Investors and Wall Street are thinking differently about the market than they have for the last 15 years. And it starts with a question that sounds almost too simple. Can an AI chatbot or AI-coded software replace what this company does?

If the answer is no, if there's no realistic scenario where Claude or GPT or Gemini eliminates the need for what a business produces, you might be holding a HALO stock.

And HALO stocks are absolutely crushing it in 2026 while everything else bleeds.

Josh Brown of Ritholtz Wealth Management coined the term in early February, and honestly, it's one of those frameworks that clicks the moment you hear it. HALO stands for Heavy Assets, Low Obsolescence. These are companies whose products and services exist in the physical world, things AI simply cannot digitize away. Think concrete. Oil refineries. Warehouses full of groceries. Airplane engines. Restaurant kitchens.

No large language model is coming for these businesses. Not this year, not in five years, probably not ever.

And the market is telling you this matters. Loudly.

There is a real divergence happening early in 2026 that should have every investor paying attention. The iShares Software ETF (IGV), the benchmark for enterprise software companies, is down roughly 22% year-to-date as of mid-February 2026. Salesforce has dropped about 26%. ServiceNow is off 28%. Intuit cratered more than 34%. The sell-off has wiped out close to $830 billion in market value from the software sector since late January. This sector has been crushed in every sense of the word.

Meanwhile energy stocks (XLE) are up nearly 23% year-to-date and just hit all-time highs. Consumer staples (XLP), materials (XLB), and industrials (XLI) have all outperformed the broader market in January and February. The S&P MidCap 400 hit new highs. Small-cap stocks with actual earnings (the S&P 600) hit new highs. The spread between equal-weight and cap-weight performance is the widest it's been since 1992.

This isn't a bear market. It's a rotation. Capital is pouring out of companies perceived as AI victims and flooding into companies perceived as AI-immune. That rotation has a name now: HALO.

To understand why HALO matters, you need to understand what's happening to the stocks on the other side of the trade.

The software sell-off has been building for months, but it accelerated violently in late January and early February after Anthropic released new tools through its Claude Cowork platform - with all legal administration features, marketing capabilities, and finance automation - under an open-source license. The market interpreted this as proof that AI companies are moving aggressively into the "application layer," directly competing with established software businesses for their most lucrative revenue streams.

Goldman Sachs strategist Ben Snider drew a chilling comparison, likening the software sector's potential trajectory to newspapers, an industry where disruptive technology caused prolonged stock declines that only stabilized after years of falling earnings. By that point, most of the equity value was already gone.

JPMorgan took a more measured view, calling the sell-off "indiscriminate" and arguing that AI is more likely to complement software workflows than replace them entirely. But it doesn't matter who's right in the long run. What matters right now is that the market is pricing in existential risk for an entire category of stocks. And when the market prices in existential risk, money has to go somewhere.

It's going into HALO.

The framework is elegant because it's so simple. Brown's litmus test boils down to one question: Will chatbots and LLMs lessen or eliminate the need for this company's product or service in the near future?

If the answer is no, you might have a HALO stock. If the answer is yes or maybe, that's not HALO. Doesn't mean it's a bad investment. Just means it's complicated for 2026.

Let me walk through how this works in practice.

Exxon Mobil (XOM) — HALO. The company produces and refines hydrocarbons. It owns massive physical infrastructure—drilling platforms, refineries, pipelines, storage terminals. AI can make Exxon's operations more efficient (better exploration models, optimized logistics), but no language model is going to eliminate the world's need for oil and gas. The stock broke above a consolidation range in late January and has gone straight up, even while oil prices haven't moved much. Investors are paying for the safety of the HALO designation, not the commodity price.

Walmart (WMT) - HALO. Almost everything they sell is physical. Their competitive advantage is the best logistics network on earth—warehouses, stores, parking lots, trucks, shelves stacked with actual products. You still need to eat dinner. You still need to get dressed. AI might help Walmart optimize inventory and reduce labor costs, making the business better, not obsolete. The stock is at all-time highs.

McDonald's (MCD) - HALO. Nobody's replacing a Big Mac with a prompt. The physical restaurant infrastructure, the supply chain, the brand—these are tangible assets that exist entirely outside the digital disruption zone.

Baker Hughes (BKR) - HALO. Oilfield services. Massive physical equipment. Drilling rigs don't run on tokens.

Caterpillar (CAT), Delta Air Lines (DAL), Hershey (HSY), Corning (GLW), Vulcan Materials (VMC), Ventas (VTR) - these are all HALO. They make tractors, fly airplanes, produce chocolate, manufacture fiber optics and glass, mine aggregates, and operate senior living facilities. No large language model is going to obviate the need for any of these things.

The common thread is the physical nature of what these businesses do and the near-impossibility of AI disrupting their core products.

Alright, let's get practical. If you're a self-directed investor trying to figure out how to position for this theme, here are the ways I'd think about it, from simplest to most involved.

If you want broad HALO exposure without picking individual stocks, sector ETFs are your friend. The sectors with the highest concentration of HALO-qualifying companies are:

Energy (XLE) - Expense ratio 0.08%. Dominated by Exxon, Chevron, ConocoPhillips, and the oilfield services companies. Pure HALO. Already up 23% YTD but energy has been in a secular underinvestment cycle for years. The dividend yield around 3% doesn't hurt either.

Industrials (XLI) - Expense ratio 0.08%. Caterpillar, Boeing, Honeywell, Union Pacific. Companies that build real things in the real world. Manufacturing, transportation, defense - all HALO.

Materials (XLB) - Expense ratio 0.09%. Mining, chemicals, construction materials. Vulcan Materials, Martin Marietta, Freeport-McMoRan. You need copper for AI data centers, you need concrete for everything else.

Consumer Staples (XLP) - Expense ratio 0.09%. Walmart, Procter & Gamble, Costco, Coca-Cola. Defensive, boring, and completely AI-proof. People are going to keep buying toilet paper regardless of what Anthropic releases next quarter.

A simple approach here is to allocate 5-10% of your portfolio to each of these sectors, funded by trimming your software and high-multiple tech exposure. You don't have to make dramatic moves. Even a modest tilt toward HALO and away from AI-vulnerable names gives you exposure to the rotation.

If you prefer picking individual names, which, if you're reading this newsletter, you probably do, the HALO filter gives you a systematic way to evaluate every stock in your portfolio and watchlist.

For each position, ask three questions:

Here are some individual names worth considering, organized by sector:

Energy: Exxon Mobil (XOM), Valero (VLO), Baker Hughes (BKR), Phillips 66 (PSX). These companies own irreplaceable physical infrastructure and benefit from the structural underinvestment in energy production over the past decade.

Industrials: Caterpillar (CAT), Applied Materials (AMAT), Delta Air Lines (DAL). Heavy machinery, semiconductor manufacturing equipment (ironically, this powers AI), and airlines all qualify. You can't 3D print a Boeing 737.

Materials: Martin Marietta (MLM), Vulcan Materials (VMC), Corning (GLW). Infrastructure spending requires physical materials. Corning is interesting because its fiber optics and specialty glass are critical for AI data centers—making it both HALO and an AI beneficiary.

Consumer Staples: Walmart (WMT), McDonald's (MCD), Starbucks (SBUX), Hershey (HSY). Boring businesses with physical products that human beings need regardless of technological disruption.

Healthcare/Real Estate: Ventas (VTR) operates senior living facilities. Aging demographics plus physical real estate equals HALO.

I wouldn't go all-in on HALO any more than I'd go all-in on anything else.

If you currently have 40-50% of your portfolio in tech and growth (which many self-directed investors do after the 2023-2025 run), consider trimming that to 30-35% and redeploying the difference into HALO sectors. That gives you meaningful exposure to the rotation without abandoning your growth positions entirely.

For individual HALO stocks, my standard position sizing rules apply: 3-5% for moderate conviction, 5-7% for high conviction, never more than 10% in any single name regardless of how HALO it is.

I'd be doing you a disservice if I didn't talk about what could go wrong. A few things:

The rotation could reverse. If AI disruption fears fade, maybe because software companies report strong earnings that prove the narrative wrong, money could flood back into tech and out of HALO sectors. Brown himself acknowledged that some software companies are getting "absurdly cheap," and investors with longer time horizons might find better opportunities in the wreckage. There are great names in the wreckage, but you have to be discerning and careful.

Energy is cyclical. HALO doesn't mean risk-free. Oil prices could collapse. A recession would hurt industrials and materials. Consumer staples tend to underperform in strong bull markets. You're trading one type of risk (disruption risk) for another (cyclical and commodity risk).

The HALO premium could get expensive. As more investors crowd into "AI-safe" stocks, valuations in these sectors could stretch. Consumer staples are already trading at the 97th percentile of their 10-year forward P/E distribution, according to Morningstar data. Paying up for safety is fine, but overpaying for safety is a different kind of mistake.

AI disruption might broaden. Bloomberg reported in mid-February that AI-driven sell-offs are spreading beyond software into logistics, real estate, insurance, and wealth management. If AI disruption fears expand to more sectors, the universe of "safe" HALO stocks shrinks.

The HALO framework isn't a permanent investment philosophy. It's a lens for understanding what's working right now in a market that's violently repricing which companies can survive the AI revolution and which can't.

Fifteen years of worshipping "asset-light" business models with high margins and recurring revenue? That entire logic is getting flipped on its head. The very qualities that made SaaS stocks market darlings - the digital delivery, scalable software, low marginal costs - these are exactly the qualities that make them vulnerable to AI competition.

Meanwhile, the boring, heavy, physical businesses that everyone ignored are suddenly the safest place to park capital.

Could I be wrong about this lasting? Of course. But here's what I keep coming back to: the companies making concrete and refining oil and selling hamburgers don't need to prove they can survive AI. They already have. The burden of proof is on the other side.

HALO. Heavy Assets, Low Obsolescence. Simple framework, real money behind it, and the market's telling you it matters.

Keep an eye on this.

The Earnout Investor provides analysis and research but DOES NOT provide individual financial advice. Jamie Dejter may have a position in some of the stocks mentioned. All content is for informational purposes only. The Earnout Investor is not a registered investment, legal, or tax advisor, or a broker/dealer. Trading any asset involves risk and could result in significant capital losses. Please, do your own research before acquiring stocks.