 2.24 at the time of writing (April 13). Inflation expectations, which jumped post-election, has flattened, with the 10-year breakeven inflation rate falling marginally in February at 2.02 versus 1.92 on April 13. It is clear that the market has cooled on such high expectations of serious economic reforms, but what does this mean for the economy moving forward?
With the hard data being slow to catch up to sentiment data and market indicators cooling from election expectations, it is vital that the economy be analyzed using a wide lens. The United States is still expanding, albeit at the slow pace it has seen since 2010, and the other hard economic data suggests the same. With the current divergence between soft and hard data, we believe that early indicators are less reliable and should be analyzed, and confirmed, together with the hard data as it is released.