Michael Cembalest Chairman of Market and Investment Strategy for J.P. Morgan Asset & Wealth Management Sep 6, 2022
Markets and multiples. Markets are pricing a Fed Funds peak of 4% by April, followed by a decline to 3% by year-end 2024. While inflation linked to food/energy, auto parts, other goods and cyclical services are rolling over, less cyclical services and housing inflation are still rising [Ex. 1 and 2; see p.7-8]. And while business surveys suggest that wage pressures have subsided from peak levels, wage growth is still extremely high [Ex. 3 and 4] and payrolls and jobless claims have only weakened a small amount. If the Fed is committed to restoring positive real interest rates (see chart, right), I find it hard to justify an increase in equity multiples.
Dot plot shows the median FOMC Fed Funds projections from June 2022 compared to the market expectation of future Fed Funds rates as implied by the overnight indexed swaps market. Markets are pricing a Fed Funds peak of 4% by April, followed by a decline to 3% by year-end 2024
Line chart shows S&P 500 price to forward earnings ratio vs inverted US real yields from 2016 to now. The lines closely track each other, demonstrating that equity markets have been trading almost lock-step with the 10 year Treasury less anticipated inflation
The recent equity rally is often compared to two prior bear market rallies, but both involved deep recessions (1973-1975 and 2007-2009) that we do not anticipate this time. To be clear, a recession is still a significant risk over the next few months. The US avoided recession four times after prior Fed hiking cycles but each time, food/energy inflation was very low which isn’t the case today [Ex. 5]. Furthermore, new orders have declined sharply relative to production, inventories are high [Ex. 6 and 7], housing is imploding and the US dollar is at a 20-year high (hurting 30%-40% of S&P revenues which are foreign-sourced). If there is a recession, I expect it to be a milder one: private sector savings are higher than before prior recessions, and a macroeconomic projection of the US business cycle points to a less severe contraction than in 2009 or 2020.
Bar chart shows private sector savings by country from Q1 2020 to Q1 2022 compared to the median savings level across countries in the run-up to previous recessions. Current levels imply that if there is a recession, it could be milder as a result of higher savings.
Line chart shows the US Manufacturing PMI Survey vs a PMI estimate based on macro variables including short rates, EM bond yields, US$ and Business Confidence Surveys. These estimates point to a mild recession next year
Bottom line for equity investors:
- While consumer price inflation may have seen its peak, it’s still elevated and wage inflation has not rolled over yet [Ex. 4]. The labor force participation rate is still below pre-COVID levels [Ex. 8], but since almost all of the gap is due to the age 54+ cohort, there’s less room for continued improvement (the LFPR for ages 25-54 is now back at pre-COVID levels). Furthermore, a measure of jobs openings plus employment vs the labor force shows the tightest labor markets in decades. So, the Fed will keep tightening. But if there is a recession, we expect it to be a milder one given the health of consumer balance sheets
- Leading indicators for the PMI manufacturing survey point to a mild recession next year (see chart above). The declining PMI is also a key input into leading indicators for S&P earnings [Ex.9]. The summer rally was due to rising P/E multiples [Ex. 10] which are unlikely to rise further given positive real interest rates and more quantitative tightening. Given weakening leading indicators and declining earnings, I expect a rollover in US equity markets this fall closer to the June lows for those looking for a better entry point
- The last Fed hike for the cycle, when it does occur, has historically been good news for investors; only during the 2000 cycle did the last Fed hike fail to result in a sustained rally [Ex. 11]
- See the two charts below: since 2009, financial repression resulted in 20%-60% of the stock market offering dividend yields above Treasury yields, and resulted in short rates that were way below asset yields. Both trends were only sustainable if inflation stayed low forever. Some of you will miss TINA investing (e.g., “There Is No Alternative” to equities since interest rates are below inflation). I won’t. It spawned a lot of mindless risk-taking and ruined fixed income markets for a generation, negatively impacting defined benefit plans, insurers and families invested in funds that reduce risk with age. Good riddance to bad policy
Line chart shows the percent of the total market cap that consists of stocks trading above the 10 year Treasury yield. Since 2009, financial repression resulted in 20%-60% of the stock market offering dividend yields above Treasury yields, and resulted in short rates that were way below asset yields. Both trends were only sustainable if inflation stayed low forever.
Line chart shows developed world asset yields vs developed world short interest rates which were very similar prior to 2009. Since 2009, short rates have been below asset yields, but that gap is closing
What it would take for the energy bill’s projected GHG reductions to actually occur, and why it matters
I read a piece in the Atlantic entitled “The Best Evidence Yet That the Climate Bill Will Work” [August 3, 2022]. The author writes: “First we got the bill. Now we have the numbers”. Really? What you have are projections from three energy modeling teams estimating that the bill could reduce GHG emissions by ~40% by 2030 vs 2005 levels, implying a quadrupling in the pace of decarbonization. There’s no discussion in the Atlantic article of what would actually have to happen to get there.
I looked at the detailed assumptions made by one of the modeling teams. These analysts are very smart and very well informed on energy transition dynamics. But: their models often assume perfectly optimal behavior by businesses and individuals based on prevailing incentives, and usually ignore frictional issues such as battery and critical mineral supply chain constraints3, interconnection delays of wind/solar power and the difficulty in siting new transmission. On the latter, House Democrats may block passage of the infrastructure project siting bill that Senate Democrats agreed to in exchange for Manchin’s Inflation Reduction Act support4.
The next page compares their assumptions and ours. The reason this is important: if you believe the energy bill modelers, the US could start enacting policies to constrain the natural gas industry since the US will need a lot less gas in 2030, and even less after that. But if you believe that the future could be closer to our assumptions, you would do no such thing for fear of ending up like Europe: energy-dependent and facing a difficult winter. With gas reserves headed for 90% by November it looks like Germany will be able to make it through the winter without Russian gas, but only if they continue to cut consumption by 15% vs normal levels.
Line chart shows estimates of greenhouse gas emissions declines resulting from the Inflation Reduction Act vs projections based on current policy. Current policy would result in a 27% decline, whereas projections of declines resulting from the IRA are between 31% and 44%.
Line chart shows the historical rate of US greenhouse gas emissions declines (0-2%) compared to the assumed annual reduction resulting from the Inflation Reduction Act.
Line chart compares Europe electricity prices to Europe natural gas from 2019 till now. The chart conveys that both prices have reached all-time highs in 2022 and continue to rise.
Line chart compares various natural gas supply-and-demand scenarios for Germany. Although Germany could reach a 90% reserve capacity by November 1st, if Russia cuts natural gas imports, Germany could run out of natural gas between January and March 2023.
Understanding the table. The table compares our assumptions for the year 2030 with one of the three modeling groups that analyzed the energy bill on behalf of the Senate. The color-coded column highlights the greatest differences between our assumptions and theirs; red = very different, green = very similar.
I emphasized the natural gas share of primary energy since it is a critical policy issue. It affects decisions on pipelines, electricity transmission, energy storage, export policy (European demand for LNG is expected to rise by 2.5x by 2030), winter heating regulations and decommissioning policies affecting coal/nuclear.
What will all those new IRS agents be doing?
The Inflation Reduction Act will dedicate $45 billion for tens of thousands of new IRS agents focused on tax enforcement, which the CBO estimates will raise $203 billion in additional revenues over a ten year period for a ~4x return. Sounds easy, doesn’t it? Not so fast; a May 2022 report from the General Accounting Office suggests this could be difficult to do, particularly if the IRS directs new agents to only audit taxpayers with more than $400k in income as requested by Treasury Secretary Yellen and as reiterated by IRS Commissioner Rettig.
- The audit rate of individual income tax returns fell from 0.90% in 2010 to 0.25% in 2019. This decline is mostly attributed to reduced IRS funding and attrition of IRS agents. Audit rates decreased the most for those with incomes over $200k, although audits are still skewed towards incomes of $1mm+ (chart, left)
- From 2010 to 2019, incremental taxes raised were highest for audits of those earning less than $200k (chart, right); the average percent change in the amount owed was highest for this category as well
- In contrast, higher incomes had the highest rate of “no change to taxes owed” after audit
- Let’s assume that the audit rate for those earning more than $400k reverts back to 2010 levels (i.e., it increases by a factor of 7x). We estimate that this would raise $8-$10 billion per year in 2031 based on data in the GAO report, far short of the $35 billion estimated by the CBO
So, I’m not sure how the Act will raise the assumed revenues if new agents only focus on wealthier taxpayers. In late August, the CBO sent a letter to the House Ways and Means Committee indicating that they had already cut their estimate of revenues raised from $203 billion to $180 billion.
Bar chart shows audit rates by income for the tax year of 2019. Individuals earning $5 million or more have a higher rate of auditing at ~2.5% than individuals who earn less than $500k at ~0.4% or lower.
Bar chart shows the estimated additional tax revenue from audits by various income levels. The majority of additional tax revenue from audits comes from individuals earning $200k or less.
Also: is the tax gap as large as advertised? What may be driving this initiative are estimates of a $300-$600 billion “tax gap” per year according to the Treasury5: the difference between what should be paid and what is paid. Most of the gap is due to underreporting of income rather than non-payment or non-filing. However, analysts at the Brookings Tax Policy Center cite several issues that result in tax gap overestimation6:
- The tax gap cited above includes additional taxes recommended by examiners even if the amount could be reversed on appeal or court challenge. And such reversals can be large: only 63% of additional taxes recommended from 2015 through 2019 were ultimately assessed after administrative appeals and abatements. That figure is likely to be even lower after further reductions following judicial review
- The IRS uses “detection controlled estimations” in its compliance studies to scale up the recommendations of all examiners to match those with the largest upward adjustments in personal income. In other words, the process is leveraged to revisions made by examiners with the largest estimates of income under-reporting. This “highest common denominator” approach is estimated to account for two thirds (!!) of the entire individual income tax underreporting gap
- The Treasury’s estimate of the tax gap is partially based on a 2007 study citing $1 trillion in US taxpayer offshore wealth and non-payment of taxes on offshore interest and dividends. However: the IRS ramped up offshore enforcement efforts in 2008 after financial institutions in Switzerland, Liechtenstein, Israel and the Caribbean turned over thousands of names to the IRS, resulting in $6 billion in penalties paid. Since then, another $11 billion in penalties were paid by 56,000 taxpayers entering into the IRS Offshore Voluntary Disclosure Program. The Foreign Account Tax Compliance Act of 2010 made it even harder for US taxpayers to hide assets, and the IRS/Justice Department have consistently prosecuted FBAR violations over last decade, penalizing financial institutions that harbor violators. Bottom line: a lot has changed since 2007
- The Tax Cut and Jobs Act of 2017 caused more taxpayers to claim standard rather than itemized deductions, and the C Corporation tax rate declined from 35% to 21%; both are likely to reduce the degree of underreporting of income
- Some components of the tax gap reflect taxpayer-IRS disputes regarding the timing of depreciation deductions. While their resolution could increase taxes due in a given year, they do not imply perpetual underreporting of income since taxes due in future years would be lower in these cases
Whatever the tax gap is, new IRS agents are likely to focus on partnerships, sole proprietorships and owners of commercial and residential rental properties7. The Treasury believes that income underreporting could be 50%-60% for these entities given less withholding and information reporting (see chart). The Brookings paper also cites underreporting of pass-through income as a factor resulting in tax gap underestimation.
Last point: if you hold a lot of crypto, they may be coming for you as well if you exchange crypto for goods and services without recognizing appreciation as capital gains8.
Bar chart shows the underreporting of income by various entities. The percent of income not reported could be ~50% - 60% for entities with income subject to little or no information reporting.
Exhibits
Exhibit 1: inflation trends
Exhibit 2: housing inflation
The indexed chart shows the rise in US home prices, US observed rent and US housing CPI from 2018 till now. Currently, all three indices are at all-time highs.
Exhibit 3: wage expectations
The chart shows the US Business Leaders Survey Wage Expectations from 2005 till now. Currently, US business leaders expect wages to decrease in the future.
Exhibit 4: consumer price and wage inflation
Line chart compares the employment cost index to PCE core from 1980-2022. The two lines follow a similar trend until the early 1990s, when the employment cost index increases; this gap closes in 2005, 2008, and 2020.
Exhibit 5: the only time the US has avoided recession when the Fed hikes is when food/energy inflation is low (as shown in the red bars)
Exhibit 6: leading indicators
The line chart shows ISM manufacturing PMI new orders less the production index from 2015 to July 2022. The line typically oscillates around 0 for this period until it spikes up to +6 in 2021, then dives down to -6 in 2022.
Exhibit 7: inventories
The line chart shows ISM manufacturing PMI inventories from 1980 to July 2022 and demonstrates that inventories have risen and are expanding
Exhibit 8: labor force
The line chart shows the labor force participation rate for the percent of the US population aged 25-54 and for the percent of the US population aged 55 and over from 2010-2022. The labor force participation rate for the 25-54 age group is 82.8% and for the 55 and over age group is 38.6% as of August 2022.
Exhibit 9: leading indicator for earnings
The line chart compares a leading indicator earnings model to actual trailing earnings growth to suggest that there is a modest year-over-year downtrend ahead.
Exhibit 10: price-to-earnings ratios
The line chart shows that forward price-to-earnings ratios for the Megacap 8 (GOOGL, AMZN, AAPL, FB, MSFT, NFLX, NVDA, TSLA), the Russell 1000 Growth Index and the Russell 100 Index declined from all-time highs in 2022, but still remain elevated relative to history.
Exhibit 11: equities and the last Fed hike
The line chart shows S&P 500 returns in the 12 months before and 18 months after Fed hikes in the following years: 1995, 1997, 2000, 2006, 2015, and 2018. S&P 500 returns are indexed to 100 = most recent Fed hike for the respective year.
1 Short covering. From the June lows, the S&P 500 rose by 12% while an index of the most shorted stocks rose by 35%.
2 Rob Portman (R-OH), a Senator I admire greatly and whose retirement is another sign of institutional decay in the Senate, released a report this year on how China's government targeted Federal Reserve employees in a decade-long infiltration campaign aimed at stealing US monetary policy secrets. My question: wouldn’t the Fed’s very poor track record in anticipating future inflation and policy rates dissuade people from wanting to steal its secrets and copy it?
3 Domestic content requirements for EV subsidies in the energy bill. Critical mineral minimum extraction/processing percentages for US/Free Trade Area countries start at 40% in 2024 and rise to 80% by 2027. Battery component requirements for manufacturing or assembly in North America start at 50% in 2024 and also rise to 80% by 2027.
4 “Prospects dim in House for Manchin’s federal permitting measure”, Rollcall.com, August 17, 2022
5 “The Case for a Robust Attack on the Tax Gap”, US Treasury, September 7, 2021
6 “The Tax Gap’s Many Shades of Gray”, Hemel, Holtzblatt and Rosenthal (TPC), September 30, 2021
7 A partial list of what the IRS looks for: allocations of ordinary income to tax-exempt partners, and allocations of deductions/long term gains to partners in high tax brackets, that are not in accordance with ownership interests; S corporations that do not pay sufficient wages to shareholder-employees, or which provide them with large non-wage distributions; work arrangements improperly structured as independent contractors so as to benefit from expanded pretax deferral options and qualified business income deductions.
8 The IRS reportedly believes that there is a large degree of non-reporting by crypto holders. Recent new rules include Executive Order 14067 instructing the IRS to focus on non-compliance; enhanced Form 1040 disclosure on crypto sales, exchanges and receipts; and Notice 2014-21/Rev. Rule 2019-24 providing guidance on crypto tax treatment. What may come next: comprehensive field exams of digital asset funds, principals and investors; enhanced information disclosures piercing some bearer aspects of crypto holdings; and assessment of economic and possibly criminal penalties which are well publicized in order to raise the bar for non-compliance.