This edition of Market Factors starts with an important interpretation of recent market volatility with significant practical investment implications. Section two covers a remarkable science fair featuring fusion power and moon infrastructure. I confess to an embarrassing hobby in the diversion and there’s quick hits as always.

Wells Fargo strategist Ohsung Kwon made the important distinction between a factor sell-off and a fundamental sell-off. The upshot is that we’re not seeing an AI pullback - it’s more complicated than that.

The common interpretation of recent volatility is that the AI market bubble is wobbling. But if this were the case, why are infrastructure, renewables and rare earths getting hit just as hard? For Mr. Kwon it means that markets are experiencing a sell-off in the momentum factor: stocks chosen for the highest price and earnings growth.

This might sound like an academic observation but it has extremely relevant practical implications for investors. It means that large investors, notably algorithmic trading funds like CTAs (commodity trading advisors, Mr. Kwon pays more attention to these than any major strategist I know), are broadly selling the stocks that were up the most. They were not specifically questioning the fundamentals future of AI stocks.

The strategist noted that June saw 2027 earnings expectations for semiconductors improve by roughly 8.0 percentage points, more than any other sector. This provides further evidence that market volatility in the tech sector was not driven by fundamentals.

Mr. Kwon offered a roadmap to clients for tech earnings season. If the hyperscalers report improved ROI (return on investment) on data centre spending and continued spending, then semiconductors should outperform and the Nasdaq should beat the S&P 500.

If ROI is strong but capex set to slow, hyperscaler stocks will outperform semiconductors and AI-related hardware. Weak ROI, higher capex means semiconductors should outperform the hyperscaler stocks and investors should begin to move assets out of tech. Lower ROI and weak capex means a risk-off environment with most asset prices headed south.

BofA Securities quant strategist Nigel Tupper added to the discussion with some promising market precedents. In a Tuesday report he noted that global semiconductor stocks averaged a 44.3 per cent return in the 12 months after a significant pullback, provided a recession did not occur. Performance averaged 11.9 per cent even if a recession happened.

My main takeaway here is that the tech rally is less threatened than I thought. This does not mean AI stocks won’t be truly tested in the months ahead, but we also do not appear on the edge of a March 2000 precipice.

As a quick aside, I keep seeing stories projecting huge increases in electricity prices where data centres are built. The concern is understandable but for investors I think it misses the point. There’s a new acronym, BYOP, that stands for bring your own power. New data centres won’t just drain the grid like giant parasites, they will also construct power generation facilities. This brings investment opportunities in sectors like natural gas turbines, (GE Vernova GEV-N is printing money selling these, even though its stock got walloped Wednesday after reporting guidance that didn’t quite satisfy lofty expectations), electric power equipment and eventually small nuclear reactors like the ones on U.S. military submarines.

BofA Securities expanded its Transforming World Conference with field trips and conference calls to facilitate what appears to be the nerdiest, most high stakes science fair ever. The potential technological breakthroughs presented are breathtaking but there’s no real way a research report summarizing the event can determine how many are viable.

I’ll ignore the AI-related presentations - we’ve been inundated with that subject for months - except to say that securities analyst Martyn Briggs believes that “the boring phase of AI is over” and that eye-popping applications in physical systems (like robots) are imminent.

It’s hard to pick highlights because there’s too many candidates. A company called Helion outlined a theoretical pathway to fusion energy. Another named Rigetti thinks quantum computing will be available within three years. Astrolab has developed mobility and infrastructure solutions for a moon base. Varda, in the same sector, is pioneering low gravity manufacturing techniques.

The list goes on. Matternet executives argued that drone delivery can significantly reduce shipper costs. VTOLs – vertical takeoff and landing aircraft – could form another layer of personal travel and cargo delivery.

It’s nice to be reminded that while daily newsflow makes it seem like the world is falling apart at the seams, really smart people are working on projects cool enough to improve living standards for everyone.

There are few dumber things to collect than watches – the phone has the correct time right there on the front – so of course I own a few pieces. No Rolexes because I draw the line at watches costing more than economy cars but I own four of them with four-figure values and another two that are close.

I always think of Hannibal Lecter when thinking about watches. In profiling serial killer Jame Gumb he tells Clarice: “He covets. That is his nature.” Watch collectors covet certain models. In my case I’m currently trying to sell my most expensive piece, a Tudor Black Bay 58, to buy a used Omega Seamaster Planet Ocean.

My covetous nature is also drawn to a Cartier Santos. In the 19^th century wristwatches were considered the exclusive province of women. Men used pocket watches stuffed in waistcoats. Louis Cartier, however, had a friend named Alberto Santo-Dumont, a Brazilian aviation pioneer who needed both hands to fly planes. The rest is history.

A used Cartier Santos with a few scratches can be had for C$5,000 which is more than I want to spend. Well, I want to spend but again, the time is right there on the iPhone.

It’s a weird hobby and one I won’t recommend. I got into it by working for a major broker dealer that paid very low salaries and very big bonuses. One year I rewarded myself with a bulletproof Sinn Diapal 756 after paying down my monstrous American Express balance. I still love that thing. But …. just don’t.

Looking for our updates on market movers, analyst actions, stock technicals, insider trades and other daily, weekly and monthly insight? Click here to visit our Inside the Market page.

Ted Dixon outlines TSX stocks with both share buybacks and insider buying in the first half of this year

How a financial services worker built up a $875,000 TFSA using a diversified selection of stocks as well as some unique portfolio manoeuvres

The AI investment boom is putting Big Tech’s free cash flow under pressure

The Globe’s Mark Rendell details how Trump’s latest tariff threats open a new front in trade negotiations

Higher bond yields bring foreign investors and higher currency values. If, for instance, Canadian bond yields were attractive relative to other developed nations, foreign investors would be buying Canadian dollars in foreign exchange markets and using them to buy bonds. Currently, domestic short-term bond yields are low relative to the U.S. (versus the historical average) and thus the loonie is not attracting bids in FX markets. BofA Securities strategist Alex Cohen is recommending his clients short sell Canadian dollars (or government of Canada bonds, it depends on the fund manager) to buy Japanese yen (or yen-denominated bonds). It’s one thing to not be attractive as a currency but another to be a funding currency to buy other countries’ assets.

There is one type of technology spending that is almost guaranteed no matter what the overall market conditions look like: cybersecurity. Stocks in this sector have been on fire recently, with many climbing more than 20 per cent in the past month, according to Morgan Stanley analyst Meta Marshall. He believes stock valuations are stretched on average but that revenues will continue to expand quickly in the third quarter. Morgan’s top picks are Palo Alto Networks (PANW-Q) and Crowdstrike Holdings (CRWD-Q). Mr. Marshall raised his target on Palo Alto Networks from US$320 to US$387 on Tuesday and also raised the Crowdstrike target from US$172 to US$227.

SpaceX (SPCX-Q) IPO’d at US$135, jumped to US$200 before falling below US$125. Companies and their financial advisers are always trying to strike while the iron’s hot to get the best stock price. This is not great for investors – the highest stock price means the company’s value is as close to peak as the issuer can manage, and thus declines from opening price are not only frequent but pretty much the rule. Asking “why are they selling?” remains a good question for investors to answer before buying any asset.

Read this week’s earnings and economic calendar here