Markit vs ISM – Which US Manufacturing Survey is Right?

The title describes the thoughts of those that closely follow US manufacturing data. While other manufacturing surveys have declined, the headline ISM continues to hold fairly steady in the high 50s, low 60s. However, the optimism shown by the headline ISM is currently out of line with the message being told by the market.

In this post I will discuss:

  • Measuring the Impact of the US Fiscal Stimulus
  • Manufacturing Data and Signs From Market Leadership
  • Cracks Developing in Semiconductors

While I tend to reference ISM Manufacturing the most, the current environment is why I analyze numerous data points when building my fundamental view, which hasn’t changed despite ISM Manufacturing reaching a new economic cycle high.

Measuring the Impact of the US Fiscal Stimulus

As I mentioned a few months ago, the US government increased discretionary spending when the budget was approved in February. When thinking about its impact on the economy, it is important to remember that the government’s fiscal year ends September 30th. Therefore, the net change in discretionary spending for 2018, $220Bn, is actually impacting the US economy in only a six month period. Said differently, the net change could potentially stimulate the US economy by approximately 2.2% of GDP in this compressed time frame.

https://www.cbo.gov/system/files/115th-congress-2017-2018/reports/53696-sequestration.pdf

When looking at the results of Q2 GDP and the estimate for Q3 from the Atlanta Federal Reserve, we see how the growth rate has significantly increased since the fiscal stimulus was passed.

Source: Trading Economics

Putting this into a historical perspective, the increase in Discretionary Spending as a percentage of GDP is rather unprecedented, especially this far into an economic cycle.

Historically, fiscal stimulus tends to not have lasting effects on the economy once it is reversed. As stated by Richard Koo (former economist at the Federal Reserve Bank of New York and currently the Chief Economist at Nomura) in a Real Vision interview published October 7, 2016:

“Of course, whenever you put in a fiscal stimulus, the economy responds. But when it’s removed, it tanks again. It doesn’t have this pump priming function that we expect from fiscal stimulus.”

Keeping this in mind, and knowing that the US Fiscal Stimulus reverses on October 1st, watching for signs of the economy slowing in the leading economic indicators and the market becomes critically important.

Manufacturing Data and Signs From Market Leadership

As stated in the title of this post, there are two different stories being told by the US manufacturing data. First, the Markit US Manufacturing PMI slowed for its fourth consecutive month.

Markit US Manufacturing PMI


Source: Trading Economics

However, the headline number for ISM Manufacturing came in at an economic cycle high. One important note though is that New Orders increased but stayed below the January peak, which is similar to what we saw in 2H14.

ISM Manufacturing (white), ISM Manufacturing New Orders (yellow)


Source: Bloomberg

When there is conflicting data, my first inclination is to see what the market is telling us. As I’ve been showing for months, cyclicals have been underperforming counter-cyclicals.

ISM (yellow), SOX (Semiconductors) Index vs Health Care (white)


Source: Bloomberg

ISM (yellow), Industrials vs Staples (white)


Source: Bloomberg

ISM (yellow), Financials vs Utilities (white)


Source: Bloomberg

I have only shown the sector rotation previously with the large caps since they are large, stable, and more established. However, we know the global economy has been slowing for months and large caps are more exposed to these economies. To get a more US specific view, we should look at the small caps, which are showing us the exact same picture as the large caps, a change in leadership to the counter-cyclicals.

ISM (yellow), S&P 600 Semiconductors vs S&P 600 Health Care (white)


Source: Bloomberg

ISM (yellow), S&P 600 Industrials vs S&P 600 Staples (white)


Source: Bloomberg

ISM (yellow), S&P 600 Financials vs S&P 600 Utilities (white)

Source: Bloomberg

With large and small caps both telling the same narrative, an expectation of slowing growth in the future, I put more weight on the Markit data point than ISM.

Cracks Developing in Semiconductors

Semiconductors are arguably the best and most consistent industry to track in regards to the business cycle. At the beginning of the cycle, they tend to trough slightly before or with the other cyclicals. More importantly to the current situation, they tend to be one of the last cyclical industries to decline.

Over the past twenty years, once the Semis begin to show weakness, the S&P 500 tends to peaks shortly thereafter. This of course leads to the question of, how do you measure weakness in the Semiconductors?

SOX Index (white), S&P 500 (yellow) (1999-2007)


Source: Bloomberg

SOX Index (white), S&P 500 (yellow)(2009- 2018)


Source: Bloomberg

As discussed in April (https://wp.me/p9vaFZ-4J), the market tends to look for growth when the economy is expanding and stability when the economy is peaking and slowing. Additionally, in that post I showed how Texas Instruments (TXN) tends to outperform Micron (MU) as we get into the later stages of the business cycle because it has a more stable business. Therefore, the recent outperformance of TXN vs MU should not be surprising.

Texas Instruments (TXN) (white), Micron (MU) (yellow), Applied Materials (AMAT) (green)


Source: Bloomberg

I bring this up because we are now seeing the price of TXN decline. Typically, once TXN begins to break down, the SOX Index isn’t far behind.

Texas Instruments (TXN) (white), SOX Index (yellow)


Source: Bloomberg

Turning back to the SOX Index, or SOXX ETF, there are very clear areas of importance on relative and absolute basis.

The SOXX ETF relative to the S&P should show weakness (or underperformance) before it does on an absolute basis. We see that the SOXX couldn’t make a new high relative to the S&P in early June and since then it has been making a series of lower highs. Once the SOXX/S&P break the lows from the past nine months, then the focus turns to the SOXX on an absolute basis.

SOXX ETF vs S&P 500


Source: Bloomberg

On an absolute basis, the SOXX ETF is still in the middle of the range that it has been in for almost a year. However, if the underperformance vs the S&P 500 deteriorates, I would expect the SOXX ETF to begin to challenge ~$166.

SOXX ETF


Source: Bloomberg

Finally, once ~$166 is broken, then the concern for the S&P peaking begins immediately.

Summary

With the fiscal stimulus set to reverse on October 1st, market leadership showing an expectation for a decline in the future growth of the US economy, and the Semiconductor Industry showing weakness, it’s tough for me to continue to recommend being long the S&P 500. Maybe there’s 3-5% of upside left in the S&P 500 before it finally peaks, but I’m growing more concerned about the risk vs reward of being overweight with the negatives growing.

When I look at positioning, as shown by freecotdata.com (@movement_cap), we continue to see that investors, or speculators, are more exposed to the long side with the S&P 500 and Russell 2000. This leads me to the conclusion that market participates are currently not positioned for the scenario that I’ve discussed.

Commitment of Traders Report: S&P 500


Source: freecotdata.com, @movement_cap

Commitment of Traders Report: Russell 2000


Source: freecotdata.com, @movement_cap

Due to my concerns, my recommendation for the S&P 500 has moved to Neutral (or 50% exposure). Since I expect the counter-cyclicals to outperform the cyclicals, I continue to recommend increasing exposure to these sectors and industries while decreasing the overall exposure to US equities.

 

Currencies

  • Overweight (or long) the USD
  • Underweight (or short) Emerging Market currencies

Bonds

  • Overweight (or long) long dated maturing US debt
  • Underweight (or short) Emerging Market debt

Equities

  • Neutral (or 50% exposure) S&P 500
  • Underweight (or short) Emerging Markets
  • Overweight (or long) US counter-cyclical sectors and industries
  • Underweight (or short) US cyclical sectors and industries