"The portfolio you can actually stick with will always outperform the one that you're gonna abandon at the wrong time." "When you add other things that are a little bit outside of your domain of expertise, start small, build intuition, operate from a position of strength." Mike Philbrick on the one lesson he would teach the average investor.

More than half of Andy Constan's work goes into figuring out how his investment thesis could be wrong. "The number one thing I wanna do is just evaluate where I could be wrong." Considering the opposite case can change his mind, strengthen his view or change how much confidence he puts in it. "I can't imagine a scenario where I could blindly assume I'm right and expect to survive."

Producer prices rose 6.3% over the year to September. Should your bond allocation be judged only against inflation? In an earlier conversation, Cullen Roche described nominal bonds as "principal stabilizers": "They should be owned to create principal stability across very specific time horizons." For him, that means "probably high quality and shorter duration" to create "a pretty high level of predictability."

A quarterly earnings number matters only if it changes what the business is worth over the long run. "What it makes in a particular three months is not life-changing." Colin McQueen holds stocks for a little over four years on average. He looks for "material information" about future earnings, then asks: "What is it worth, not how it's gonna react to the news flow."

Cutting risk after losses can quietly weaken a good trading strategy. "For a long time, if I lost money five days in a row, I'd be, 'I better take it easy today, go to 30%.'" Brent Donnelly realized he was reducing risk after drawdowns and increasing it after unusually good runs. "If you have a good process ... and you know the difference between variance and actual bad trading where you're not following your process, then you can maintain a way steeper slope of your P&L."

Tech spending doesn't have to collapse for tech stocks to lose leadership. Jim Paulsen compares technology-related spending with its long-run trend. His concern is the speed of the recent expansion: "It's just too extreme, too much, too fast." "Even if it stays fairly strong, if it just is less strong, I think that's gonna have an impact on leadership."

An AI chip can still work and already be uneconomic. "It's not that the chip broke and stopped working. It's uneconomic relative to buying more of these." Rob Arnott draws on Chris Brightman's research: what if newer chips make old ones uneconomic after three years, while the purchase is depreciated over six? "I'm not saying the AI revolution isn't real. It's very real."

"We're never gonna run out of data to process." Ed Yardeni sees data as a new factor of production alongside land, labor and capital. Cheaper processing, he argues, creates more demand for computing. "The question is: how do you make money on that?" His approach is to own the Nasdaq-100: "Some companies in there are really gonna make a lot of money" and "the other ones will just go away."

"I think it largely gets the direction right, but the magnitude wrong." Expensive companies can grow faster and still fail to justify their prices. With earnings expectations helping push stocks to record highs, Cliff Asness's research asks the relevant question: did growth over the next three or five years match what investors paid for?

Governments and the AI buildout are competing for the same pool of money. "They all are enormous sponges for capital." Ben Hunt on the crowding out effect that AI borrowing creates.

A passive portfolio can hide an active bet. "You're really actively choosing to bet the farm on one thing." Jason Hsu worries that investors buying the S&P 500 for diversification may underestimate their exposure to the same AI growth story.

Demanding revenue too early can damage the business you invested in. Jack Raines noticed that some pre-revenue investments in his seed-stage fund performed better than businesses already making sales. Pushing for a top-line number too soon "might actually kill the ability for the rocket growth later."

If SpaceX joins the S&P 500, index funds could rush to buy from holders who do not want to sell. "Like a herd of elephants trying to get through a revolving door at the same time." Rob Arnott argues that funds trying to match the index's closing price can crowd into the same trade. Minimizing tracking error does not necessarily mean getting a good purchase price.

"All too often, people assume that something as complex as financial markets requires a complex solution, and it really doesn't." "The key to success is just minimizing mistakes and not interrupting compounding." Peter Lazaroff on the one lesson he would teach the average investor.

Tom Maher on when basket trades overpower company fundamentals. "People hate company B, and this is in a basket with company B, so it's getting dragged down." These moves can persist for "six, nine, 12 months, not one or two days."

Last week's AI selloff puts an old investing problem back in focus: a technology can change the world without rewarding the people funding it. Ben Inker's earlier warning: "You get way too much investment and you destroy that ROI." "It doesn't mean that the technology doesn't change the world."

Better or worse may matter more than good or bad. Jim Paulsen on why changes in expectations matter more to asset prices than whether the economy looks strong in isolation.

"The surefire winners of the internet boom were Yahoo, AOL, and Cisco." Chris Davis sees a warning in treating today's AI leaders as inevitable winners.