AI Stocks Fell 25%. Your DCA Plan Didn't Blink.

Twenty-five percent. That's how much value the Philadelphia Semiconductor Index has shed since June 22, when AI-chip stocks were still being priced as though the spending could only go up. Five weeks later, that trade is sitting in a textbook bear market — the kind measured in percentages, not vibes — and it dragged the rest of the tech-heavy Nasdaq down with it.

On July 28, the Nasdaq-100 crossed the 10% threshold that officially defines a correction, its fourth straight losing session driven by chipmakers. A day later the Dow fell 1,153 points, its worst single day since April 2025, after the Federal Reserve held its benchmark rate at 3.50%–3.75% in the most hawkish way possible: three officials voted for a hike instead of a hold, and Chair Kevin Warsh told reporters there is "no soft implicit target" on inflation. Middle East tensions flared the same week and pushed oil higher, giving equities one more reason to sell.

None of that is abstract if your 401(k) or brokerage account holds an S&P 500 index fund, a target-date fund, or anything else leaning on the handful of AI-adjacent names that now make up a large share of the index. The question worth answering this week isn't whether the selloff is real — it is — it's what a scheduled monthly contribution should do about it, and whether a drop this size is one you can actually sit through without selling at the bottom.

The Bear Market Hiding Inside the "Safe" Index Fund

The mechanic driving this selloff is straightforward, even if the dollar amounts are not. The largest hyperscalers — Alphabet, Microsoft, Amazon, Meta and their peers — are on pace to spend roughly $785 billion on AI infrastructure in 2026 and close to $1 trillion in 2027. Alphabet alone raised its 2026 capex guidance to $205 billion, more than double the $91 billion it spent in 2025. Investors spent over a year rewarding that kind of spending. Now they're starting to ask what it's actually buying.

Alphabet had its worst trading day of the year the day after that guidance landed, despite beating earnings estimates. Meta's stock fell even after reporting record quarterly profit. That combination — good numbers, bad reaction — is the tell that this isn't a fundamentals problem. It's a positioning unwind. "There's a lot of panic around the AI investment, and the panic appears to be indiscriminate," D.A. Davidson analyst Gil Luria said of the selloff. Indiscriminate selling is exactly what shows up in a plain S&P 500 index fund, since AI-adjacent names are concentrated enough in the index that nobody who owns one is fully insulated, no matter how boring the rest of their portfolio looks.

This is where the Fed decision compounds the problem instead of cushioning it. A rate cut would have handed the market a reason to look past the AI jitters. Instead, three Fed officials pushed for a hike, and Warsh's insistence on hitting 2% inflation "on this committee's watch" told traders that borrowing costs aren't coming down to rescue growth stocks anytime soon. Add Middle East tensions pushing oil higher, and you get a selloff with three separate causes stacked on top of each other — exactly the kind of week that tests whether an investing plan survives contact with real fear.

What a Real Correction Actually Buys You

Here's the part the panic obscures: a fixed monthly contribution doesn't care whether the market is up or down — it only cares about the price per share the day it lands. When prices fall, the exact same $500 buys more of the fund than it did a month ago. That's the entire mechanical case for dollar-cost averaging, and this correction is a live demonstration of it rather than a textbook example.

Run the math on the semiconductor slide specifically. A $500 contribution that bought one "unit" of a chip-heavy fund at the June 22 peak now buys roughly 1.33 units at today's 25%-lower price — the same dollar amount, a third more shares. On the broader Nasdaq-100's milder 10% pullback, that same $500 buys about 1.11 units instead of one. Neither number requires the market to do anything special going forward. Both are locked in the moment the contribution clears, just by continuing to invest on schedule.

Extend it further. Two months of a $500 contribution made during a 10% dip — $1,000 total — buys shares that only need an 11.1% recovery to get back to where they started, because that's the math of clawing back a 10% loss. If the index eventually retraces to its prior high, as it has after every correction of this size in the last decade, those specific shares are worth roughly $1,111 the day it happens — an $111 gain embedded purely by not stopping, on top of whatever the market does afterward. An investor who paused those same two contributions out of fear doesn't just miss that $111. In most cases, they miss the discount and the recovery entirely, because the money that would have gone in during the dip either sits in cash or gets deployed later, after prices are already back up. That's the whole case for staying on schedule: skipping contributions to dodge a correction doesn't eliminate the risk — it just trades a temporary paper loss for a permanent missed gain, because time in the market beats timing it, every cycle.

How Much of This Can You Actually Stomach

The math above assumes you'll actually keep contributing through the drop, and that's the part a spreadsheet can't verify. Whether a 10% or 25% paper loss feels survivable or terrifying depends on where you land on the risk-tolerance spectrum, and most people don't find out until they're already looking at a smaller number than they expected.

If the $12,500 number in that middle row makes your stomach drop, that's useful information, not a reason to panic-sell. It just means your actual risk tolerance is lower than a "buy the dip" headline assumes, and the honest way to find out is to check it against a real number while the fear is still real — not after the market has already recovered and the answer feels obvious in hindsight.

The Difference Between a Correction and a Reason to Change Your Plan

Not every drop deserves the same response, and conflating them is how people end up making bad decisions at the worst possible moment. A garden-variety 10% correction, which is what the Nasdaq-100 is in right now, happens on average about once a year and has historically recovered within a few months. The semiconductor sector's steeper 25% slide is a different animal — sector-specific, driven by a legitimate debate over whether $785 billion in annual AI infrastructure spending will generate a return, not a broad economic downturn.

The distinction matters for what you do next. A garden-variety correction is exactly the environment dollar-cost averaging is built for — you keep buying, the discount is temporary, and the math above holds. A genuine reassessment of AI spending's payoff is a longer, slower story that plays out over quarters, not days, and it's a reason to check how concentrated your specific holdings are in that trade, not a reason to sell everything on a single red week.

Your Next Contribution Matters More Than This Week's Headlines

Nothing about this week changes what you actually control, which is whether the next scheduled contribution still goes in. The chip-sector bear market is real. The Fed just told you rate relief isn't imminent. Middle East tensions could easily add another rough week. None of that changes the mechanics of what a fixed dollar amount buys at a lower price.

The decision that actually matters is smaller and more boring than the headlines: keep the contribution on schedule, or don't. One of those choices locks in a permanent discount. The other one doesn't.