When Big Spending Becomes a Problem

Oracle Corp (ORCL) reported earnings that should have been good news. Instead, the stock is getting hammered, and it's dragging down the entire AI trade with it. Welcome to the bizarre world where spending money to make money suddenly looks like a red flag.

Here's what happened: Oracle posted earnings of $2.26 per share versus the $1.64 consensus estimate. Sounds great, right? Except that beat came from a one-time gain selling Oracle's stake in Ampere to SoftBank Group Corp (SFTBY). Strip that out, and the picture looks different.

Revenue came in at $16.06 billion versus expectations of $16.19 billion, up 14% year-over-year. The backlog grew to $523 billion, which is genuinely impressive. But none of that matters because analysts are fixated on two numbers that have everyone spooked.

The Numbers That Broke the Market

Oracle spent $12 billion on capex last quarter. Analysts expected $8.4 billion. That's a pretty significant miss if you're looking at it from a "we thought you'd spend less" perspective. Then the company raised its full-year capex guidance from $35 billion to $50 billion. That's a $15 billion increase.

Oracle's management tried to get ahead of the concerns on the conference call. They emphasized that Oracle will maintain its credit rating, that analyst estimates suggesting the company will need to borrow $100 billion are too high, and that there's a tight process in place to control spending. The tone was confident and positive.

It didn't work. The Nasdaq futures dropped in early trading as a direct result of those two capex figures. When a company increases its spending projection by $15 billion mid-year, it raises questions about whether everyone in the AI infrastructure race is building too much, too fast.

The Bull Case and the Bear Case

Here's the thing about Oracle's aggressive spending: if it pays off, the stock could go much higher. The opportunity in AI infrastructure is real, and Oracle is borrowing aggressively to capture it. In a very long-term scenario where Oracle executes successfully, the stock could reach $516 to $533.

But there's a darker scenario. If rumors start circulating about a potential credit downgrade, the stock could fall to $150. As of premarket trading, Oracle was at $193.68, so there's meaningful downside risk if the credit story deteriorates.

The momentum crowd stepped in to buy the dip aggressively in early trading. That's what the momentum crowd does—every tiny dip must be bought. The VUD indicator (which measures net supply and demand in real-time) showed net demand for stocks in early trading, with green representing demand and orange representing supply.

The Ripple Effect Across AI Stocks

Oracle's capex bombshell has rekindled fears of AI overbuilding, and it's hitting a bunch of related stocks. NVIDIA Corp (NVDA), Advanced Micro Devices Inc (AMD), CoreWeave Inc (CRWV), Nebius Group NV (NBIS), and IREN Ltd (IREN) are all feeling the pressure.

The question investors are asking: if Oracle is spending this much and still missing revenue estimates, what does that say about return on investment for all this AI infrastructure spending? And if Oracle stumbles, what happens to the GPU makers and cloud infrastructure providers counting on this spending wave?

How Oracle's stock moves from here will have a major impact on the entire stock market. That's not hyperbole—the AI infrastructure theme has been one of the key drivers of market performance, and Oracle just threw a wrench into the narrative.

All Eyes on Broadcom

Broadcom reports earnings after the market close today, and this report has taken on outsized importance. Broadcom has the potential to either extinguish or exacerbate AI overbuilding fears. The company is Alphabet Inc (GOOGL)'s partner for TPUs (tensor processing units), so its commentary on AI demand will carry significant weight.

If Broadcom reports strong results and confident guidance, it could calm the nerves Oracle just rattled. If Broadcom shows any signs of caution or spending fatigue among its hyperscale customers, watch out below.

Other Market Movers

Initial jobless claims came in at 236,000 versus 191,000 prior, showing an increase. However, continuing claims fell to 1.838 million from 1.937 million previously. Taken together, this data doesn't present any immediate concerns.

In an interesting development, Walt Disney Co (DIS) is making a $1 billion investment in ChatGPT maker OpenAI and will become a major customer. This isn't great news for Google. Disney is also sending a cease and desist letter to Google, adding another layer of tension between the two companies.

Here's a sobering note for Fed watchers: the 10-year Treasury had a yield of 3.65% on September 18, 2024, just before the Fed started this easing cycle with a 50 basis point cut. After 175 basis points of cuts by the Fed, the 10-year Treasury yield is now at 4.14%. As the Fed has lowered short-term interest rates, long-term interest rates have risen. That's the Achilles' heel of this easing cycle.

Money Flows in the Magnificent Seven

Most portfolios are heavily concentrated in the Mag 7 stocks at this point, so daily money flows in these names matter. In early trading, money flows were positive only in Apple Inc (AAPL).

Money flows were negative in Amazon.com Inc (AMZN), Alphabet Inc Class C (GOOG), Meta Platforms Inc (META), Microsoft Corp (MSFT), NVIDIA Corp (NVDA), and Tesla Inc (TSLA).

Money flows were mixed in SPDR S&P 500 ETF Trust (SPY) and Invesco QQQ Trust Series 1 (QQQ).

Commodities Corner

The momentum crowd is extremely aggressively buying silver, which hit a new all-time high. For those tracking precious metals exposure, the most popular ETF for gold is SPDR Gold Trust (GLD), and for silver it's iShares Silver Trust (SLV).

Bitcoin (BTC) is range bound, showing no clear directional movement.

For those interested in oil exposure, the most popular ETF is United States Oil ETF (USO).

Portfolio Strategy Considerations

In this environment, consider continuing to hold good, very long-term existing positions. Based on individual risk preference, consider a protection band consisting of cash or Treasury bills or short-term tactical trades, as well as short to medium-term hedges.

Protection bands can be determined by adding cash to hedges. The high band of protection is appropriate for those who are older or conservative. The low band is appropriate for those who are younger or aggressive. If you don't hedge, total cash levels should be more than stated but significantly less than cash plus hedges combined.

A protection band of 0% would be very bullish and indicate full investment with zero cash. A protection band of 100% would be very bearish and indicate a need for aggressive protection with cash and hedges or aggressive short selling.

It's worth remembering that you can't take advantage of new upcoming opportunities if you're not holding enough cash. When adjusting hedge levels, consider adjusting partial stop quantities for stock positions (excluding ETFs), using wider stops on remaining quantities and allowing more room for high beta stocks—the ones that move more than the market.

Bond Allocation in a Rising Rate Environment

For those following a traditional 60/40 portfolio (60% stocks, 40% bonds), probability-based risk reward adjusted for inflation doesn't favor long duration strategic bond allocation right now.

If you want to stick to the traditional 60% allocation to stocks and 40% to bonds, consider focusing on only high-quality bonds with five years duration or less. Those willing to bring more sophistication to their investing may consider using bond ETFs as tactical positions rather than strategic positions at this time.

The market is at an inflection point. Oracle's spending has raised uncomfortable questions about AI infrastructure returns. Broadcom's report tonight will either validate the spending thesis or confirm the fears. Either way, it's going to be an interesting evening for anyone with exposure to the AI trade.