What to Do When the Entire Market Looks Expensive

Plus: AI music is now charting weekly  ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌

Read Online In partnership with 90% In Finland, nearly 90% of people say they trust others — one of the highest rates globally. Only 30% of Americans agree that "most people can be trusted." How to invest in a market where nothing looks cheap America's housing crisis is somehow getting worse How China is pulling ahead in AI Newsletter Exclusive: Will AI reshape the music industry? When Even Defensive Stocks Look Risky, What's the Move? Last week, NYU professor Aswath Damodaran, Scott's favorite blue-flame thinker on valuations, said that for the first time ever, he's thinking about moving his money into cash and collectibles — saying there's no place to hide in the stock market. He's not alone. Goldman Sachs strategists predicted last week that U.S. equities will likely underperform global peers for the next decade. They recommended that investors "Diversify beyond the U.S., with a tilt toward emerging markets." The S&P 500's forward P/E ratio is 22x, above its historical average of 17x. The index has traded above 22x only twice since 1985: during the dot-com bubble and the COVID-19 pandemic. The market fell sharply both times. Even traditionally boring, "defensive" stocks look expensive. Consumer staples like Walmart and Costco — mature, low-margin businesses — are trading at Nvidia-level valuations. Alternative investments are also at all-time highs.  Gold is up more than 55% this year, on pace for its best yearly performance since 1979. Bitcoin reached an all-time high of $126,251 in October 2025. So … what should investors do? First, don't panic and try to time the market. That almost never works. Over the past 35 years, missing just the 10 best days would have cut your total return in half. Instead, think long term: Ten-year returns for the S&P 500 have been negative just 6% of the time since 1929. Invest on a schedule: Set a fixed amount to invest regularly. This is called dollar-cost averaging, and it reduces the risk of buying at a peak. Diversify: The classic 60/40 portfolio (stocks/bonds) is a solid baseline; younger or higher-risk investors can lean more heavily into equities.  Within stocks, diversify globally: Roughly 25% to 30% in developed non-U.S. markets is a good starting point, and the MSCI World ex-USA Index is a straightforward option. For U.S. equities, this may be a good time to consider equal-weight S&P 500 ETFs. Goldman analysts expect them to return 8% annually in the next decade, beating the traditional S&P 500 by 5 percentage points per year. For bonds, consider the iShares 0-3 Month Treasury Bond ETF, which tracks the performance of short-term U.S. Treasury bonds with a low expense ratio.     There have been 10-year periods in U.S. stock market history with negative returns. In the 10 years before 2009, U.S. stocks lost about 37% of their value. Also in the decade before each of these years: 1974, 1939, 1921, 1857, and 1842.  It's very infrequent, and it's very unlikely in the large scheme of the U.S. stock market — but it is possible.  So even if you disagree with us about the current market valuation, you should at least consider the possibility that over the next 10 years, the stock market could have a negative, or at the very least lackluster, return.  How do you prepare for that possibility? Start to think about de-risking. If you're overexposed in any one company, start to trim in very modest increments over time — not to eliminate the position but just to make the portfolio more balanced. You shouldn't have more than 5% of your portfolio in any one stock.       I'm contemplating buying another home in Manhattan, because of all of my assets, it's been the worst-performing one. I think the existential threat of the new mayor has been vastly overrated. All these idiots who say, I'm leaving. I'm like, OK, fine, see ya. New York is still singular.  The greatest arbitrage in economic history is really talented young people who are willing to work their asses off and take risks. That's the secret sauce of any economy, and there's more of that secret sauce poured over Manhattan than any place in the world.     What Happens When Three-Quarters of Households Can't Afford a Home Somehow, the U.S. housing crisis is getting worse. New data revealed that the median first-time homebuyer is now 40 years old, the oldest on record, and first-time buyers accounted for just 21% of purchases last year — roughly half the historical average. In 1985, the median home cost roughly 3.5x the median income. Today, the median home costs more than 5x household income.  Nearly 75% of U.S. households cannot afford to buy a home in 2025. That's up from 56% in 2002. The Trump administration is pitching a 50-year mortgage as the solution, claiming it would make ownership more affordable for younger buyers. But that wouldn't fix the underlying supply problem. The U.S. is short between 1.5 million and 5.5 million homes, roughly 3.7% of the current housing supply.  Congress has proposed some solutions. The "ROAD (Renewing Opportunity in the American Dream) to Housing Act," a bipartisan bill that just passed the Senate, offers grants to communities that build more affordable housing, removes regulatory barriers to construction, and modernizes financing programs such as the Low-Income Housing Tax Credit.  The bill is now being considered in the House of Representatives before it is passed on to the president.  The private markets are working on solutions of their own. Modular construction involves building components of homes off-site and assembling them at their final destination.  This process reduces building costs by 20% to 30% and can speed up construction timelines by up to 50%.  Modular construction made up 5% of total U.S. construction and 9% of apartment starts last year.    The Trump administration's 50-year-mortgage proposal makes no sense.  Take a $450,000 mortgage. For a 30-year mortgage at a 6.25% interest rate, your monthly payment would be around $2,800. Over the lifetime of the loan, you'd pay about $547,000 in interest.  Now, take a 50-year loan at the same rate. Your monthly payment would fall to around $2,450, so you'd be saving about $350 a month. Sounds great, right? Not really. In total, you'd pay about $1 million in total interest — roughly 87% more than the 30-year loan. What we really need to do is make it easier to build. This is mostly a regulatory problem that we seem to be fixing. Most of this year's New York City ballot measures were related to affordable housing, and they passed.  Mamdani also won the mayoralty, and he ran on the affordability crisis. But his solution of "rent freezing" doesn't work. It disincentivizes construction, sending rents everywhere else up. A good example of this policy not working is Argentina. One of President Javier Milei's best moves was scrapping all the rent controls in Buenos Aires. Since he did that, supply has nearly tripled and rents have fallen by about 40%.       Sweden has demonstrated how modular construction can actually work. Nearly half of new homes are built in factories, and its housing shortages are now milder than in many peer countries.  Lindbӓcks, a leading Swedish modular home manufacturer, uses assembly lines to produce more than 25,000 square feet of housing per week (that's the equivalent of 28 average-sized New York City apartments). Regulatory process is also key to the Swedes' success. The Swedish government lets builders figure out what materials should be used and how, and regulators check their work.  In the U.S., building codes mandate specific materials and methods. The Swedish performance-based system makes building more cost-effective and expedient, compared with our prescriptive approach.     ____________sponsored content ____________ The only company recycling 400+ waste streams TerraCycle builds profitable recycling models for "unrecyclable" products and packages by working with hundreds of major brands, like Coca-Cola, Gillette, Whole Foods, and Taco Bell, to recycle razors, sauce packets, coffee pods, and so much more. TerraCycle has been profitable for 10+ years, turning their mission-driven recycling solutions into revenue-driving success; increasing revenue 75% between 2020-2024. First-half 2025 revenue is up 11% vs. last year. Previous investors have already seen between 17–20% of their investment back in dividends. Now it's your turn. Invest in TerraCycle by 11/20 and get 15% bonus shares. Invest in TerraCycle This is a paid advertisement for TerraCycle's Regulation CF offering. Please read the offering circular at https://invest.terracycle.com/ ____________sponsored content ____________ China's AI Models Are Getting Cheaper — and Better A Chinese startup called Moonshot AI just released a new open-source model, called Kimi K2 Thinking, that outperformed OpenAI and Anthropic models across several benchmarks. Plus, it's 4x cheaper to run at scale than GPT-5.  A week later, Alibaba announced it would cut the costs of its biggest AI model, Qwen3-Max, in half.  Airbnb CEO Brian Chesky has said that his company relies heavily on Qwen because it's cheaper and faster than American models.  Nvidia CEO Jensen Huang has suggested that China's recent gains in AI reflect their structural advantages. He told the Financial Times: "China is going to win the AI race," citing the benefits of their looser regulations and cheaper, abundant energy. Moonshot AI took off after launching its first ChatGPT-style chatbot, Kimi, in October 2023. The startup, now valued at $3.3 billion, is backed by Alibaba, Tencent, Meituan, and HongShan (formerly Sequoia China).  Moonshot's newest model is more efficient than ChatGPT and other American models because of its Mixture-of-Experts design. Instead of the whole model activating for each task, only a few smaller parts turn on at a time.  In addition to being more efficient, Kimi K2 Thinking is reportedly less sycophantic than other AIs. In a recent Reddit AMA, the Moonshot team said that was an intentional design choice.   The most important part of this story is the price disparity: Kimi K2 costs $2.50 per million output tokens — that's 4x cheaper than GPT-5 for comparable performance. Chinese AI models from Alibaba and DeepSeek are also roughly 10x less expensive than American alternatives.  Why are Chinese AI models so cheap? Energy. China produces 2x more wind power, 3x more solar, and 6x more hydroelectric power than America. That, plus government subsidies, makes their electricity half as expensive as ours.   This administration is only making our energy problem worse. They're actively gutting renewable projects and making the permitting process way more difficult. If you want to build more solar energy in America today, there's a new rule where you have to get a personal sign-off from Secretary of the Interior Doug Burgum. It's just a bureaucratic nightmare.       The reason we don't produce enough energy is we have politicized energy. Republicans have somehow figured out a way to feminize renewable energy, playing into Trump's distorted version of masculinity where coarseness and cruelty signal strength, while anything progressive gets dismissed as weak or feminine.  China doesn't gender their power generation. They're building renewable capacity at scale because they want cheap energy to maintain their competitive edge.  Our gendered framing of renewables is hurting us. Oil prices are genuinely unpredictable. It could spike to $200 a barrel or crash to $10. But it's a fairly safe prediction that solar and wind will be cheaper in a decade than they are today. That's not "woke ideology," it's economics. So here's the question: How do we make smart energy policy patriotic again?     Newsletter Exclusive: AI Music Is Now Charting Weekly Three AI-generated songs topped the Billboard and Spotify charts last week. It wasn't the first time. At least one AI or AI-assisted artist has charted each week of the past month.  In September, Hallwood Media signed a reported $3 million contract with Telisha "Nikki" Jones, the lyricist behind Xania Monet, an AI-generated artist.  According to Billboard, there was a bidding war. Jones, a poet from Mississippi, creates the music by prompting an AI music generator called Suno. The company, which is being sued, has acknowledged in a court filing that it is using music found all across the internet to train its model. Major record labels are also getting into the AI-generated music market. Universal Music Group (UMG) recently settled a copyright lawsuit with AI music creator Udio and agreed to co-build a commercial AI music-creation-and-streaming service. The product, which launches next year, will compensate artists and give subscribers access to a model trained on UMG music.  More deals are expected soon. According to the Financial Times, Universal and Warner are in talks with Spotify, Google, and AI startups, including ElevenLabs and Stability AI. Dealmakers insist that artists have nothing to worry about. Romel Murphy, Xania Monet's manager, told CNN: "AI doesn't replace the artist. That's not our goal at all. It doesn't diminish the creativity and doesn't take away from the human experience." Creativity aside, artists should be worried. The threat of AI-generated music is that it will dramatically increase the volume of music on streaming platforms, making it even harder for emerging artists to break through and make a living.  It's already challenging enough. Twelve million artists uploaded songs to Spotify last year. Less than 0.2% made more than $50,000.  One study predicted that people working in the music sector will lose almost a quarter of their income to AI within the next four years. The same study predicts that AI music will account for about 20% of music streaming platforms' revenues by 2028. Can moral objections to machine-made creativity influence the economics of music? Listeners care about whether a song is AI generated, but there's little evidence that the industry does.  Most people say they prefer human-generated music, and some say they would skip AI music altogether.  One study of 9,000 people in eight countries found that 40% said they would skip AI-generated songs entirely. The problem is that 97% could not distinguish between AI-made and human-made music.  Will streaming platforms label AI-generated music clearly enough to make a difference? Spotify says it will adopt an industry standard for identifying and labeling AI music in credits — but will those disclosures be hidden in the fine print? Should it even matter? What if AI-generated music is just its next evolution, like electric guitars and DJs? Music is a human experience … but AI has become a human experience too. There is going to be crazy sh*t coming out of the White House over the next seven days. With the Epstein files coming out, they need a nuclear weapon of mass distraction.   In the latest episode of China Decode, Alice Han and James Kynge unpack China's energy dominance and a new pilotless electric air taxi from EHang Holdings that can fly more than 100 miles on a single charge.   40% of young women say they would move out of the U.S. permanently if they had the opportunity  Excellent charts from Michael Cembalest on the shift from old to new media  Resy's 2025 retrospective on the top dining trends    Comment on this post   Update your email preferences or unsubscribe here © 2025 Prof G Media. All rights reserved. 382 NE 191st St Miami, FL 33179, United States of America Terms of Service