MacroVoices #540 Adam Parker: Beyond the AI Bubble: Diversifying Portfolios in an Earnings-Driven Market

Adam Parker expects US equities to grind higher over the next six to twelve months, driven by strong corporate earnings even as price-earnings multiples contract modestly. He does not believe the US is in a stagflationary environment and would not pre-position an equity book for a growth scare. He argues the market is pricing a distribution of 2031 revenue outcomes, not ignoring fundamentals, and supports this by noting that companies where out-year revenue revisions are rising have outperformed those where they are not.

Parker argues the S&P 500 currently functions like an AI semiconductor ETF, and that genuine diversification requires measuring AI exposure across the full revenue chain. He estimates approximately 265 of the top 3000 US equities by market cap have meaningful AI revenue, defined as at least 5 percent of revenue or an imminent new product announcement, with every sector except consumer staples represented. Stocks such as Eaton, Caterpillar, and GE Vernova appear to be in industrial and utility sectors but carry approximately 0.9 correlation to AI semis, meaning they do not provide the diversification investors might assume. Memory and semi-cap equipment stocks are the most volatile cluster within the chain, and Trivariate's 10,000 simulations on Micron found it trades at roughly four to five times peak earnings and ten times normalized earnings, which Parker views as probably too cheap.

Parker recommends tech, healthcare, and energy as the offensive core of a US equity portfolio. He estimates a 30 to 40 percent probability that healthcare is the best performing sector over the next five years, versus near zero implied by current market pricing, noting healthcare revenue per share has grown every year for more than 30 consecutive years and that healthcare stocks are very anti-correlated to AI semiconductors. Energy equities represent approximately 4 percent of the S&P 500 and sit at a 40-year low in their correlation to the tech sector, making them a diversifying feature. Parker recommends owning more than benchmark weight in energy, citing above-average estimate achievability because changes in oil price are highly correlated to changes in net income for energy companies. Power, utilities, and energy stocks are also a meaningful part of the AI revenue chain, and Parker is bullish on that space partly because the power constraint on AI infrastructure prolongs the cycle by preventing it from overheating.

Valuation-based stock picking has not generated returns for approximately 15 years. Price-to-earnings, price-to-forward earnings, and price-to-book have no useful information for stock selection over timeframes under three years. Buying stocks that just became more expensive on price-to-forward earnings has a higher probability of beating estimates in the current regime, and stocks that beat estimates once have a higher conditional probability of beating a second time than the unconditional base rate. High-quality stocks as scored by Trivariate have not beaten junk stocks for six years, and approximately 82 percent of current S&P 500 market cap is top-half quality stocks, meaning a long-only investor could own two-thirds top-half quality names and still be significantly underweight the index. Running 50 or more names has produced better performance for institutional managers over the last few years, as drawdowns from concentrated portfolios have been too extreme even for 75th-percentile stock pickers.

Parker's analysis of April earnings call transcripts found that Hormuz and Iran oil commentary had limited net impact on earnings estimates, with the most negatively affected companies concentrated in lower-quality consumer discretionary names already beaten down. He does not expect a Hormuz re-escalation to cause a meaningful sustained market sell-off. On market structure, S&P and Dow futures were noted to be at the 100th percentile of long positioning on a one-year lookback, and any rebalancing of crowded CTA systematic positions was described as potentially disorderly, with systematic sell tripwires estimated at 7,300 to 7,350 on the S&P 500.

On bonds, large speculators covered 85,000 contracts of 30-year bond shorts in a single week, moving positioning from near the bottom to the middle of the one-year range, while commercials flipped from net long to net short in the same period. This short-covering episode was described as the first real crack in the consensus trade of being short the long end of the curve. Parker argues that incorporating interest rate views into equity positioning has historically been a loser's game and does not believe there is a specific 10-year yield level that magically caps equity performance, arguing that when growth is strong, multiples expand and earnings grow simultaneously.

Patrick Ceresna proposed expressing the energy sector rotation through the XLE ETF, trading around 55 and a quarter at the time of recording, using an August 21, 2026 collar structure buying the 50 put for 30 cents and selling the 65 call for 20 cents for a net debit of 10 cents, providing approximately 10 dollars of upside potential with hedge protection starting roughly 5 dollars below the current price.