Sentiment is Bottoming, the Pain Trade Will Be Higher Equity Prices

Three months ago (https://wp.me/p9vaFZ-81) I recommended reducing allocations to equities based upon risks that were growing in the market. While my fundamental view hasn’t changed, sentiment is beginning to bottom. Based upon history, when a change in sentiment occurs after a period of significant stress, there is a high probability for a strong counter-trend move to take place regardless of fundamentals.

In this post I will discuss:

  • Quick Significant Declines in Market Multiples
  • US Recession? Short Answer… No
  • Measuring Sentiment with Volume and Volatility

A change in sentiment can quickly gather steam as investors rapidly change from fear to a fear of missing out (FOMO). When comparing the current development to similar periods, the probability for a 15% move (+/- 5%) is high enough that I’m increasing my recommended allocation to Equities and reducing the strong USD exposure.

Quick Significant Declines in Market Multiples

In the market, there are two items that determine price: fundamentals and the multiple. I look at the multiple as a way to view the market’s opinion, or sentiment. Currently, most market bears have been focused on the multiples being high. However, just because multiples are high doesn’t mean that the market is destine to decline. Said differently, the level of the multiple is never a catalyst for the market to increase or decrease.

Specifically looking at Enterprise Value to the last 12 months of EBITDA (EV/TTM EBITDA), which has not been impacted by the lower corporate tax rate, we see that it is near 30 year highs. This simply means that the market believes that each dollar of revenue, gross profit, operating income, etc is more valuable than it was previously. Investors have to decide whether they want to fight the market, which is a losing battle, or embrace the valuation being applied by the market.

EV/TTM EBITDA

Source: Bloomberg

Since I’ve already said that valuation is never a catalyst for the market to move, what I look for are sharp sudden moves in valuation. Historically, when moves of about 10% or greater occur to the downside, the market has been higher three months later unless we’re in a recession, which is a key point. These moves have been noted in the next three charts.

1990 – 2001: EV/TTM EBITDA (black), S&P 500 (red)

Source: Bloomberg

2003 – 2008: EV/TTM EBITDA (black), S&P 500 (red)

Source: Bloomberg

2010 – 2018: EV/TTM EBITDA (black), S&P 500 (red)

Source: Bloomberg

The current move of 11-12% is a rather significant decline compared to the last 30 years. Additionally, it’s worth noting that Emerging Markets have seen a multiple decline typically only seen during a US Recession.

EV/TTM EBITDA (black), S&P 500 (red), US Recession (blue outline)

Source: Bloomberg

US Recession? Short Answer… No

In the eyes of the market there is a big difference between being in a recession and the economy readjusting to a lower growth rate, which is what I believe we are seeing. In September (https://wp.me/p9vaFZ-81) I discussed how the fiscal stimulus would end that month and detailed how the market was showing signs of adjusting and pricing this in.  Furthermore, in October (https://wp.me/p9vaFZ-8w) I detailed how the economic data continued to peak and decline. 

Looking at a very broad view of the US economy, ISM Manufacturing and Employment are not showing signs of the US being in a recession.

Starting with the leading indicator of ISM Manufacturing, New Orders, while the trend is lower the readings are far too high to think that we’re in a recession.

ISM Manufacturing New Orders

Source: Bloomberg

The same is true when looking only at the headline number as well. 

1990 – Now: ISM Manufacturing (green), S&P 500 (red), US Recession (blue)

Source: Bloomberg

1960 – 1990: ISM Manufacturing (green), S&P 500 (red), US Recession (blue)

Source: Bloomberg

Turning to the employment data, we see the unemployment rate (U-3) is very low relative to the past 60 years. Applying a similar thinking to the multiples on the S&P, just because it’s low doesn’t mean that it has to turn higher. 

1960 – Now: Unemployment Rate, U-3 (red), Recession (blue)

Source: Bloomberg

Also, we always see Continuous Jobless Claims Y/Y turn higher well before U-3. Currently, CJC isn’t turning higher so the probability of U-3 sustainably turning higher very low.

Source: St Louis Federal Reserve, FRED

To summarize, yes, the economy is slowing but we are definitely not in a recession and not going into one in the next three months. Just a reminder, the US is not a boom bust economy like it was in the 60s and 70s so a recession takes much longer to develop in a consumer driven and more stable economy.

Measuring Sentiment with Volume and Volatility

In the next three charts, I’m seeing the market price in and adjust to the uncertainties in the economy. In August (https://wp.me/p9vaFZ-7t) I detailed how I use Volume and Volatility so I’ll refrain from repeating it.

With Volume, history shows that volume tends to slowly rise with growing uncertainty and decline once the uncertainty is behind it or less of a concern. An important point to keep in mind, Volume doesn’t gradually move higher once we’ve seen a significant spike like we did in October. Currently, we are seeing less volume in the market than in October. Similar to 2010 and 2011, it hasn’t declined much but we are seeing lower highs.

Total Market Volume

Source: Bloomberg

When we look at similar periods of declining Volume on a chart of the 2ndmonth futures contract of Volatility, we see it stabilizing once more at an elevated level. During periods of high uncertainty, we tend to see this pattern until the uncertainty has been priced in.

Volatility – Continuous 2ndMonth Futures Contract

Source: Bloomberg

Additionally, after the large increase in market stress in early October, the Volatility Spread has been putting in higher lows. While still negative, and therefore showing signs of high stress, the signal is that the Spread is becoming less negative as the uncertainty is becoming less of a concern for the market.

Current Chart: Generic 2-1 (blue, right), S&P 500 (red, left)

Source: Trading View

Comparing this to similar economic periods when growth (i.e. ISM Manufacturing) was peaking or slowing, Volume was declining after a large increase, and 2ndMonth Volatility was stabilizing, we see similar patterns in the Volatility Spread.

Early 2018: Generic 2-1 (blue, right), S&P 500 (red, left)

Source: Trading View

2014-2015: Generic 2-1 (blue, right), S&P 500 (red, left)

Source: Trading View

Mid – 2010: Generic 2-1 (blue, right), S&P 500 (red, left)

Source: Trading View

Early 2007: Generic 2-1 (blue, right), S&P 500 (red, left)

Source: Trading View

Mid – 2006: Generic 2-1 (blue, right), S&P 500 (red, left)

Source: Trading View

Finally, an indicator that is used commonly is the VIX/VIX3M (i.e. Spot VIX divided by the 3 Month Implied Volatility). As you can see on the next chart, the level the ratio reached on Thursday morning has only been exceeded five times since 2005. This is a very good example of the high levels of pessimism currently in the market.

VIX/VIX3M (white), Thursday AM reading of 1.18 (red line)

Source: Bloomberg

Since the market is still seeing uncertainty, the market has remained under high stress. However, the uncertainty and stress is waning.

Summary

There are three risks to my overall view of sentiment bottoming. 

First, I’m too early, which very well could be the case. There is plenty of headline risk at the moment with the UK vote and ECB meeting this week, the Fed meeting next week, Mueller-Trump, etc. The way that I’m handling this is by gradually reducing my shorts and adding to my longs on days of weakness. As I start to see the market become less stressed then I’ll continue to make the transition with the remaining segments of the portfolio.

Another risk is that this bounce is similar to the bounce in 2014-2015, which was less than 10% and only lasted three months. Since I’m not seeing a driver for fundamentals to change its longer term trend, the probability of this occurring is higher than I would like. However, the market multiple has declined much more now than in 2014-2015.

The third is that I’m reading the market wrong and the bears have every right to be excited. Since the probability of being in a recession now or in the next three months is low, then the probability of the market behaving like we’re in a recession is low.

All together, my view on sentiment leads me to change my recommended allocation. Since the USD has been strong for the past 8 months, a pullback should be expected in a strong counter-trend rally. Without a change in the fundamentals, I’m only changing to Neutral. Since I’m changing my view on the USD, then I am also changing my recommended exposure for Emerging Market Bonds and Equities to Neutral.

Currencies

  • Neutral (or 50% exposure) USD
  • Neutral (or 50% exposure) Emerging Market currencies

Bonds

  • Overweight (or long) US debt 
  • Neutral (or 50% exposure) Emerging Market debt 

Equities

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

Update on Stress in the Market

Two weeks ago I described how I analyze stress in the market (https://wp.me/p9vaFZ-3z). Based upon previous periods, Friday, March 2nd was not the end of the market stress we have witnessed over the past month. In this post I’ll make the case for my analysis by:

  • Analyzing characteristics in recent times of stress (2015 and 2016)
  • Comparing the previous periods to what occurred on Friday, March 2nd
  • Current outlook and expectation

The move in the VIX spread continues to tell me that we have not transitioned from the turbulent times of the past month. Also, the recent lack of improvement in the spreads leads me to believe we have more risk ahead of us.

Analyzing Recent Times of Stress (2015 and 2016)

When comparing previous moves in the spread during times of stress, March 2nd was not the bottom for the market.

When looking at what occurred from August – September 2015, the move on Monday, September 28th was a large move to the downside (i.e. a move with conviction). A move like this is typically associated with margin calls (i.e. forced sellers), which leads to a “flush” of the market to the downside and then a fast recovery higher.

S&P 500 (yellow), VIX Contract X 15 – VIX Contract V 15 (white)
Source: Bloomberg

A similar type of move can be seen in February 2016. In this time period we had a change in the front month contract, which complicates historical analysis. Also, by not having intraday data on the chart, it doesn’t show that at the worst point of February 15th, the market was down ~2.5% from the close on February 14th.

S&P 500 (yellow), VIX Contract G 16 – VIX Contract F 16 (white), VIX Contract H 16 – VIX Contract G 16 (white)
Source: Bloomberg

When there is a change in the contracts, as with 2016, it truly shows how important it is to analyze the individual contracts during times of increased stress instead of relying on continuous, or generic, contracts.

S&P 500 (yellow), Generic 2nd Month Contract (UX2) – Generic 1st Month Contract (UX1) (white)
Source: Bloomberg

Comparing the Previous Periods to What Occurred on Friday, March 2nd

When I compare 2015 and 2016 to what occurred on Friday, March 2nd, I see an orderly move lower in the spreads compared to the “flush” in the previous periods.

The first chart shows the 2nd month minus the 1st month ending on March 2nd. The move that began five days earlier descended in a linear fashion, which is not indicative of a “flush” or forced selling.

S&P 500 (yellow), VIX Contract J8 – VIX Contract H8 (white)
Source: Bloomberg

The same decline can be seen on the next chart showing the 3rd contract minus the 2nd contract.

S&P 500 (yellow), VIX Contract K8 – VIX Contract J8 (white)
Source: Bloomberg

To summarize, the lack of the “flush” last Friday has me believe that we have yet to see the bottom in the spreads, and the market, for the near term.

Current Outlook and Expectation

When looking at the charts of the spreads, we are once again seeing them stagnant, which is a concern. The concern is heightened because we have peaked at a lower high while still in backwardation.

S&P 500 (yellow), VIX Contract J8 – VIX Contract H8 (white)
Source: Bloomberg

S&P 500 (yellow), VIX Contract K8 – VIX Contract J8 (white)
Source: Bloomberg

As shown in the February post “Update on Macro and Markets” (https://wp.me/p9vaFZ-2T), we know volatility tends to first pick up when there is a change in market leadership (i.e. cyclical to defensive). Knowing this, I continue to closely watch the Euro and Oil.

It seemed that during the first bout of volatility, market participants began to question their positions since they are no longer making new highs.

EUR/USD (white), Oil (yellow)
Source: Bloomberg

However, positioning has only declined slightly. Therefore, these are still very crowded trades.

EUR/USD Commitment of Traders (COT)
Source: freecotdata.com (@movement_cap)

WTI Commitment of Traders (COT)
Source: freecotdata.com (@movement_cap)

In order for the market to complete the change in leadership from cyclical to defensive, the positioning in these trades will need to change.

Whether it occurs in the next bout of market stress or the one after that, I can’t be sure. However, with global manufacturing declining for the second consecutive month, it seems that we are getting closer to this occurring.

Source: Markit, JPM

Measuring the Level of Stress in the Market

According to WedMD, there are two types of stress:

  • Acute (temporary) stress, which your body recovers from quickly
  • Chronic (long term) stress, which can lead to serious health problems

Applying this rational to market declines:

  • Acute stress: no recession is imminent and the market should recover its losses
  • Chronic stress: recession risk is high and a large decline in equities is expected

Currently, the market is going through a higher level of acute stress. Please note, I do mean, “going through” because the market has not yet exited this period of increased stress.

In this post:

  • Process for measuring the level of stress in the market
  • Historical analysis of the VIX spread
  • The current level of stress in the market
  • Longer term expectation of volatility

Process for Measuring the Level of Stress in the Market

I measure the amount of stress in the market by analyzing the VIX futures curve, specifically the 1st and 2nd month contracts. On Bloomberg, these would be UX1 and UX2 for the generic (continuous) contracts.

Looking at the chart below, I would use the following checklist to analyze the spread between UX2 and UX1:

UX2-UX1 (white), S&P 500 (yellow)

Source: Bloomberg

ISM Manufacturing

Source: Bloomberg

Historical Analysis of the VIX Spread

The following is a brief description of what has occurred since the VIX futures began trading in 2004.

Keep in mind, when bouts of stress appear in the market, the periods of high stress have occurred after ISM has peaked and is declining. There are occurrences of low stress periods when ISM is increasing but the market tends to absorb these incidents easier. I don’t show nor discuss the direction of ISM in the following analysis so please reference the previous chart.

March 2004 – January 2006

After a volatile start to trading, the spread peaked in August and gradually decreased until reaching backwardation in March 2005.

The spread recovered until July 2005 when it began to decline once more until October 2005. October was also the first time we saw a retest in the lows of the spread a few weeks later, which we typically see in higher stress periods.


Source: Bloomberg

December 2005 – January 2008

After peaking in December, spreads declined until reaching a low of -2.00 before retesting the lows of the spread eights weeks later. This marked the first high stress period for the market during the decline of ISM.

After recovering, the spread decreased for another ten months before falling to -4.00. The spikes in the spread seen in September and October 2007 are examples of the problem associated with only using continuous futures in high stress periods. More on this later.


Source: Bloomberg

January 2007 – October 2008

After recovering from August 2007, the spread stayed at or near 0.00 until March 2008. The higher lows seen in January and March 2008 and again in July 2008 showed that the market was beginning to heal despite the S&P making new lows. However, this all changed September 2008 with Lehman Brothers.


Source: Bloomberg

October 2008 – January 2010

After seeing the large decline in the spread in October 2008, the stress in the market slowly began to dissipate and grind its way out of backwardation. If an investor was only looking at the price of the S&P, they would have missed the healing process the market was undertaking.


Source: Bloomberg

January 2010 – January 2012

Exiting the recession, we had a short-term scare with the spread declining to -1.00 before seeing a much larger decline in 2011 associated with the US Debt Ceiling Crisis and the European Banking Crisis.


Source: Bloomberg

November 2011 – January 2014

This time frame had little to no stress, similar to what the market went through from February 2016 until January 2018. The area circled is the Taper Tantrum.


Source: Bloomberg

July 2013 – June 2015

This time frame showed more stress in the spread than the previous period but again, nothing strenuous. October 2014 was the US Treasury Flash Crash, which took place only a few weeks after Mario Draghi announced that the ECB was going to begin Quantitative Easing (QE). Additionally, we had the Russian Ruble Crisis peak January 2015.


Source: Bloomberg

March 2015 – May 2017

The spread gradually moved lower from the March highs before seeing a tremendous drop after China devalued the Yuan. A few months later, we had another decline in the spread when there was a growing concern about the stability of the European Banks. Finally, the decline in June 2016 was when the UK voted to leave the EU.

An important note is that each decline in the spread was not as low as the previous (i.e. Aug ’15 > Sept ’15 > Jan ’16 > June ’16 > Nov ‘16) and ISM bottomed in January 2016. Meaning, the market and economy was gaining strength and was able to absorb the new concerns easier than previously.


Source: Bloomberg

December 2015 – February 2018

The spread increased from the August 2015 lows and eventually peaked in July 2016 after the UK vote. After peaking, we saw the spread slowly decline and face little to no stress for the next 18 months, which is similar to what we saw after the European Banking Crisis in 2011.


Source: Bloomberg

Current Level of Stress in the Market

Following the continuous (generic) futures contracts can be good when looking for a general idea of stress in the market.

As noted previously, there are problems with only using the continuous futures. First, there can be sudden spikes in the data as Bloomberg transitions from one set of contracts to the next. Second, when we are in stressful periods, like we are now, it is best to look at the specific contracts for “healing points”.

To solve this issue, during periods of concern or stress I look at the two nearest spreads so I can see how the current and upcoming spread is developing.

The next chart shows the current 2nd month contract (UXJ8) minus the current 1st month contract (UXH8). The immediate short term healing point that needs to be exceeded is 0.20. More importantly, the longer term healing point is 0.60.

UXJ8 – UXH8
Source: Bloomberg

The next chart is the current 3rd month contract (UXK8) minus the current 2nd month contract (UXJ8). The immediate short term healing point is 0.15 and the longer term healing point is 0.60.

UXK8 – UXJ8 

Source: Bloomberg

Based upon prior periods that had the spread decline to -2.00 along with no recession (i.e. 2006, 2011, 2015), we will not pass 0.60 without retesting the lows of the spread. During these previous periods, we saw 3-4 weeks of calm from the lows before the market began to succumb to the sustained level of stress in the market. Meaning, the probability of a short-term market peak is increasing since the spread bottomed February 5th.

Longer Term Expectation of Volatility

When ISM peaked in the past, we tended to see bouts of high stress every 8-12 months. Between these periods of high stress, volatility declined back to low levels and the spread would increase to 1.00 or higher.

The only way we will see high stress remain constant is if we are heading into a recession, which I don’t foresee at this time.

Therefore, my base case is the first scenario since ISM peaked in September 2017. Updates on the volatility spread will be discussed in future posts since this has become a relevant topic once more.