Bank of America stays bullish on AMD (AMD), keeping its buy rating and raising its price target to $720 from $620. That marks about 14% upside from the prior close. The core argument: the AI compute opportunity runs on both GPUs and CPUs, and CPUs are a key part of the puzzle where AMD (AMD) fits. Bank of America points to Anthropic's $11.6 billion deal with Okami as proof that agentic AI will drive demand for CPU-heavy cloud infrastructure.
The stock hit a new all-time high, and CEO Lisa Su spent the week in the spotlight - at a White House dinner with President Trump and China's Xi Jinping, and ringing the New York Stock Exchange opening bell alongside First Lady Melania Trump.
The Bull Case: Momentum and CPU Demand
Shorting a stock with this much momentum is hard. AMD (AMD) is up 40% in the past month. The question to ask: will it keep going?
An S&P Global article compared factors against yields. When treasury yields rise fast, the factors that perform best include high beta. AMD (AMD) trades at a beta around 2.4, so it clearly counts as a high-beta name. This matters because investors are turning into stock pickers rather than buying the index broadly. Market breadth backs this up: the share of S&P 500 stocks trading above their 50-day average sits around 25%, and it is below 50% for the 200-day. The play is to find the winners and ride the trend. Given the momentum and the CPU demand story, there is no reason to look for bearish trades in a stock showing this much strength.
Edge AI as a New Profit Stream
Edge AI is the practice of running machine learning and AI models directly on local hardware devices instead of the cloud. This drives more CPU demand, and that demand is set to keep exploding. When the deal news hit, AMD (AMD) rallied because the company sees the picture and the race to lock up CPUs in long-term deals. The same upside applies to Intel (INTC) and Nvidia (NVDA), which are also moving into the CPU business. Edge AI opens a whole new profit stream for these companies.
Trade One: A Slightly Bullish Diagonal
Both traders start by checking relative implied volatility over the past year. The current implied volatility percentile rank is 15% (and dropping - later cited near 13%), so volatility is relatively cheap versus the past year. When volatility is that low, selling spreads is unattractive, so the move is to buy a time spread.
The chosen structure is a diagonal: sell the October 5th expiration 660 call, buy the October 16th expiration 630 call, for about $2,360 at pricing. It works like a covered call but with a far smaller cash outlay. A covered call on $620 stock costs about $62,000 minus the call premium; this diagonal costs about $2,360 for a neutral-to-bullish position.
The trade profits as the stock moves up toward the 660 strike, with diminishing returns past that point. Because it trades below the width of the strike (the $30 gap between the 630 and 660 strikes), the profit graph curtails somewhat above 660 rather than cratering, which would happen if it were trading above $30. It gives room to capture upside if the stock keeps rallying. This is a short-term trade - selling October 5th, buying October 16th, not looking out 90, 100, or 150 days. It needs some upside movement toward the 660 strike.
Trade Two: A Bearish Put Calendar
A diagonal sits in the calendar spread family but has a vertical spread embedded in it, giving the $30 difference between the long and short strikes, plus time value. If the market allows, duration can be extended to chip away at the debit and lower it.
The second trade is a straight calendar spread. Like Meta Platforms (META), AMD (AMD) is getting overbought on the relative strength indicator. After measuring the expected move, the structure is a one-week put calendar: buy the October 9th, sell the October 2nd $600 put. Both trades buy Vega at the 13 percentile - a key point, since the strategy should match the volatility level: know which strategies to use when implied volatility is high and which when it is low, and a calendar spread suits low volatility.
The put calendar gives two things: a directional play (short about 5 deltas) and about 14 Vega, so it benefits from a downward move and from any rise in implied volatility. It was priced around 540 and traded near 550 as the stock backed off slightly. The lesson: understand implied volatility, buy it low and sell it high. The diagonal substitutes stock for the longer-dated option and is the more aggressive bullish play; the calendar is a less aggressive bet on a move down toward 600.
Both trades have defined risk - the most you can lose is the money spent on them. Care is needed near expiration if the stock lands close to the strike.
Choosing a Side
It comes down to one question: does the stock have room to grow, or will it consolidate after its big move? A case can be made either way. It has already rallied 40% in the last four or five weeks, so it could take a breather, sit flat, or pull back. If it pulls back toward the 600 strike, the put calendar works out extremely well. The diagonal does not need a large upside move, but it does need some upside. That trade-off is the bull-versus-bear choice between the two positions.
