November 18, 2026
What You Need to Know Today:
What were these bankers thinking when they made these AI loans?
On Monday, the Trump administration announced a new rule that would require data center operators in Singapore, Malaysia, Indonesia and Thailand to obtain licenses to purchase advanced chips made in the US. The change takes a significant step toward closing a loophole that has been an important driver of Chinese demand for computing power outside the mainland.
Banks and private credit funds that lent billions to fund data centers and AI chip purchases in Southeast Asia are being forced to confront a long-feared scenario: the loss of critical Chinese customers. The threat that Chinese companies could be cut off from Southeast Asian data centers has long loomed over dealmakers, but financing has continued despite the risk. Data center companies based in Asia Pacific have raised about $49 billion in loans since 2023, according to figures compiled by Bloomberg. Customer contracts were often used to secure those deals, with chips themselves sometimes used as collateral.
It’s not just AI or tech: bank stocks near a technical death cross
Shares of large U.S. banks are close to forming a rare “death cross” technical pattern, a bearish momentum signal. That’s not good. Even if it is understandable. Disruptions like that at JPMorgan Chase and rising borrowing costs have crushed sentiment toward the sector. The KBW Bank Index fell as much as 2.4% on Tuesday, sending the index’s 50-day moving average to almost cross below its longer-term 200-day moving average. Chart watchers on Wall Street keep tabs on this downtrend pattern as it may indicate more market weakness, at least in the short term. On Monday the Invesco KBW Bank ETF (KBWB) closed just above a death cross with its 50-day average only about 0.7% above its 200-day average.
What were these bankers thinking when they made these AI loans?
On Monday, the Trump administration announced a new rule that would require data center operators in Singapore, Malaysia, Indonesia and Thailand to obtain licenses to purchase advanced chips made in the US. The change takes a significant step toward closing a loophole that has been an important driver of Chinese demand for computing power outside the mainland.
Banks and private credit funds that lent billions to fund data centers and AI chip purchases in Southeast Asia are being forced to confront a long-feared scenario: the loss of critical Chinese customers. The threat that Chinese companies could be cut off from Southeast Asian data centers has long loomed over dealmakers, but financing has continued despite the risk. Data center companies based in Asia Pacific have raised about $49 billion in loans since 2023, according to figures compiled by Bloomberg. Customer contracts were often used to secure those deals, with chips themselves sometimes used as collateral.
Part 2, the election and–AI stocks
I’m not sure that the stock market, yet, fully appreciates the pressure that the election put on AI stocks. And it couldn’t come at a much worse time, when Wall Street is questioning when AI companies, especially data center companies, will start producing the profits to justify their enormous capital spending. The pressure comes from two directions. From Congress and from the states. I’m much more worried about the impact of the elections at the state level right now because I think it will have a more immediate impact on the timing of revenue growth at hyperscalers and data centerr companies.
Saturday Night Quarterback says (on a Sunday), For the week ahead expect…
This coming week will be all AI, all the time. On Tuesday Nvidia (NVDA) reports earnings.
On Wednesday, Anthropic’s record-setting IPO is expected to prIce with the first trade of ANT on Thursday. Think we might get a little market volatility one direction or the other?
I’m buying on the dip–but gold and silver, not tech stocks
If you’re a short-term trader maybe you can make a case for buying AI and chip stocks on the recent rout. If you’ve got a time horizon that stretches for longer than the next few weeks or months, I think the dip you want to buy is the big one in gold and silver.
And then there was one–maybe–AI company with a path to profits
In the distant past–you know like maybe 6 months ago, the conventional Wall Street wisdom divided AI stocks into two camps. There was the large group that included exciting shoot-the-moon companies like OpenAI, Anthropic, and Coreweave that were out on the leading edge of these technologies. They might well capture the future of a global economic revolution. On the basis of that promise, they had raised hundreds of billions of dollars to build out the infrastructure of AI. The big question was Would they generate enough profit to pay the interest on all that debt or would they run out of new investors to cover their losses before they generated substantial profits
And then there was a smaller group of big Internet companies–you know the names: Microsoft, Amazon, Alphabet, Meta Platforms–that stood to make huge profits building the “compute” that AI needed, but who had significant other businesses that generated huge positive cash flows so that they could finace their AI build out internally. In the last week or so, however, this wisdom has run into a huge problem.
This market is just too damn expensive given the risks–Special Report 10 picks to stay (mostly) invested while cutting your risks
Could stocks move higher from here even though they’re more expensive than a year ago? Of course.
Could they gain another 29.8% in a year? It’s possible. But it strikes me as unlikely.
That sums up the current market puzzle very neatly, I think. Stocks are expensive and don’t seem to be pricing in potential very real risks. But stocks have returned almost 30% in the last year. And it’s very hard to walk away from the possibility of that kind of gain. Even if it is unlikely. What I’d suggest right now is a stock market strategy that takes into account both a solid appreciation of the excessive valuations in this market (along with a failure to price in risk), and the realities of investor psychology.
It is too difficult to go cold turkey on stocks right now and move 100% out of the market. On the other hand, you should be moving to reduce your exposure to the riskiest parts of the market–high multiple stocks–and raising your exposure to counter-cyclical or non-cyclical stocks. In this post, I’m going to give you 10 stock picks to use in executing that kind of strategy–a way to stay in the market and yet to reduce your risks.
Special Report: The Next Big Things and how to invest in them–Part 1 Quantum Computing; Part 2 Nuclear Fusion
A suggested quantum computing portfolio. If you want a piece of this Next Big Thing, but with less risk and less upside than a pure-play quantum stock, I’d suggest Alphabet/Google (GOOG). Among pure plays I’d include D‑Wave Quantum (QBTS), up about 235% year‑to‑date as of late 2025; Rigetti Computing (RGTI), up34% YTD by late December; and IonQ (IONQ), up around 25% year-to-date by late December.
Special Report: 3 Stock Market Bubbles: When will they burst? What to do now? Part 1, the AI bubble, Part 2, the debt market bubble, and Part 3, the cheap money bubble
This is a very difficult stock market. Even as stocks climb to new record highs. On the one hand, even investors who are all in, maybe even overweight to the long side, worry that this rally isn’t sustainable for much longer. By most historical standards valuations are off the charts. I get a steady stream of stories and posts asking whether XYZ stock has climbed to faro fast. Volatility on somedays can be downright scary with relatively minor events leading to big market moves. It’s simply very hard to stay on board this rally. On the other hand, it’s very hard to get off the train. I see lots of Wall Street analysts cutting recommendations from “buy” to “hold” on valuation fears, but I see almost no one saying “sell.” FOMO–fear of missing out–is just too strong. Which is totally understandable. The Standard & Poor’s 500 index was up 25.02% in 2024 and was up another 18.11% in 2025 to date through October 27. Market leaders have racked up even bigger gains. AI chip icon Nvidia (NVDA) was up 171% in 2024 and has gained another 39% in 2025 through October 27. It’s insanely difficult to walk away from those kinds of gains. So what’d you do?
Special Report: How to invest in our 3 energy crises–First 8 picks JCI, BEPC, LNG, SMNEY, GNRC, CCJ, EQNR and GVE
You don’t need the Department of Energy or the Energy Information Administration to tell you we have an energy crisis. (Good thing since they’re shut down with the rest of the Federal government today.) All you need to do is look at your electricity bill. This summer monthly home electric bills jumped in Trenton, New Jersey, for a typical home by $26. In Philadelphia, it increased about $17. And in Columbus, Ohio, it spiked $27. And your monthly bill doesn’t capture the full damage. In California,residential electric rates are up 62% in five years. In Maryland residential rates are up 54% in five years. Most frustratingly–and most importantly for investors–those bills don’t explain the nature of the crisis.
Or more accurately “crises.” Because we’re the middle of three, overlapping and interlocking energy crises. That are playing out on different timeframes that range from NOW to the next 5 to 10 years. It’s that last point that’s critically important for investors. Because to make money–and let’s be clear: like in all crises there’s money to be made investing in these three crises–you’ve got to understand the nature of each crisis and buy into it at the right time. Not so early that you sell in disappointment because your profits haven’t arrived yet. Not so late that all themes tasty profits are gone. This Special Report is about untangling the 3 energy crises, giving you a timeline for investing in each, and then calling out 10 picks you cause to profit from theses crises. Ya, ready?








