The Economy’s Weekly Recap

8/19/24 - 8/26/24

Raymond Lin

Vanishing jobs? More like vanishing newsletter!

Apologizes for the delayed release. I’ve been busy recently and wasn’t able to write this week’s edition of Phi Fiscal until this week- Raymond

This Week’s Prominent Events

Vanishing Jobs

• A major metric used to see how the US economy is doing, which also helps guide the Federal Reserve’s policy on interest rates, is net job additions. Calculating the number of jobs added every month is a complex task though, so the Bureau of Labor Statistics that collects that information uses two different surveys and frequently refines its estimates through revisions. 

• Last week, the BLS issued a rather large revision, the scale of which hasn’t been since the Great Recession. This revision cut the jobs added from March 2023 to March 2024 by 819,000, lowering the net jobs added from an initial estimate of 2.9 million to a new  preliminary estimate of just 2.1 million. This revision was much higher than the expected revision of just 500,000, suggesting the US labor market may have been weaker than originally anticipated. Additionally, the revision of 819,000 is a 0.5% reduction in total employment, the largest reduction since 2009 following the Great Recession. For reference, the 10 year average of revisions is just +/- 0.1%. 

• Although the data won’t be finalized until February 2025, the revision supports the recent trend of slowing economic and job growth in the US. If August’s jobs report tells a similar story as the revision, then the odds of a September interest cut are very likely. 

• When looking at Federal Reserve Chairman Jerome Powell’s recent comments at the annual Jackson Hole conference, this sentient seems especially valid. During the conference, he said “The time has come for policy to adjust”, citing the direction of the economy and the risks of keeping rates high for longer. With this statement in mind, it seems practically guaranteed that September will see a rate cut, although the scale of the rate cut is not certain yet. It could be 25 basis points or 50 basis points, although the former is more likely as the Fed doesn’t want to agitate declining inflation. 

Ford’s EV Rollback

• Once touted as a game changer, EVs have seemingly lost their luster as demand for them slows and innovations like AI capture public attention. This phenomenon can be observed in the slowing rate of EV sales, which grew 46% in the first 7 months of 2023 but only 7% in the first 7 of 2024.  This slowing growth comes at a time when many automakers are losing money on their EVs, such as Ford’s $2.5 billion losses in its EV department during the first 6 months of 2024. With slowing growth and continued unprofitability, some companies have decreased investment in EVs. For example, Ford decreased the share of EV spending in its capital budget from 40% to 30%.

• Ford doubled down on this rollback of EVs last week by canceling its 3 seat EV SUV and delaying the release of its EV pickup truck until 2027, which altogether Ford anticipates could lead to a $1.9 billion loss. In the meantime, Ford plans on creating smaller and more affordable hybrid cars, believing that smaller hybrids give convenience and could be more profitable as battery costs are a smaller problem. This is a reversal from the traditional ICE playbook, where large SUVs and pickup trucks are more profitable. 

• However, this does not mean EVs are endangered. For example, Tesla is an EV only company that remains profitable and remains dedicated to the industry. However, companies like GM also remain committed to EV production, with it recently saying that EVs would be profitable once 200,000 are produced during Q4. Such continued EV efforts can only be observed in the battery industry, where over $100 billion has been invested in EV battery plants across the US. This will help reduce battery costs and support the positive long term trajectory of the EV market, which the International Energy Agency expected to entail 50% of global car sales by 2035.

Apartment Price Fixing

• While it lacks a public presence, the company RealPage has an impact on millions of Americans' lives. It’s a software company that recommends what landlords should charge for rent by using AI and confidential data about occupancy and rent. It has advertised that it can help landlords earn between 3% to 7% more, leading to RealPage helping set rents for more than 3 million US apartments. 

• Unfortunately for RealPage and its PE owner Thoma Bravo, a 2022 ProPublica investigation shed light on the role RealPage had on rent prices across the country, which has now culminated in a DOJ antitrust lawsuit. In the lawsuit, the DOJ alleges that RealPage facilitated price fixing that boosted rents for millions of people, highlighting the role of RealPage’s algorithm. It also accuses RealPage of monopolizing the market for software that helps landlords set rent prices. Interestingly, the DOJ’s lawsuit is actually late to the party as RealPage has been sued twice by DC and Arizona in November and February over illegally conspiring to raise rents. 

• However, while RealPage does recommend prices, it may be hard for the DOJ to prove it within the framework of the century old Sherman Antitrust Act. For one, RealPage only advises prices and landlords can reject recommendations. Only 40% to 50% of rents advised by RealPage fall within 1% of RealPage’s recommendations, suggesting it is often ignored. 

Data Center Construction Boom

• A mundane title right? Well, it could perhaps be better phrased as the AI construction boom as many of the data centers that have been constructed recently are primarily dedicated to training or operating AI services. 

• When looking at data center energy usage, a surrogate for the growth of data centers, 500 megawatts of new data centers were created in the first half of 2024. As things stand, data center energy usage and presumably construction have jumped 70% from a year ago. Most of this increase has been concentrated in 8 areas: Northern Virginia, Dallas and Fort Worth, Silicon Valley, Chicago, Phoenix, the New York Tri-State Area, Atlanta, and Oregon. 

• The reason for this concentration? Resource availability. These data centers require 2 major types of infrastructure: water and energy. In Virginia, one of the top places for data centers, water usage has increased 66% between 2019 and 2023. Energy demand is expected to grow as well, with Goldman Sachs predicting data center power demand to grow 160% by 2030, although it will only become around 4% of US electricity consumption

• Nonetheless, this amounts to a sizable increase that could constrain future data center construction and AI advancement as suitable places to build data centers become limited and natural resource usage leads to opposition.

Chinese Loan Collapse

• While the US economy has been quite strong following the pandemic, China’s has struggled to match its pre-Covid growth. For example, China’s unemployment rate increased by 0.2% in July, reaching 5.2%. For those 16 to 24, it was 17%, the highest figure since China began releasing the data last December. The economically vital real estate sector saw new home prices fall 4.9% from last year, the fastest drop in 9 years. However, retail sales did rise 2.7% annually in July, mostly due to seasonal reasons.

• Generally though, the economy is still struggling, a fact evident in the decline of China’s new loans. In July, new loans fell 88% from last July as they came in at just $36.28 billion, missing expectations by around 40%. This lower loan demand is the result of decreasing demand for credit and lower consumer spending, both of which are likely to perpetuate China’s relative economic stagnancy. 

• A symptom of this decreasing loan demand and general economic apprehensiveness has been increased investment in China government bonds, which have seen a rally in recent times that has led to multi decade low long term yields. As more investors and institutions pour into bonds, a Silicon Valley Bank esque risk emerges since, if interest rates increased, the value of those bonds would decrease. This could lead to a growth in financial instability and the rapid collapse of banks like what happened to US banks in 2023. 

• However, this outcome is unlikely as Chinese bond demand remains strong and interest rates remain low, with the interest rates related to mortgages and corporate borrowing being cut by the Chinese central bank recently. Nonetheless, it highlights the unenvious position the Chinese economy is currently in. 

Future Events

AI Training Lawsuits

• AI has progressed extensively in capability and availability since OpenAI’s ChatGPT began an AI frenzy. However, this AI growth has been heavily reliant on training data, which is often used in a morally and legally dubious fashion. As a result of this, along with the fear of the economic impact AI may have on certain fields, many individuals and companies that produce content used in training have taken action.

• For example, the New York Times sued Microsoft and OpenAI last year due to their intellectual property infringement during AI training. In April, the Chicago Tribune and 7 other newspapers sued Microsoft and OpenAI as well. Authors like John Grisham, George R.R. Martin, and Jodi Picoult have also grouped together to sue OpenAI. 

• And as AI continues to grow and affect other industries, more companies and individuals will likely seek compensation for their intellectual property being used in AI training. Just last week, authors Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson sued Anthropic, the creator of Claude, in a class action lawsuit alleging that Anthropic used “copyrighted content sourced from pirate websites like Bibiliotik”. The authors want Anthropic to pay proposed damages and to prevent the company from using copyrighted material in the future. If lawsuits like the one against OpenAI or Anthropic succeed, it could trigger a flurry of lawsuits, financial losses, and restrictions on AI training, which would severely hinder the development of AI. 

• While the outcomes of the aforementioned lawsuits are pending, AI companies have taken action to limit future lawsuits by employing content licensing deals. OpenAI has partnerships with Reddit, Time Magazine, the WSJ, the New York Post, The New Yorker, Architectural Digest, Vanity Fair, Wired, and other companies in the last few months, allowing it to stand on firm legal footing for some of its AI training. Unicorn AI company Perplexity AI released a revenue sharing model with media outlets like Fortune, Time, and WordPress.com where, if an article is used in a response, a percent of advertising revenue is given to the media outlet.

US Real Estate

• Like China, the US real estate market has undergone a recent decline as the last few months saw continued declines in home sales and increasingly high home prices. However, unlike China, the causes for the sluggish real estate market are not because of a lack of demand. Instead, it’s mostly a symptom of poor affordability and low supply. Fortunately though, there are some indications that the real estate market may begin to grow faster in the coming months. 

• In July, previously owned home sales increased 1.3% from the prior month, reaching 3.95 million and breaking a 4 month streak of falling sales. July’s sales were the lowest for any July since 2010 though and were down 2.5% from last July, but the trajectory remains a sign that the market is improving. Additionally, when considering the September rate cut and recent falling mortgage rates, it is likely the cost of a mortgage will decrease in the coming months, which would increase home sales. 

• In addition to increased demand, another positive sign for the real estate market is supply. There were 1.33 million US homes for sale or under contract in July, up 0.8% from June and 19.8% from July 2023. This means there is a 4 month supply of homes, the low end of a balanced market for buyers and sellers. With both increasing supply and demand, the US real estate market seems to have a positive future. 

Presidential Promises On Taxes

• This story serves as a continuation of last week’s because, as the election approaches, understanding both candidate’s economic policies will be important. In terms of taxes, both the Trump and Harris campaigns have made significant promises. And even though they are not set in stone or necessarily achievable without an electoral victory in Congress, they are still worth covering to understand the possible direction the US is heading in. 

• Some of Harris’ key tax positions are… 

  • Increasing the corporate tax rate from 21% to 28%, which is estimated to bring in $1.3 trillion over the next decade. Prior to Trump’s 2017 tax cut that reduced it to 21%, the corporate tax rate was 35%.

  • Increasing taxes on stock buybacks from 1% to 4%.

  • Forcing companies to pay taxes worth 21% of the income they report to investors, up from the current 15%. This would prevent companies from shrinking their tax liabilities via deductions or tax credits.

  • Increasing the top marginal income tax rate to 39.6%, up from 37%. Combined with an increase in medicare surtaxes from 3.8% to 5%, top earners could see a top marginal rate of 44.6%

  • Taxing the assets at the owner’s death, which would help counter the inheritance loophole for taxes where taxes are only calculated based on the time of inheritance and the final sale of assets.

  • Americans making above $1 million would see investment earnings taxed at the same rate as regular income instead of the lower capital gains rate.

  • Those with over $100 million in wealth, a vague descriptor, would have to pay 25% on income and unrealized capital gains, the value of the appreciation of unsold assets like stocks, bonds, and real estate.

  • A more generous child tax credit, which could cost around $2 trillion over the next decade. 

  • Maintaining some of the measures from Trump’s tax cut, particularly for those making less than $400,000.

• Trump’s tax promises are a bit more vague as he has not laid out an economic plan, but some important fragments are…

  • Extending his 2017 tax cuts, with the possibility of further cutting taxes, which is anticipated to add $4 trillion to the deficit over the next decade. However, this measure could help spur economic activity and growth.

  • Similarly, Trump has floated the idea of cutting the corporate tax rate, possibly down to 15%.

  • Eliminating taxes on tips and social security, which is expected to add over $1.2 trillion to the deficit in the next 10 years. Again, this could increase the personal income of households and spending, but its cost is harrowing.

• Of course, these policies mean nothing until they’re implemented, but analysis by the Penn Wharton Budget Model suggests that Trump’s economic proposals could add up to $5.8 trillion to the deficit over the next decade while Harris’ could add $1.2 trillion. Both of their plans, however, are dependent on control or compromise, barring the unlikely scenario that one party wins in both houses of Congress and the Oval Office.

Weekly Question

True or False: Cheyenne, a city in Wyoming of 65,000 residents, elected a candidate promising to delegate decision making to an AI

False. While Victor Miller did run on that promise, he received just 3% of the vote and OpenAI shut down his AI. While just a funny story, it does raise the serious concern of the role AI will play in future decision making, whether it be personal, corporate, or political.