Intro
Technical strategist David Nicoski walks George Noble through a deck of charts covering gold, energy, copper, natural gas, health care, the Mag 7 and the data-center build-out names, and explains why he is short CoreWeave. Noble adds his own fundamental read alongside it and presses him on whether health care beats the semis over the next year or two.
Guest: David Nicoski, technical strategist
Host: George Noble
Published: 28 August 2026 on The Noble Update Podcast
Show notes | 31 min
Key Takeaways
The money that left tech did not leave the market
"the divergence between biotech and the SOX and techs since mid-June is 50%" — Nicoski
The semiconductor index itself is down 30%
Great companies at the wrong price is the whole opportunity
Dick's at a P/E of 24 three weeks ago against Abercrombie & Fitch at eight
Gold cannot be in a bubble that almost nobody owns
"From the 1981 high, you're down 90%, nine zero." — Nicoski, on gold against the S&P 500
Health care beats the semis over the next year or two
"I think that healthcare is going to outperform the SOX, yes." — Nicoski
Energy is still beating technology measured from the COVID lows
The 10-year yield can clear 4.8 and then run to six
Short CoreWeave, with the credit default swaps at 764
Investment grade sits below 100; the debt yields north of 12%
The data-center build-out names are breaking down
Cummins through its 200-day; Sterling down 50% with the market still at new highs
There is nothing to do in the Mag 7 in either direction
Meta is down 40% from its peak against the S&P 500
Natural gas is basing around 250, and the data centers need it
The cost inflation AI created is now hitting the companies building AI
Own the green squares on the heat map and avoid the red, per Noble
The Money Leaking Out of Tech
Nicoski's opening claim is that tech may have made a top, and that the more useful half of the story is where the money went: "a lot of that money is leaked out into other sectors"
Health care is where it landed. "biotechs have been absolutely on fire" — Nicoski
The gap he puts at the center of the episode: "the divergence between biotech and the SOX and techs since mid-June is 50%"
Noble's reaction: "50, 50, 50. Five, zero. Jesus."
"That's with SOX melting down and XBI going up." — Nicoski
The semiconductor index is the other side of it — Nicoski points at "looking at the SOX index, which is down 30%"
He says he had been calling the shift through June: "every single sector is inflecting bullishly against the SOX", which told him where the money was going
His framing for the whole episode is positional, not directional: "it's more important to be bullish in the right areas"
Great Companies at the Wrong Price
Nicoski, who says up front that he is a technician rather than a valuation analyst, ran through the spread inside consumer names. Dick's had a P/E of 24 three weeks ago; Abercrombie & Fitch was at eight.
"And one was priced for perfection, and the other wasn't."
"Walmart back in February hit its high, trading at a PE of 45." Costco did the same thing.
"Target's up 100% in the last eight months versus Walmart." Target was on a P/E of eight when that started.
The conclusion he draws from the pair: "the market is misallocating capital to get into the best names. And the best names could be great companies at the wrong price."
He says he can find consumer discretionary names with ten years of history trading on P/Es of eight, and he keeps the hedge on it — "Now, I can't tell you they're going to go up, but there's certainly mispricing"
The example he keeps returning to is a restaurant chain: "Take Cheesecake Factory that was trading at a PE of 8. It's up 100% in the last four months."
A Market of Stocks, Not a Market Number
Noble's fundamental gloss on the same point: "the stock is not the company, the company is not the stock. They're two different things."
He adds that what an investor holds is closer to a rental than an ownership claim — "I hate to say we're renting pieces of paper" — even though it is a claim on the company's assets
"It's a market of stocks. It's the golden age of stock picking." — Noble, who says both of them believe it
Noble's account of the regime change: until the fall of last year you bought the Mag 7 and ignored everything else, and since then it has been exactly the opposite. He jokes that in the old regime the question was "who needs to pay Nicoski's research fees?", and that now there is a real chance to add value.
Nicoski says the index call is optional. "we could be sitting here going sideways for six months. Where's the alpha?"
"you got to weed through the marsh" and find the names floating to the top
Two lines on why he ignores the index number entirely: "If I told you the S&P is going to 8,000, it doesn't tell you where to invest." and "Just like when the GDP number comes out, I can't invest in a GDP."
The trade he says was sitting there for anyone watching the sector spreads: "You could have done a pair trade and said money's leaking out of SOX, it's going into healthcare or financials."
Noble's own three-part frame for what a strategist owes a listener: "There's what's descriptive. It tells you what has gone on in the market. There's what's predictive, what you think will go on." The third is prescriptive — what people are actually supposed to do with their money, which he says is what the two of them are trying to answer.
The Indices: Sitting on Support, and a Pair of Twos
Nicoski marks up the broad indices first. The S&P is sitting on a support level with a divergence in RSI.
The level he names for anyone trading futures: "7600 level is an important level, at least from an index if you're trading the futures"
On the Nasdaq, using the QQQ, its performance against the broad market has been falling since early summer — "you really peaked out in that June, third week of June, roughly"
He will not turn that into a bear case: "you can say the index is short-term bearish, but you're also above the 200-day, which is also going to provide ample support"
Noble's summary of the setup: "So basically, it's a pair of twos." Nothing to say either way.
The 10-Year at 4.8, and a Dollar Breakout That Failed
On the 10-year Treasury yield: "This still has the potential to move above the 4.8 level. And if you move up there, you're going to six" — Nicoski, who says the target comes from how deep the base underneath it is
The dollar index went the other way from the consensus. "And we broke out and failed. And now you broke the uptrend."
It is oversold enough short-term to bounce, but "There's so much overhead supply on that when you look back at a two-year chart."
Noble on the chorus that got the dollar wrong: the amateurs at home were all shouting breakout, and "they've got strangely quiet the last few weeks"
Gold Cannot Be in a Bubble Nobody Owns
"I keep hearing how gold is in a bubble" — Nicoski, who answers it with the NYSE ARCA Gold Bugs Index over the long run
The pattern he points to is a break out of a base and what he calls a Gann fan — breaking the third downtrend, then a large move. He says the index is now above a major base.
His test for a bubble is capital absorbed, not price: a bubble "also had to soak up a massive amount of capital", and against the S&P 500 the gold miners are nowhere near a multi-decade high the way technology is
Back in 2011 the ratio was around 180; it has just bounced off the 25 level
On gold the commodity, measured against the S&P 500: "From the 1981 high, you're down 90%, nine zero."
"That is not a bubble. That's not how bubbles get made."
He says the position is still held in disbelief by almost everyone — "They don't know what a shiny rock does." — and that the same disbelief ran through Bitcoin
The portfolio point behind the ratio: "when you're 10% of the value that you were in 1981, that means that that holding is a lot less in everyone's portfolio as well"
Gold's Setup: Oversold, Then a Bull Flag
He puts up the unchanged slide from his previous call, where he had gold very oversold and marked "Current support resistance at the 4,100 level", with the metal down to 10% above its 200-day moving average
The call he made on that slide: "if the secular bull is not over, this is a good period to take positions in a very unloved asset"
Where he thinks it goes next: "I think it's going to back and fill and create what's called a bull flag." He is a buyer on pullbacks rather than here.
He declines to argue the macro case behind it — "I don't know how we fix the debt crisis. I don't want to go through that."
Noble's rule for exactly this moment: "trade what you see, not what you think"
Energy: The Chart He Would Buy Out of a Coma
"I can't tell you where crude oil's going, but I can tell you where this chart's going" — Nicoski, on the XLE
He says he was super early getting bullish on energy stocks, and that his own product covers far more names than the XLE, which only holds the S&P 500 constituents
The comparison almost nobody accepts, in his account: "from the COVID lows, energy is still outperforming the S&P by a dramatic margin. In fact, it's outperforming technology."
"but no one wants to believe it. They just wanna continue to expand their weighting in tech"
"If I woke up from a coma that I had been in 20 years, I'd wake up and go, hey, this is a bull market." — without knowing there was a war or a supply problem
The base is the argument: the sector got washed out at the COVID lows and has been making higher lows over a long period since, and the long stretch of losing ground to the market is breaking.
Semis: Bearish Near-Term, With 1999 as the Template
"I'm bearish near-term, I'm not bearish long-term until you break the longer-term uptrend" — Nicoski on the SOX
He will not try to call the low. "RSI is extremely oversold", and he would rather come back to it at the 200-day.
On the Nvidia print: "I don't know what Nvidia is going to report. I think you can take it as a trade here." The index still has to rally a long way to clear its downtrends.
"So I'm just going to absolve myself and look for other things that are more appetizing at this point in time."
The long-term chart is where his real view sits. He points back to the 1999 and 2000 blow-off top, the rally that followed into July, and its failure to make a new high.
"As of this point in time, I do not think the SOX will go take out its high."
Copper, a Weak Dollar, and Kicking the Can
Copper is above a big base and starting to gain ground on the broad market, per Nicoski
"If we get weakness in the dollar, I think you're gonna see commodities take a move higher."
"I don't see any reason for the dollar to stay as strong as most people think that it should be"
His reason is policy, not positioning: "We're trying to manipulate bonds here now, right? To get out of the problem, we're kicking the can down the road. We're not fixing anything."
The Mag 7, One Chart at a Time
The Mag 7 index is still losing ground to the market. Nicoski allows that the pattern can be read as a symmetrical triangle, but says the performance is not there, and that the trade was the downtrend break in April of last year.
Noble is blunter: "It just doesn't do anything. It's like there's nothing there. It's like, don't waste your time with this."
Microsoft draws the driest line of the episode: "It's quite normal for big market cap companies to jump 40%, isn't it?"
"if you got closer to 350, it's a buy", and he says Meta is showing the same pattern
"But I'm a buyer on callbacks. I'm not buying up here." Its leadership peaked in August of last year.
On Apple, which he does like: he had warned in his previous deck that it was stretched near the highs, and posted a chart on X saying "every time they have a parabolic move, the preceding two weeks are not good for the stock". It then took a tumble.
He rates it fine at this level, but "I'm finding 10 times better things to do outside of the Mag 7"
Meta is a trade rather than a holding in his read: "It's down 40% from the peak versus the S&P 500."
The rest of the roster gets a sentence each — Google looks decent at the 200-day unless it starts beating the market again; Tesla he does not like, with major resistance overhead; and Amazon is "it's me trying to read, hieroglyphics. I didn't go to school for that."
Nvidia is the flattest of the lot: "It hasn't outperformed the S&P, and it hasn't underperformed. It's performing in line."
The Heat Map: Own the Green Neighborhoods
Noble stops to explain Nicoski's group heat map to the audience: "I've been reading this stuff for 30 years. You want to be in the green."
Correctional facilities sit at the top, ranked a one, having been a five or a six a few weeks ago
Noble's version of the rule: "You want to own houses in the good neighborhood, hopefully the best houses in the best neighborhood." If a portfolio is full of names in the red, he says, take a sharper look at it.
Employment services is the call Nicoski says they nailed: "Human Resources, we caught all of that move." Robert Half and Manpower are among the names he lists.
The rest of the green side: tax preparation services, casual dining restaurants, refining, men's apparel retail, diagnostic and test services, housewares and specialties, hotels and resorts, and small caps
The red side is where the AI build-out lives — engineered buildings, nuclear power generation, batteries, energy storage and robotics. "Well, what the hell happened there?"
The reversal he flags on himself: "The new one to the group that I was really early getting bullish on in June, July of 22, was the E&C companies. And I'm now bearish them, by the way."
Health Care Beats the Semis for the Next Year or Two
Nicoski says the health care call started with a reversal in early summer — "we caught that reversal back in June, breaking the shorter term downtrend" — and that the long-term downtrend is now going too, which pushes him to overweight the sector harder
The biotech-against-semis line is the number he keeps coming back to: "That's up 50% from its low.", measured since June
Noble asks him to commit past the next month. Over the next year or two, is long biotech against the semis a winner?
"I think that healthcare is going to outperform the SOX, yes."
"I reserve the right to change my mind when the trend changes. I'm a servant to the trend."
On medical devices, he takes IHI against the Philadelphia semiconductor index and says "that's up about 33% since the end of June", with a good pattern, now flirting with its 200-day
Natural Gas Basing Around 250
The area he calls interesting and compelling is natural gas, where both the price downtrend and the underperformance are breaking. He says the soft grains are showing the same thing.
"you can see that there's a bottom right there around the 250 level"
He runs several producers behind it: CRK is starting to reflect the bottom, and RRC is reversing
The demand argument is the data centers: "I'm seeing a lot of names in that natural gas area that the data centers really need natural gas"
The hedge stays on: "I can't tell you how long it's gonna go on"
The Data-Center Build-Out Names Are Breaking
A name he had highlighted in his last packet reported, "They reported, gapped up, went above 900." — and went straight back to where the uptrend had broken. It now trades lower than it did before the report.
Cummins is breaking its uptrend and going through the 200-day
"this has me concerned because a lot of these names go into the data center build out" — Nicoski, who says the breakdowns are showing up across the group
"You'd almost think this Dycom chart is the Dick's chart, right?" He calls it a massive break.
Noble's read is that the fundamentals have not moved and the stocks have. He points at Sterling: "The thing bottomed at 100, went up to 1,000."
Nicoski: "This thing's already down 50%. The market's still at new highs. Like where did that market cap go?"
He notes he turned bullish on the entire group in July of 22 and has the research to prove when he made the call
Walmart, Costco, and the AI Play Joke
Nicoski says he called Walmart in February, saying then that "this is ridiculously priced", and it double topped
"Now you went from all-time relative strength highs versus the S&P in April to two-year relative strength lows that fast"
Costco did the same thing off its own new high, "and they still have PEs of 40, 40 plus."
Noble's deadpan: "Dave, you don't get it. They're AI plays."
The other side of the trade was the name everyone hated: "This was Target trading at a PE of eight. Walmart's at 45, which one do you wanna own?" Target is up 100% against Walmart since the December low.
While money crowded into Walmart and Costco for safety on fears the economy might fall apart, "Cheesecake was valued at eight times earnings. Now it's at 23."
Short CoreWeave at 764 on the Credit Default Swaps
"I'm short CoreWeave. I bought the triple short CoreWeave at $2 when they beat and gapped up." — Nicoski, who says he took the trade that day
His evidence is the credit market rather than the chart. The credit default swaps, he says, "They're trading at 764.05."
For scale: "anything that's investment grade is below 100"
"This is at 764, which tells you that the expected potential for default in five years is 50%. Their yield on their debt is north of 12%."
On the engineering and construction names, he says he asked AI what the companies themselves were saying, and got "cost controls, inflation in the area of hiring, electricians, HVACs and plumbers was greater than they expected" — costs they could not drop to the bottom line
The loop he draws out of it: "you're seeing the inflation that AI is causing is now affecting the companies that are actually creating the inflation"
Software, the Products, and What Charts Are For
Noble sums the episode as a stock picker's market with precious metals, energy and biotech working and tech uninteresting. Nicoski qualifies the last part.
"There are some software names that look good. And I think maybe AI is overblown for software at this point in time." He says the group was beaten too hard and nobody gets to see the five-year trajectory of what AI does to it.
Asked about his retail products, Nicoski says he has so many that he needs to make a training video explaining what each one is for, and that users need to know what is a long-term signal and what is a trade
When he marks up 40 charts, the useful information is how many names cluster in one area — restaurants, consumer, industrial or financial
On the employment services names: "It didn't matter which one you bought, they were all up 40 to 50%."
Noble puts his own fundamental neck out where Nicoski will not: "I expect your gold and silver energy calls are going to be up and to the right. I think the tech is going to be down big, down bigly."
They do the show roughly monthly and expect to be back in September
Noble on why he reads charts at all: "The thing I love about charts is they take you out of your personal thinking. This is what's happening. It's not what you want to have happen. It's what's happening."
Nicoski's closing argument is about timing, not valuation: "behavior is a big part of this market. And if you can time behavior, you're going to do a lot better than just looking only at fundamentals because timing is everything."
"You can short a stock and it can go up 20% against you." Find the reversal and the negative inflections on the chart instead.
Nicoski's bottom line is that the index number is beside the point: the money that left the semiconductors is already sitting in gold, energy, natural gas and health care, and he would rather own those and short the AI build-out's weakest credits than try to pick a bottom in the SOX.
Products, Companies & Tools Mentioned
XLE and XLK (The energy and technology sector funds; Nicoski uses the XLE as the energy proxy while noting his own work is far broader than the S&P 500 names it holds)
XBI and the SOX (The biotech-against-semis spread that anchors the episode — "the divergence between biotech and the SOX and techs since mid-June is 50%")
NYSE ARCA Gold Bugs Index (The long-run gold miner chart he uses to argue gold is not in a bubble — around 180 against the S&P 500 in 2011, just bounced off 25)
Walmart and Costco (Both hit highs on P/Es in the forties and went from record highs against the S&P 500 in April to two-year lows against it)
Target and Cheesecake Factory (The hated side of the same trade — Target on a P/E of eight and up 100% against Walmart since December, Cheesecake rerated from eight times earnings to 23)
Dick's Sporting Goods and Abercrombie & Fitch (P/Es of 24 and eight three weeks ago — "one was priced for perfection, and the other wasn't")
CoreWeave (His short, taken through a triple-short product on the post-earnings gap up; the credit default swaps are "trading at 764.05" against a sub-100 investment grade threshold)
Cummins, Dycom and Sterling (Data-center build-out names breaking down; Sterling "bottomed at 100, went up to 1,000" per Noble and is now down 50% with the market at new highs)
Microsoft, Apple, Meta, Google, Nvidia, Tesla and Amazon (Chart by chart, all rated somewhere between neutral and untouchable; Microsoft a buy nearer 350, Meta down 40% from its peak against the S&P 500, Nvidia "performing in line")
CRK and RRC (Natural gas producers reflecting the bottom he sees around the 250 level)
IHI (Medical devices, up about 33% against the Philadelphia semiconductor index since the end of June and flirting with its 200-day)
Robert Half and Manpower (Employment services names from the group call he says they nailed, where "It didn't matter which one you bought, they were all up 40 to 50%.")
QQQ (The Nasdaq proxy, short-term bearish but holding above its 200-day)
Copper (Above a big base and starting to gain on the market, with a weaker dollar as the catalyst he wants)
Bitcoin (Cited only as a precedent for the disbelief that surrounds gold now)
Books & Resources Mentioned
Vital Signs and Booster Shots (Nicoski's own research publications, where he says the group calls including employment services were flagged to clients)
His firm's Substack (The other place he says those calls were published)
A chart posted on X (His warning that Apple's parabolic moves are followed by two poor weeks)
Get the latest market chatter and takes as they happen:
X | Threads | Instagram | YouTube | TikTok | Facebook