Is Applied Materials a Buy Before Earnings? Not at 43x Into a 7.5% Move

TL;DR

• No, not at $548.15. Applied Materials closed up 4.29% on Wednesday, spending part of the good news before the print. At roughly 43x forward earnings against Nvidia's ~24x, I am not paying up for a cyclical toolmaker the day it reports.
• The quarter itself is pre-agreed. Management guided $8.95bn plus or minus $500m and $3.36 plus or minus $0.20; consensus sits at $3.39 on $9.01bn, a whisker above the midpoint. The FQ4 guide is the only number that moves the stock.
• Street mean target is about $630, roughly 15% above Wednesday's close, and the stock still sits about 26% below its $739.67 June 30 record. There is a good company here at a price I do not want.
• The volatility looks cheap, and that is the actual opportunity here. Options price roughly ±7.5%. This stock has printed +13.42% (June 25) and about +15% (July 30) single-day moves inside seven weeks.
• Cisco's Wednesday-night call helps the demand case: it took ~5 points of Q4 revenue growth from price increases on memory-heavy hardware. Memory scarcity that durable is what keeps memory capex, and AMAT's order book, going.

Is Applied Materials a Buy Before Earnings?

Not at Wednesday's close. I like the business into 2027 and I would rather own it about 8% lower, which is roughly where the down side of the implied move puts it.

My objection is to the price, and I have no quarrel with the business. Applied Materials is the toolmaker for the memory capex cycle this site has tracked since July, and that cycle keeps producing corroboration. But the stock rallied 4.29% to $548.15 on Wednesday into a two-sided event, and a 43x forward multiple on a company whose customers set capex one year at a time leaves very little room for a merely adequate FQ4 guide.

The Board

A cyclical toolmaker priced like software, going into a print that can move it 15%.

43x Is the Whole Problem

Strip out the AI narrative and look at what is being paid. 43x forward for semiconductor capital equipment is a multiple usually reserved for software. Nvidia, which sells into the same buildout with far better margins, trades near 24x. Lam Research and KLA have historically carried mid-teens to low-20s multiples through the fat part of a cycle.

That is not an argument that AMAT is expensive in the abstract. It is an argument about who is left to buy. The Street's mean target of roughly $630 is only 15% away, and several of the loudest bull targets, B. Riley at $790, Jefferies at $770, Wells Fargo at $740, were struck in late June with the stock near $627 and before the 40% summer drawdown. Those targets have not been the marginal buyer since.

What Cisco Said Last Night Actually Helps

Cisco reported Wednesday after the close and gave the memory thesis its most concrete confirmation yet. Roughly 5 points of its 18% Q4 revenue growth came from price increases on memory-heavy hardware, with another 4 to 5 points expected in fiscal 2027, and non-GAAP gross margin still fell 210 basis points to 66.3%. CFO Mark Patterson put memory at 15-20% of Cisco's bill of materials, against roughly two thirds for server makers.

Read that as a demand signal rather than a Cisco story. A networking vendor with modest memory exposure raising prices for a second straight year means the shortage is not clearing on its own. It clears when someone builds fabs, and the equipment in those fabs is what Applied Materials sells. Samsung has said the shortage runs to 2028; TSMC lifted 2026 capex to $60-64bn. The full preview of Thursday's print is here, and Cisco's results are broken out here.

Two Double-Digit Days, and the Market Prices 7.5%

Here is where I part company with the setup. Bloomberg-compiled data puts the implied move at about 7.5%, or roughly $41 on a $548 stock. One other read, from TipRanks, priced 11.03%, though it was struck against a $527.48 close rather than Wednesday's, so I am using the 7.5% figure and flagging that the two disagree by an unusual margin.

Against that, the realised record. AMAT rose 13.42% on June 25 when it unveiled six chipmaking systems aimed at AI memory, then about 15% on July 30 in the memory relief rally. Both of those are roughly double what Thursday is priced at, and they happened seven weeks apart in a stock that has since gone nowhere in particular.

This site got exactly this wrong in July. The reflex was that implied looked expensive, and realised moves beat implied repeatedly: Roblox -29%, SanDisk +26.29%, Reddit -20.76%. Applying that lesson here points one way. I would rather own the move than the stock.

The Options Angle

• Options are the better expression than shares this week. The directional bet requires a view on the FQ4 guide that I do not have; the volatility bet only requires the move to be bigger than 7.5%, which it has been twice since June.
• A straddle at the money is the clean version. It needs a move past roughly $591 or $509 to pay, and the down-side breakeven sits close to the ~$508 level AMAT traded at in early August.
• The specific risk to this play: a pre-agreed print plus a guide that merely reiterates the 30%-plus systems growth already promised produces a 2-3% shrug, and the straddle loses most of its premium overnight. That has happened in this name too.
• I could not source a live option chain, so the cost below is derived from the reported implied move rather than a quoted price. At the 11% read the play is far more expensive and I would pass on it.

Trade log

#	Stance	Structure	Strikes and expiry	Cost or credit	Spot at writing	Implied move	Breakeven
1	Long volatility	Long straddle	$550 straddle, Aug 14 weekly	~$41, ~7.5% of spot (no live chain; derived from implied)	$548.15, Aug 12 close	±7.5%	$591 up / $509 down; needs >7.5%
2	Pass	Long shares into the print	n/a, Aug 13 close	n/a	$548.15, Aug 12 close	±7.5%	Waiting for ~$505 or a confirmed FQ4 guide

The One-Line Read

The memory cycle is real and Applied Materials sells the tools, but 43x after a 4.29% rally into a two-sided print is someone else's entry. I want the move, not the shares.