Stocks cratered this week in one of the weirdest market moves in recent history. But the harder investors look, the more confident they should be about buying the dip -- as long as they pick the right ways to do it.

The midterm elections were supposed to be the "clearing event" the market needed to resolve its jitters and head higher. While the initial reaction to the Democratic sweep was positive, the good feelings soon faded, with the Dow Jones Industrial Average on track to end the week down 0.6%, the S&P 500 index falling 2.3%, and the Nasdaq Composite slumping 3%. The action presents itself as a question for investors: Did the election mark the end of the AI trade and the bull market?

We don't think so. Sure, it's a mess, and we've entered a more volatile environment that will probably carry us through Christmas. But demand for artificial intelligence should continue, even if the buildout of data centers slows, while a strong economy should continue giving the non-AI stocks tied to that strength a boost. Rather than fearing a bear market, investors should try to untangle the web of political aspirations, economic incentives, and technological necessities to find the dips worth buying in a very strange selloff.

Here's what we know. Democrats won massively in Tuesday's elections, taking control of both chambers of Congress, as well as several important governorships; in addition, voters overwhelmingly voted to restrict data centers in ballot measures, which shows just how unpopular the AI buildout has become. The results led to initial speculation that tariffs would be challenged, the Iran war would soon come to an end, and data-center construction would slow -- a cocktail that explains why oil plunged on Wednesday, AI buildout stocks like Nvidia and CoreWeave slumped, and the consumer discretionary sector surged 4.3%, driving an overall S&P 500 gain of 1.6%.

After Thursday's open, things got really hairy. While AI names led the losses, the entire market was in free-fall from the opening bell to 10:52 a.m. ET, and after a brief respite, stocks closed at the day's lows. The facts hadn't changed since Wednesday, but the vibe certainly had. With little news to explain the decline, the stock market's tumble was likely driven not by fundamentals, but by forced selling. Indeed, news soon emerged that several hedge funds were liquidating positions, meaning that they were shutting down their books (or portion of their businesses) and getting out at any price. And what were they getting out of? The small set of stocks that had surged all year, which essentially means AI buildout plays such as Caterpillar, Micron Technology, and the like.

On Friday, the selling slowed -- the Nasdaq rose 1.3%, the S&P 500 was up 0.6%, and the Dow slipped slightly -- as investors surveyed the wreckage. And a bounce could be in the offing.

"If there's no financial motivation outside of getting the open positions to zero, then you must take the other side," says Dennis Davitt of Millbank Dartmoor Portsmouth.

And let's not forget that valuations, once a concern for this market, have fallen so much that they look downright attractive. The S&P 500's forward price/earnings ratio is 17.9, the lowest it's been since 2023, per FactSet data. Nvidia's forward P/E is 13.2, while Micron's is 3.8. You could say that earnings estimates, and actual earnings, will fall from here, and fair enough. But it takes some serious elbow grease to argue that investors are living in a bubblicious la-la land.

The Nasdaq, meanwhile, is now down more than 10% from its Oct. 6 peak, putting us in correction territory. Since 1971, there have been 53 corrections -- not including the current one -- and 20 of these progressed into bear markets, meaning that the Nasdaq fell 20% or more. Since the median progression from correction to bear took just 28 days, we ought to know pretty soon whether we're entering one.

It is hard to make that case. Not only has any residual stock market euphoria been terminated with extreme prejudice, but the economy remains almost shockingly strong. U.S. gross domestic product grew at an annual rate of 3.4% during the third quarter -- a far cry from the economic shrinkage that marks a recession. On Friday, we learned that the U.S. added 165,000 jobs in October, and that the unemployment rate fell to 4.1%. Carol Schleif, chief market strategist at BMO Wealth Management, calls the jobs report "another indication that there's a lot of really good, solid momentum in the U.S. economy."

In particular, she points to the dynamic of "AI layoff boomerangs," in which companies that have cut employees with the aim of replacing them with AI tools have found the need to hire again. Some might see this as a kind of admission of failure, if it indicates that AI isn't doing an adequate job of replacing these employees' tasks. But Schleif takes the more optimistic view that companies have actually proven successful at integrating AI, and are now growing quickly enough that they can boost hiring again, often in new roles that involve collaborating with AI agents. "These labor market numbers are indicative of an economy transitioning into the next iteration of the information age, and it's really healthy," she says.

Whatever you think is driving labor market growth, the fact remains that when the unemployment rate is at 4.1%, even Nostradamus would find it hard to argue that a recession is nigh. And while a serious slowdown in AI-related capex wouldn't help the economy, Dennis DeBusschere at 22V Research points out that since consumer expenditure, not AI, is the main driver of economic growth right now, "AI capex plateauing would more likely result in a benign slowdown than a recession."

This insight has a serious impact on allocation decisions. When recession risk is rising, you want to buy defensive stocks such as consumer staples and utilities -- or, if you're confident the economy is rolling over, simply flee for the safety of bonds. But that's not the environment we're in, which is why DeBusschere favors non-AI cyclicals over defensive sectors.

We don't want to minimize the situation. The biggest mistake investors tend to make after a market drop -- well, besides just selling everything -- is transposing yesterday's environment onto today's prices and concluding that everything is a bargain. Some things really have changed. The Democrats' incredible electoral gains, particularly in Texas, could slow down the buildout of data centers. In addition, the AI buildout remains dependent on investors pouring in more capital to fund the construction of new data centers, the buying of chips, the training of models, and -- here's the big one -- the pursuit of profitable applications.

Drawing this much cash was already a tough prospect. "Recent significant capital raises at other notable companies, and others soon anticipated to take place, to fund the same AI growth -- for example, Alphabet, SpaceX, Anthropic, and OpenAI -- have resulted in an increasingly challenging environment for raising capital," Meta Platforms wrote in an October filing. Anthropic told investors it is looking forward to "an age of miracles" that will see AI cure cancer and help businesses devise "new products and new ways of working." It still has a long way to go, which is why Anthropic is looking to raise some $100 billion in an IPO valuing the company at $1.5 trillion. But on Thursday, as markets tanked, the chance of Anthropic going public this year fell from 90% to 64%, per bettors on Kalshi.

There's no way investors are as enthusiastic today about AI as they were a week ago, much less a month ago. Since AI expansion is still being driven by capital expenditures and not profits, this harsher environment could significantly slow down the development of the technology itself.

That has a wide range of implications for the different players. Companies like Amazon.com and Alphabet that have solid core businesses and easy access to capital markets could be relatively insulated -- they'll just slow down their spending on AI capex. Companies like Oracle and CoreWeave, which are more highly leveraged and more dependent on continued growth in AI spending, look more tenuous.

Apple, meanwhile, seems to be playing from a position of tremendous strength. The big knock on the company was that it was way behind in AI, was selling expensive products to a weak consumer, and was paying through the nose for memory chips, since memory prices have skyrocketed on the back of incredible AI-driven demand. If AI growth slows while the consumer stays strong, Apple is the place to be, especially if it can get its margins expanding again. And traditional consumer-led cyclical stocks, like Ford Motor, Starbucks, and Walt Disney, would look even better.

But don't count memory out. The Roundhill Memory exchange-traded fund got absolutely creamed on Thursday, and even after a bounce on Friday, is still down 28% since Oct. 28. Micron was sitting on a year-to-date gain of 325% in June; after falling more than 40% since then, it's now up 140% in 2026. Its chart really only makes sense if you consider investor dynamics. Micron was part of a set of hot stocks widely held by hedge funds and retail investors alike -- and both used leverage to juice their risk. On Thursday, both pros and mom-and-pops sold, by force or by choice, in a Caligulan orgy of selling. Leveraged ETFs, which need to reduce their positions when prices fall, sold more than $60 billion worth of equity exposure into Thursday's close, with 90% of that tied to tech, according to J.P. Morgan estimates.

The swift selloff had some investors scratching their heads. The data-center buildout might indeed slow as a result of the election, but this is "a slight adjustment, rather than a sea change," says Divya Mathur, portfolio manager at ClearBridge Investments, adding that memory stocks "look extraordinarily cheap, so long as nothing crazy happens on AI infrastructure."

The market for memory chips remains vastly undersupplied, with Micron, Samsung Electronics, and SK Hynix selling out capacity months or even years ahead at previously unheard-of prices. "There's an enormous disconnect between what customers are prepared to spend and what the equity market seems willing to believe," says David Fetherstonhaugh of VistaShares. And while markets sometimes move before the fundamentals can catch up, it's a different story when hedge funds are selling off their positions at fire-sale prices.

Indeed, we think a barbell strategy of holding cyclical stocks plus memory companies has merit. You benefit from continued strength in the U.S. economy, and while you're not exposed to the most speculative aspects of the AI buildout, holding memory stocks does give you schmuck insurance -- at a very reasonable price.

At the end of the week, the big questions about AI remain unanswered -- and will likely stay that way for a while. Over the next two years, politicians of both parties will likely race to oppose data centers, and perhaps artificial intelligence in general, which certainly won't help matters. But while AI has been driving the stock market, it hasn't really been driving the economy. The market is now shifting, violently, to a new leadership regime. Investors who haven't reassessed their exposure to the AI buildout need to think deeply about how the companies in their portfolio will be impacted.

Just don't confuse a shakeout with a bear market.