Weekly Energy Update (August 25, 2022)

by Bill O’Grady, Thomas Wash, and Patrick Fearon-Hernandez, CFA | PDF

Crude oil prices remain under pressure on fears of a deal with Iran and weakening economic growth.

(Source: Barchart.com)

Crude oil inventories fell 3.3 mb compared to a 2.5 mb draw forecast.  The SPR declined 8.1 mb, meaning the net draw was 11.4 mb.

In the details, U.S. crude oil production fell 0.2 mbpd to 12.0 mbpd.  Exports fell 0.8 mbpd, while imports were unchanged.  Refining activity rose 0.3% to 93.8% of capacity.

(Sources: DOE, CIM)

The above chart shows the seasonal pattern for crude oil inventories.  Clearly, this year is deviating from the normal path of commercial inventory levels although the past two weeks are consistent with seasonal behavior.  We will approach the usual seasonal trough for inventories in mid-September.

Since the SPR is being used, to some extent, as a buffer stock, we have constructed oil inventory charts incorporating both the SPR and commercial inventories.

Total stockpiles peaked in 2017 and are now at levels last seen in 2003.  Using total stocks since 2015, fair value is $106.51.

With so many crosscurrents in the oil markets, we are beginning to see some degree of normalization.  The inventory/EUR model suggests oil prices should be around $64 per barrel, so we are seeing about $24 of risk premium in the market.

Market news:

 (Source:  Strategas)

 Geopolitical news:

 Alternative energy/policy news:

(Source:  Adam Tooze)

  View PDF

Asset Allocation Bi-Weekly – The Inflation Surprise (August 22, 2022)

by the Asset Allocation Committee | PDF

July inflation came in below expectations.  On a yearly basis, the market expected an overall rate of 8.7%, while the actual reading was 8.5%.  For core CPI, the actual was 5.9% compared to expectations of 6.1%.  Perhaps the most bullish part of the report was the monthly change, where the overall rate was unchanged compared to June and the core rose by 0.3% compared to expectations of a 0.5% rise.  This was the first monthly reading for the overall rate that was zero or less since May 2020.

The reaction of financial markets was swift and bullish.  The S&P 500 rose over 2% on Wednesday, August 10.  The dollar fell, commodities mostly rose, and the short end of the yield curve rallied.  Why the strong reaction?  The July data was the first time in months we have seen a modest rise in inflation and there is hope that we may be past “peak inflation.”  This hope may be fulfilled.  It’s clear the U.S. economy is slowing and there is evidence that supply chains are slowly improving.

Is this optimism justified?  Although the news on inflation was positive, comparing the policy rate relative to CPI and unemployment suggests the FOMC still has a long way to go before achieving a neutral rate.

In the above chart, the independent variable is CPI less the unemployment rate.  The blue line shows the difference from the model’s estimation of what the policy rate should be based on the difference between CPI and unemployment.  Since 1957, the Federal Reserve has never conducted a policy this easy.  Although recent tightening has somewhat narrowed the gap, the FOMC needs to see either a rise in unemployment or a sharp drop in inflation.

Financial markets, in contrast, are anticipating an end to this tightening cycle.  Comparing the implied interest rate from the Eurodollar futures market, two years into the future, suggests the Fed will be easing in the coming months.

Eurodollar futures are already projecting a lower rate over the next two years.  We have not reached the extremes of earlier tightening cycles, but another rate hike of 50 bps will push the deviation into easing territory.

So, why the difference in the two models?  The financial markets are anticipating a change in economic conditions, either falling inflation, rising unemployment, or both.  Why?  In part, it’s because the financial markets are anticipating recessionary conditions.  For example, the yield curve has inverted.   The chart below depicts a study that looks at 10 different Treasury yield curves.  History shows that when more than six of these curves invert, a recession occurs, on average, in 15 months.  In July, eight of the curves inverted.  As an aside, note that it is common for these curves to steepen after the inversion but before the recession develops.  One should not take any comfort in fewer yield curves inverting once we exceed six.

What is unknown is how fast inflation will fall or unemployment will rise.  Our concern is that the financial markets are anticipating a rapid adjustment in the first chart’s indicator that may not be borne out by the data.  If that outcome occurs, the FOMC may tighten more than the market expects and maintain that tightness longer than anticipated.  That outcome would likely be negative for equities, but for now, hopes of falling inflation and an easing response from policymakers have lifted risk assets.  If this positive outcome fails to materialize, a downturn in risk assets is likely.

View PDF

 

Weekly Energy Update (August 18, 2022)

by Bill O’Grady, Thomas Wash, and Patrick Fearon-Hernandez, CFA | PDF

Crude oil prices remain under pressure on fears of a deal with Iran and weakening economic growth.

(Source: Barchart.com)

Crude oil inventories fell 7.1 mb compared to a 0.3 mb build forecast.  The SPR declined 3.4 mb, meaning the net draw was 10.2 mb.

In the details, U.S. crude oil production was steady at 12.2 mbpd.  Exports rose 2.9 mb, while imports were unchanged.  Refining activity dipped 0.8% to 93.5% of capacity.

(Sources: DOE, CIM)

The above chart shows the seasonal pattern for crude oil inventories.  Clearly, this year is deviating from the normal path of commercial inventory levels although this week’s outsized decline is consistent with seasonal behavior.  We will approach the usual seasonal trough for inventories in mid-September.

Since the SPR is being used, to some extent, as a buffer stock, we have constructed oil inventory charts incorporating both the SPR and commercial inventories.

Total stockpiles peaked in 2017 and are now at levels last seen in 2004.  Using total stocks since 2015, fair value is $104.45.

With so many crosscurrents in the oil markets, we are beginning to see some degree of normalization.  The inventory/EUR model suggests oil prices should be around $64 per barrel, so we are seeing about $24 of risk premium in the market.

Market news:

 Geopolitical news:

 Alternative energy/policy news:

  • With the Inflation Reduction Act now signed into law, a backlash against economists for pushing for a carbon tax is developing. To some extent, this makes sense as using other tools can be more politically popular.  However, from an efficiency standpoint, a carbon tax would still be a superior policy, but obviously, if you can’t get it passed, holding on to that policy to the exclusion of all others makes little sense.  What the bill is really all about is industrial policy.  Government shaping the economy is nothing new but is generally considered legitimate only in cases of clear market failure.  Since a carbon tax was never implemented, it really hasn’t been proven that a market failure exists.
  • There is great excitement in the environmental community over the new measures but one potential concern is the lack of workers to build out the plan.
  • Germany is extending the life of its three remaining nuclear power plants.
  • Any commodity activity disturbs something. Whether its drilling, ranching, farming, or mining, something or someone gets disturbed.  As demand for lithium rises, opposition to mining or brining has emerged.  Although such opposition may be overcome, higher costs are likely to result.
  • As we noted last week, the price of EVs continues to climb. Ford’s (F, $16.18) announcement of substantial price increases on its F-150 “Lightning” EV pickup is the most recent example of this issue.
  • There is growing evidence that the Arctic is warming faster than other parts of the world. The impact is difficult to estimate, but we would expect greater weather variability from this situation.
  • Much of the Midwest, parts of the Southwest, Florida, and the Atlantic coast could become subject to extreme heat events in the coming decades. But the real worry is heat in areas unprepared. The linked map shows the areas of installed air conditioning.
  • Hot weather just isn’t an inconvenience. The drought and warm weather is affecting industrial activity in Germany.  In China, power shortages, caused by hot weather, are causing car and battery plants to suspend operations.  Tech firms have also temporarily shut down.
  • Delays of utility-scale solar projects are steadily rising. These delays may be tied to trade restrictions which have recently been eased.
  • California looks ready to extend the life of the Diablo Canyon nuclear power plant that was scheduled for decommissioning.
  • Increasingly, we are seeing an “all of the above” strategy in energy investment. Investing in renewables doesn’t necessarily preclude investing in fossil fuels.
  • Although wood burning is not necessarily environmentally friendly, it appears Germans are considering it in the face of rising fossil fuel prices. Wood pellets are also seeing rising demand in Europe and Asia.
  • This recent report from the Peterson Institute details China’s dominance in rare earths processing.

  View PDF

Asset Allocation Bi-Weekly – The Devil Is in the Details (August 8, 2022)

by the Asset Allocation Committee | PDF

U.S. policymakers used deregulation and globalization to corral inflation from 1966 to 1982.  Unfortunately, that policy was at odds with America’s superpower role, which required the U.S. to act as global importer of last resort.  If the U.S. didn’t consume all the goods the world wanted to sell to Americans, the world economy would face a liquidity crisis.  Policymakers addressed the issues of containing inflation and providing global liquidity by deregulating financial services, which made it easier for households to borrow money.  Although deregulation and globalization slowed real income growth, the ability to borrow allowed households to absorb global imports, holding the international system together.  After 1995, this lending was increasingly attached to residential real estate, which was considered safe.  Sadly, one of the key economic imbalances that was revealed during the Great Financial Crisis was excessive household debt.  Since the crisis, the economy has been trying to address this debt overhang.  There has been some progress as household debt peaked at 129.4% of after-tax income in Q1 2008 but fell to 84.4% in Q1 2021.  Since then, it has risen to 96.5%.

Although policymakers haven’t targeted this issue, we believe that addressing this debt situation is not only key to improving the health of the economy, but the austerity required to reduce debt may be a factor behind political polarization.  The last time the U.S. had a similar debt issue was in the late 1920s when the Great Depression was the resolution, although the situation wasn’t fully addressed until WWII.  From a financial perspective, WWII finally resolved the private sector debt problem by placing that debt on the public balance sheet.  The debt relative to the size of the economy was reduced on the public balance sheet through financial repression.

One of the difficulties in discussing debt is proper scaling; in other words, how do we know when debt is “too high”?  Economists often use nominal GDP or some sort of income measure to scale debt.  The problem is that both GDP and income are “flow” data, meaning that they measure a quantity calculated over a period of time, while debt is “stock” data, which is a level at a specific time.  In terms of debt, income or GDP may or may not tell us much about the ability to service the debt.

Accounting often creates ratios that measure stock or flows.  For example, assets divided by liabilities are two stock numbers that give us some idea about the balance sheet of a firm or household.  Clearly, if the assets exceed liabilities, it suggests solvency.

From 1970 to 1990, American households had more cash than debt.[1]  After 1990, household leverage rose, peaking with debt exceeding cash to the tune of nearly $6.0 trillion.  The difference narrowed after the Great Financial Crisis by more than 50%.  The huge injection of fiscal aid to households during the COVID-19 pandemic finally led to cash exceeding debt for the first time in three decades.

So, have we resolved the household debt problem?  Perhaps, but the Federal Reserve’s Distributional Financial Accounts, which examines household balance sheets by income, suggests that the debt situation hasn’t necessarily been fixed.

We divide households into three groups: the top 10%, the middle 40%, and the bottom 50%.  The top 10% has seen its cash levels rise relative to debt for most of this century, but this difference widened dramatically during the pandemic.  Since the upper income brackets were mostly excluded from direct cash payments, it’s likely that this group liquidated appreciated assets.  We do note that all three classes took on more debt, but in the case of the top 10%, the cash accumulation far exceeded these new liabilities.  The middle 40% saw cash rise relative to debt into Q1 2021, but over the past year, liabilities grew modestly relative to debt.  However, for the bottom 50%, net debt continued to rise even with the influx of pandemic transfer payments.

We don’t have a data series by income prior to 1989, so we can’t compare what occurred during and after WWII, but, given the high marginal tax rates of that period, we suspect that the lower income classes saw their balance sheets improve.  What can we take away from the above chart?  First, as interest rates rise, consumption may fall since the bottom 50% increased their leverage during the pandemic.  Consumption will then have to come from the upper 50%.  Second, given the massive cash balances of the top 10%, asset prices could find support in the coming months.  Although higher cash yields from rising interest rates might keep this cash on the sidelines, we suspect this level of cash will eventually find its way into the equity, commodity, and debt markets.  This flow may depend on signs that the FOMC is near the end of its tightening cycle, but once such a catalyst emerges, the conditions for a strong financial market recovery are in place.  The great unknown, of course, is which market the potential flows will favor.

View PDF


[1] This dataset, from the FRB’s Financial Accounts of the U.S., includes households plus non-profits that service households.  Thus, strictly speaking, this isn’t just households, but data suggests the non-profit contribution is relatively minor.

Weekly Energy Update (August 4, 2022)

by Bill O’Grady, Thomas Wash, and Patrick Fearon-Hernandez, CFA | PDF

Prices are testing long-term support near $90 per barrel for crude oil.

(Source: Barchart.com)

Crude oil inventories rose 4.5 mb compared to a 1.5 mb draw forecast.  The SPR declined 4.7 mb, meaning the net draw was 0.2 mb.

In the details, U.S. crude oil production was steady at 12.1 mbpd.  Exports fell 1.0 mb, while imports rose 1.2 mbpd.  Refining activity declined 1.2% to 91.0% of capacity.

(Sources: DOE, CIM)

The above chart shows the seasonal pattern for crude oil inventories.  Clearly, this year is deviating from the normal path of commercial inventory levels.  The fact that we are not seeing the usual seasonal decline is a bearish factor for oil prices.

Since the SPR is being used, to some extent, as a buffer stock, we have constructed oil inventory charts incorporating both the SPR and commercial inventories.

Total stockpiles peaked in 2017 and are now at levels last seen in 2004.  Using total stocks since 2015, fair value is $104.04.

With so many crosscurrents in the oil markets, we are beginning to see some degree of normalization.  The inventory/EUR model suggests oil prices should be around $65 per barrel, so we are seeing about $30 of risk premium in the market.

Gasoline Prices

One of the challenges in determining “how high” gasoline prices truly are is to find the proper scaling variable.  The most common metric is to deflate the current price by some price index.  Although that method is sound (it compares gasoline prices to other items), it doesn’t, in our opinion, capture the centrality of gasoline prices.  After all, a commuter often has limited ability to substitute either a fuel (cars tend to only use gasoline) or a travel method (car sharing requires coordination, public transportation varies by locale).

As an alternative to using a price index, we take the nominal hourly wage for a non-supervisory worker and divide into that wage the current average national price of gasoline.  This measure calculates how many gallons of gasoline a worker can purchase for an hour’s worth of work.  The scaling method isn’t perfect as it doesn’t take into account taxes and benefit costs, for example, but it does allow us to compare the current price to history.

The below chart shows that calculation along with the Presidential Approval Ratings.  On the left axis, the higher the number, the more gallons one can buy with an hour’s worth of work.  So, the higher the reading, the better off the worker is.  Over time, the average is 8.6 gallons.  As the chart shows, we are well below that level at present.  It’s also clear that Presidential Approval Ratings and the number of gallons one can buy “rhyme.”  The two variables are not perfectly correlated.  There are several spikes in approval tied to wars or other geopolitical events that sometimes occur with lower gallon per hour readings.  But, overall, President Biden’s low approval ratings is not being helped by falling gallon per hour numbers.

 

Market news:

 Geopolitical news:

 Alternative energy/policy news:

  View PDF

Bi-Weekly Geopolitical Report – Political Crises for Top U.S. Allies (August 1, 2022)

by Patrick Fearon-Hernandez, CFA | PDF

It’s been a hot summer for some of the United States’ key allies, and we don’t just mean the weather.  We’ve seen a major political assassination in Japan, the scandal-driven ouster of a prime minister in the United Kingdom, and a failed vote of confidence that led to the resignation of the prime minister in Italy.  Since all these political crises happened in such a short amount of time, we decided to explain each of them in this report and discuss their common features and geopolitical implications.  As always, we end with a discussion of the ramifications for investors.

View the full report

Don’t miss the accompanying Geopolitical Podcast, available on our website and most podcast platforms: Apple | Spotify | Google

Business Cycle Report (July 28, 2022)

by Thomas Wash | PDF

The business cycle has a major impact on financial markets; recessions usually accompany bear markets in equities.  The intention of this report is to keep our readers apprised of the potential for recession, updated on a monthly basis.  Although it isn’t the final word on our views about recession, it is part of our process in signaling the potential for a downturn.

The Confluence Diffusion Index declined for the second consecutive month. The latest report showed that seven out of 11 benchmarks are in expansion territory. The diffusion index declined from +0.8789 to +0.6364 but remains well above the recession signal of +0.2500.

  • Heightened recession worries weighed on the financial indicators.
  • Tightening financial conditions and elevated inflation have led to production slowdowns.
  • Employment indicators suggest that the labor market is starting to loosen.

The chart above shows the Confluence Diffusion Index. It uses a three-month moving average of 11 leading indicators to track the state of the business cycle. The red line signals when the business cycle is headed toward a contraction, while the blue line signals when the business cycle is in recovery. The diffusion index currently provides about six months of lead time for a contraction and five months of lead time for recovery. Continue reading for an in-depth understanding of how the indicators are performing. At the end of the report, the Glossary of Charts describes each chart and its measures. In addition, a chart title listed in red indicates that the index is signaling recession.

Read the full report

Weekly Energy Update (July 28, 2022)

by Bill O’Grady, Thomas Wash, and Patrick Fearon-Hernandez, CFA | PDF

It appears that oil prices are settling into a broad trading range between $125 and $95 per barrel.

(Source: Barchart.com)

Crude oil inventories fell 4.5 mb compared to a 1.5 mb draw forecast.  The SPR declined 5.6 mb, meaning the net draw was 10.1 mb.

In the details, U.S. crude oil production rose 0.2 mbpd to 12.1 mbpd.  Exports rose 0.8 mb, while imports fell 0.4 mbpd.  Refining activity declined 1.5% to 92.2% of capacity.

(Sources: DOE, CIM)

The above chart shows the seasonal pattern for crude oil inventories.  Clearly, this year is deviating from the normal path of commercial inventory levels.  Although it is rarely mentioned, the fact that we are not seeing the usual seasonal decline is a bearish factor for oil prices.

Since the SPR is being used, to some extent, as a buffer stock, we have constructed oil inventory charts incorporating both the SPR and commercial inventories.

Total stockpiles peaked in 2017 and are now at levels last seen in 2004.  Using total stocks since 2015, fair value is $103.88.

With so many crosscurrents in the oil markets, we are beginning to see some degree of normalization.  The inventory/EUR model suggests oil prices should be around $65 per barrel, so we are seeing about $30 of risk premium in the market.

Gasoline Demand

Last week, we noted that gasoline demand was weakening and that has been putting pressure on oil prices.  This week, we take a look at the number of licensed drivers in the U.S.

The data is annual, and the last available year is 2020 so we don’t know how this metric was affected by the pandemic.  On a trend basis, the number of drivers rose above trend in the early 1970s as baby boomers reached driving age.  As population growth slowed, the number of drivers fell steadily back to trend.  However, after the Great Financial Crisis, the number of drivers fell rapidly and has yet to recover.  It doesn’t appear to be the boomers.  In looking at the number of licensed drivers aged 65 and over, the growth rate is still 3% per year.  What we do find is that the number of drivers aged 29 years and younger declined over the past three years, falling 2.2% in 2020.  Social media might play a role, or the urbanization of younger households could play a role as well.

In observing the above chart, it again begs the question, “Why would anyone build additional refinery capacity if the number of drivers is falling below trend?”

 Market news:

Geopolitical news:

  • Although China is buying more Russian energy, its investments into Russia through the “belt and road” project have fallen to zero. Beijing is trying to avoid Russian sanctions.
  • One of the reasons given for President Biden’s trip to the Middle East was to reduce Chinese influence in the region. Washington fears that China is attempting to expand its footprint in the region as the U.S. slowly withdraws.  One area of interest for Beijing is Iraq.  However, China will have to navigate increasing intra-Shiite tensions.
  • Iran announced it has arrested suspected Mossad spies who were conspiring to attack Iran’s “sensitive” sites. This action occurs just as evidence appears to show that Iran’s nuclear program is expanding rapidly.
    • Iran’s foreign minister canceled a trip to the U.N. for meetings on nuclear non-proliferation. The IAEA wants Iran to turn on monitoring devices on its nuclear facilities; Iran has refused.
    • The last of the former administration’s nuclear negotiators has been replaced, reducing the odds that Iran will return to the 2015 JCPOA.
  • China and Australia have been at odds over pandemic policy. The PRC has retaliated against Australia by cutting imports.  But, as coal inventories fall, China may be forced to ease import bans on coal.
  • Cryptocurrencies have been used in cybercrime and sanctions evasion for some time. The Treasury Department is investigating whether a U.S. crypto exchange facilitated Iranian transactions in violation of sanctions.

Alternative energy/policy news:

  View PDF

 

Asset Allocation Bi-Weekly – The Puzzle of the Labor Force (July 25, 2022)

by the Asset Allocation Committee | PDF

In the June 27 report, we discussed the idea that the FOMC may focus on reducing job openings rather than raising the unemployment rate as a way to ease labor-market tightness.   The consensus from the establishment survey portion of the June payroll report, which questions employers, was that hiring remained elevated.  The other part of the survey, the household survey, comes from households, and generates the unemployment rate among other statistics.  The household survey showed a decline in employment, but the unemployment rate held steady because the labor force fell by a similar amount.

Although the payroll survey is considered by most economists to be the most reliable data from the employment report, the unemployment rate, which compares employment to the labor force[1] from the household survey, garners the most public attention.   Our concern is that changes in demographics may lead to an unemployment rate that may not reflect a weakening labor market.

The chart above shows the percent of the “working age population” relative to the total population.  To capture the past 120 years of history, the Census Bureau includes 15-year-old citizens, which are not working in the modern economy.  But using this age ranges allows us to compare a longer history when these teens were more commonly part of the workforce.

The rise of immigration and improvements in infant mortality rates led to an increase in the working-age population into 1940.  Before this, restrictions on immigration which began in the mid-1920s, World War II, and the Great Depression all negatively affected birth rates. The “baby boom” after World War II ended initially led to a massive decline in the working-age population which started to reverse in 1960 when the first “boomers” born in 1946 began to enter the workforce age range.  The steady progression of baby boomers entering this age bracket led to a steady rise in the working-age population until it peaked in 2006.  Since 2006, the working-age population has been declining.  We have added a vertical line on the chart to differentiate between actual and the Census Bureau’s forecast.  The current and forecast data suggest the number of available workers could decline in the coming years.

How could this fate be avoided?  There are essentially three ways.  First, immigration could rise.  The political environment suggests that this is unlikely.  Second, increasing the number of workers in the prime-age category might boost the work force; currently 74.4% of those in the age category of 16-64 are actively participating in work.  Additional training, subsidized childcare, relocation assistance, and other policies might encourage higher rates of employment.  Third, encouraging older workers to continue working could also increase employment.  For the most part, the third tactic was mostly used until the pandemic.

But, since the pandemic, participation of those over 65 years-of-age has declined and is not showing signs of recovering.

So, how could a falling labor force lead policymakers astray?  To capture this issue, we have created an indicator that is the difference between the yearly change in CPI and the unemployment rate.  In modeling this indicator relative to the policy rate, it shows that the FOMC’s policy is still too easy.

The model suggests that the fed funds target should be 10.3%; such a rate is unfathomable.  We do note, however, that with the exceptions of 2006-07 and recessions, fed funds have been below the fitted rate this century.  Still, to move fed funds to the one standard error level (the lower parallel line on the deviation axis) would require a fed funds in the neighborhood of 7%.

To narrow the deviation, the unemployment rate needs to rise, and inflation needs to fall.  Inflation has remained stubbornly high but will likely start to decline at some point.  But as the above demographics suggest, it might take a sharp drop in employment to raise the unemployment rate in a world where the labor force isn’t likely to grow very fast.  The chart above highlights the dilemma.

The chart above measures the labor force relative to its peak.  A reading of one indicates a new peak in the labor force.  Note that in the early 1950s, when the working-age population was declining relative to the population as a whole, it was not uncommon for the labor force not to rise to the earlier peak in expansions.  But, as the working age population rose, new peaks became common.  The pandemic led to a massive decline in the labor force, and although it has steadily recovered it remains well below the earlier peak.  The forecast we created is based on the Bureau of Labor Statistics estimates of labor-force growth this decade.  Based on this forecast, we won’t make a new peak before mid-2024.  If a recession occurs before that point, an event that appears increasingly likely, we won’t make a new peak in this expansion, which is something we haven’t done since the early 1950s.  That factor will keep the unemployment rate lower than it otherwise would be and may lead the FOMC to tighten policy more than expected.

[1] Defined as those working and those unemployed but actively seeking employment.

View PDF