Did The Market Just Break? | Lance Roberts

Adam Taggart opens by describing a painful week for technology stocks. About $4 trillion in market capitalization has been erased from the Magnificent Seven, four of those seven stocks are in bear-market territory, and more than 60% of S&P 500 stocks are at least 20% below their yearly highs. Lance Roberts agrees that the weakness is real but says it is concentrated in growth and technology rather than spread evenly across the market. Intel was down nearly 30% after earnings, while Google, Amazon, and Meta were also under pressure; in contrast, Eli Lilly, Walmart, Costco, Berkshire Hathaway, and value stocks generally performed well. Lance says the dispersion reflects a rotation that began October 5, when his model reduced growth exposure and added value. Over the week, value stocks were up roughly 1.5%, while growth stocks were down about 1%. Only 32% of stocks were advancing, but the equal-weighted index and measures of stocks above their 20- and 50-day moving averages were holding up better than the headline indexes. He attributes part of the sharpness of the selloff to margin debt—described as around $1.2 trillion—options trading, leverage, and zero-day-expiration options. Google was more than four standard deviations oversold, which Lance views as potentially contrarian and closer to a buying opportunity than a reason for a panic sale. He cautions investors against selling after a sharp decline, buying back after a rebound, and then being caught in a retest.

Adam asks whether the technology selloff reflects the familiar end-stage pattern of a hyper-bull market: companies continue reporting strong earnings, but the results are not good enough to justify extreme expectations. He also raises a more specific concern about artificial intelligence cannibalizing Google’s search business. AI systems increasingly give users direct answers instead of sending them through blue links that generate cost-per-click advertising revenue, and Google’s latest earnings discussion appeared to show some of that effect. Lance says investors should be careful with broad valuation arguments because valuation differs by company. He cites Google as an example: despite the selloff, it was growing revenue about 25% year over year, trading at roughly 18 times forward earnings, and had a forward price-to-earnings-growth ratio near one. He says the bigger problem is that Wall Street kept ratcheting expectations higher rather than lowering them before earnings, so strong results still disappointed. Adam counters that Google may be inexpensive by some measures but remains dependent on a search business that historically gave it an extraordinary moat. Google search generated roughly $243 billion in the latest figures Lance cited, alongside about $77 billion from cloud, $51 billion from subscriptions, $42 billion from YouTube ads, and $30 billion from the Google network. Lance responds that search is now about half of revenue rather than the entire business, and Google has a long record of turning capital allocation into revenue. He expects the company could eventually attach advertising to Gemini or otherwise use AI across its ecosystem, though Adam emphasizes that Google faces much more competition in AI than it did in search.

Although Lance’s model moved from growth to value in early October, he says it is approaching a possible reversal. Value has become extremely overbought while growth is extremely oversold, so the model could switch back toward growth as soon as the following week, although Lance stresses that its technical rules—not his personal forecast—will determine the timing. He wants to get through the election first, after which the market will have a known result regardless of the outcome. Lance identifies several potential year-end tailwinds: corporate buybacks may return, earnings season will be behind the market, liquidity is typically stronger into year-end, and portfolio managers may rebalance so that their books show the companies that performed well. He says 72% of companies were guiding earnings higher as of Monday, an all-time record in the data he cited. Third-quarter economic growth was roughly 3.2% in the first estimate, although he notes that the first GDP estimate largely compiles Wall Street economists’ projections and is often revised lower. Consumption remained around 70% of GDP and was essentially unchanged from the prior quarter; the growth instead came largely from imports of AI and semiconductor equipment and from business capital expenditure. If the technical signals line up, Lance would trim value stocks back to target weights, add technology exposure, and potentially raise equity exposure from 60% to roughly 65%–67% going into year-end. He repeatedly says this is a conditional plan, not a guarantee.

Adam shifts from AI as an investment theme to AI as a work tool. He says he recently began using Grok agentically rather than merely as a search engine or writing assistant. He gives the system descriptions of recurring business tasks, asks it to perform them on a schedule, and lets it alert him when a problem requires his attention. In one example, Grok examined a prior daily output, reproduced it using the same inputs, and agreed to complete a sequence of tasks before notifying Adam for review. Adam says this frees him to spend more “meta time” on decisions such as choosing future topics, planning interviews, and developing Thoughtful Money. Adam’s conclusion is that people who refuse to understand AI risk being left behind, comparing them to buggy-whip manufacturers who ignored automobiles. He also says the experience made him question why businesses would hire inexperienced workers for routine tasks when AI can perform much of that work faster, more accurately, and cheaply. Lance cites an MIT study involving students, professors, and research assistants: students may use AI without learning the underlying material, professors may stop assigning homework because it can be generated automatically, and researchers may no longer need assistants for basic research. Adam describes this as eroding the lower rungs of the career ladder, because internships and entry-level work are where people traditionally gain skills. Lance says he still writes his own articles but uses AI to find sources, challenge his assumptions, and test research. He believes that debate reduces confirmation bias and has improved his portfolio-management work. He says AI helped him build and backtest a family-office portfolio model in roughly weeks rather than the two or three months the process might previously have taken.

Adam asks about the reported Binance-related hack, noting that the incident occurred while technology stocks, indexes, precious metals, and Bitcoin were all under pressure. Lance clarifies that Bitcoin and Ethereum themselves were not hacked; the vulnerability involved an exchange and storage solution. He says the apparent North Korean theft highlights weaknesses in the surrounding ecosystem and the difficulty of tracing funds because crypto’s intended decentralization can also make policing harder. Lance does not believe the incident materially changes the convictions of crypto supporters, many of whom still believe crypto will replace fiat currency. However, he cites roughly 150 major hacks and protocol exploits in 2025, according to the security firm TRM Labs, in addition to personal wallet compromises over the prior five years. He expects AI to make cyberattacks more sophisticated and says investors should consider cybersecurity exposure. His own approach is not narrative-driven: his firm uses a crypto model that responds to whether the asset is rising or falling, and crypto is currently one of its worst-performing thematic exposures. He makes the same point about portfolio diversification. Gold and silver, previously a hot trade, were hit particularly hard, while normal diversifiers have not always behaved as expected because inflation, oil-price shocks, real rates, and other forces have disrupted traditional relationships. Lance warns investors not to stay attached to a losing position simply because they hope it will recover. Instead, they should assess where capital is flowing and move toward stronger opportunities, with the expectation that they can return to a weak asset later if its behavior improves.

Adam notes that bonds have recently done what investors traditionally expect: they provided stability while equities struggled. He says investors often dismiss a 5% bond yield when stocks are delivering double-digit gains, only to appreciate that income when equities fall. Lance agrees that bonds are returning to their portfolio role after a period in which rising rates made them look ineffective, but he cautions against chasing them immediately. He says he reduced a longer-duration bond position from 5% to 4% after rates fell and bonds became technically overbought. The 10-year Treasury yield was around 4.95% and approaching resistance, while upcoming inflation data could temporarily push yields higher and bond prices lower. Lance recommends patience for investors seeking duration or current yields, arguing that a better entry could develop over the next couple of weeks. He also directs listeners to educational bond material at Real Investment Advice, including an e-guide and articles on bonds at 5%, the 10-year Treasury as a potential buy, and the role of bonds in a balanced portfolio. He argues against replacing the bond allocation in a 60/40 portfolio with Bitcoin or gold because bonds are intended to provide relatively low-volatility safety and income. For a family-office portfolio he is building, he is using municipal bonds because the client wants tax-free income. His broader recommendation is to create an investment policy statement before investing; without a blueprint, investors accumulate a random collection of assets and often panic when the market changes.

Lance’s technical analysis places the S&P 500 near its 100-day moving average after trading in a broad range since June. The index has gone largely nowhere for five or six months despite repeated AI and macro headlines. He explains that two standard deviations contain roughly 96% of potential movement and three standard deviations about 99%, so extreme moves can signal overbought or oversold conditions. The market is not fully oversold yet, but momentum has weakened and much of the earlier overbought condition has been worked off. The key level for the following week is whether the S&P can defend its 100-day moving average. A breakdown from the narrow consolidation could expand the Bollinger Bands and produce a retest of the 200-day moving average near 7,300. A successful defense followed by a move above roughly 7,630 could support a year-end rally toward 8,000–8,100. The mega-cap ETF MGK has suffered visible damage, while the equal-weight index has performed better, reflecting the rotation from growth into value. Lance says his actively managed 60/40 portfolio reached an all-time high during the week even though the broad index was flat; he currently holds about 10%–15% cash and made only one major trade, trimming bonds. He is waiting through the election before making larger changes. The episode closes with Adam and Lance emphasizing controllable responses rather than forecasts. Adam quotes the idea that people cannot control what happens but can control their response. Lance describes putting unsolvable problems “in a box” until circumstances make them actionable. Both argue that preparation should happen before a major downturn, without allowing fear to create the opportunity cost of sitting permanently in cash.