My Stop Losses Got Hit. Here's What I'm Doing Next. — Transcript
Full transcript
- 0:00Five AI stocks in my portfolio fell
- 0:02together this week, and my stop losses
- 0:04triggered for two of them. That was the
- 0:05plan. Three of the five are now at the
- 0:07price where I'd normally start adding.
- 0:10I'm waiting on all three. One of the
- 0:12five is down more than half since June.
- 0:15On my chart, it's dirt cheap, and it's
- 0:18the one I'm being most careful with. So
- 0:20I ran all five from the same four-step
- 0:22checklist I use every red day. By the
- 0:25end, you'll have the checklist, the
- 0:27exact stop rule that fired, the two
- 0:29chart lines I use to pick my prices, and
- 0:32the prompt I use to run all of it. I'm
- 0:35not a financial advisor. I own a
- 0:37position in some of these. I landed in
- 0:39London off a red eye to a red market.
- 0:42The first thing I did was nothing. I
- 0:44went for a walk. If a red day makes your
- 0:47stomach drop, that's information. It
- 0:50usually means you're holding more risk
- 0:52than you're actually comfortable with.
- 0:53My red day checklist has four steps.
- 0:56Step one, every stock gets three
- 0:58questions. Why did I own it? Did the
- 1:00business change or just the price? Is it
- 1:03still growing? Step two, if a stock
- 1:06moves 5% on a normal day, I stop adding
- 1:09to it. Step three, trailing stops on
- 1:11every position set before the drop. A
- 1:14trailing stop follows the price up
- 1:16and sells if the stock falls a set
- 1:18amount from its high. Mine is 15%. If
- 1:22you're new to stops, this is how I'd set
- 1:24one up. Say a stock is at $200. The stop
- 1:28sits at 170. If the stock climbs to 250,
- 1:31the stop climbs to $212.50.
- 1:34If it falls back, the stop stays put.
- 1:37I put stops on part of each position,
- 1:39less on Nvidia because I've owned it
- 1:41since 2017.
- 1:43The rest stays in in case the stock
- 1:45swings back up. And I start in my IRA
- 1:48where my Nvidia sits because selling
- 1:50there doesn't trigger a tax bill right
- 1:52away.
- 1:53How much you cover depends on how much
- 1:55risk you can handle, and that's
- 1:57different for everyone.
- 1:58If last week kept you up at
- 2:00cover more. Those stops are why I didn't
- 2:02have to make a single decision in a
- 2:04panic this week. If you set a stop this
- 2:06week, it trails from this week's price.
- 2:08It protects what you have now. Step
- 2:10four, write down your dirt cheap prices
- 2:12while you're calm. My starting point is
- 2:15one of two lines on the chart. The first
- 2:17is the 200-day EMA. That's roughly the
- 2:19average price over the last 200 trading
- 2:21days with recent days counting more. The
- 2:24second is the 61.8% Fibonacci level.
- 2:27Take the stock's last big run from the
- 2:30low to the high. When it has given back
- 2:3261.8% of that run, it's at the level.
- 2:35When a stock reaches one of those lines,
- 2:37I go back to step one before I add a
- 2:39dollar. So for these five AI stocks,
- 2:41demand held up this quarter. I see that
- 2:43demand in my own work. I use Claude and
- 2:46ChatGPT every day, and I've handed them
- 2:48a lot of my admin and logistics. If one
- 2:51person gets that much out of it, picture
- 2:53what a big company can do sitting on
- 2:55years of its own data.
- 2:57What changed is how these companies pay
- 2:59for the build. Three of these five are
- 3:01borrowing or spending more than they
- 3:03bring in. Keep that in mind because it
- 3:05decides which ones I trust most right
- 3:08now. Amazon first because it has the
- 3:10cleanest chart. Why I own it.
- 3:13AWS, Amazon's cloud business, which
- 3:15rents computing power to everyone else.
- 3:18And the best explanation of why I'm
- 3:20bullish came from chatting with a former
- 3:21AWS employee. Now the big models are
- 3:24closer in quality. So companies ask a
- 3:27different question. Where does our data
- 3:28already live? If it's already on AWS,
- 3:31AWS usually wins the AI work too.
- 3:34A lot of companies buy Claude through
- 3:35Amazon instead of straight from
- 3:37Anthropic because their data never
- 3:39leaves AWS. The part I like most, that
- 3:42former employee believes Amazon makes a
- 3:44higher margin on open models, the free
- 3:47ones anyone can run, than it does on
- 3:49Claude. Claude gets the customer in the
- 3:51door. Once their data is on AWS, they
- 3:54use the cheaper models too, and Amazon
- 3:56keeps more of each dollar. Did the
- 3:58business change? The spending did. Over
- 4:01the last 12 months, Amazon spent $42
- 4:04billion more than it brought in.
- 4:06It plans about $220 billion on the
- 4:08build-out this year and says it'll spend
- 4:11more next year. Is it still growing?
- 4:13Yes, and faster.
- 4:15AWS grew 39% last quarter, the sixth
- 4:18quarter in a row it's sped up.
- 4:20And customers have signed $578 billion
- 4:23of contracts it hasn't delivered yet. On
- 4:26the chart, Amazon is still above its
- 4:28200-day EMA.
- 4:30That line is about 6% lower.
- 4:33The Fibonacci level is about 12% lower.
- 4:35It's down less than 10% from its high.
- 4:38So the stop didn't fire. I still hold
- 4:41all of it. The next one past the growth
- 4:42question even more.
- 4:44It's how it's paying for that growth
- 4:46that caught my eye. Alphabet, Google's
- 4:49parent. Why I own it.
- 4:51Search and YouTube pay for everything
- 4:54else. And Google makes its own AI chips.
- 4:57And I think the same data advantage
- 4:59protects Google Cloud. Companies already
- 5:02running on Google tend to keep their AI
- 5:04work there. Did the business change? How
- 5:07it pays for things did. Since December,
- 5:10its long-term debt went from about $47
- 5:12billion to $125 billion. Last quarter,
- 5:16it sold $12 billion of new stock and
- 5:19bought none back.
- 5:20Its free cash flow for the quarter was
- 5:22negative $9 billion. Is it still
- 5:24growing? Very fast. Google Cloud more
- 5:27than doubled to almost $31 billion in
- 5:30one quarter.
- 5:31And it has $561 billion of signed cloud
- 5:34contracts waiting.
- 5:36On the chart, Alphabet is already below
- 5:38its 200-day EMA. So by my rule, it's at
- 5:41my starting point. That sends me back to
- 5:43step one.
- 5:44The growth passes, but it's borrowing to
- 5:46pay for it.
- 5:48My stop didn't fire because I forgot to
- 5:50renew it and it had expired.
- 5:52That's on me. My reason for owning
- 5:54Alphabet hasn't changed. So I still hold
- 5:56all of it. I'm not adding right now, but
- 5:59it's getting pretty close. AMD is where
- 6:01step two kicks in. Why I own it.
- 6:04It's the main alternative to Nvidia for
- 6:07AI chips.
- 6:08Did the business change?
- 6:10Its debt tripled this year
- 6:12to about $7 billion.
- 6:14Is it still growing? Yes. Data center
- 6:16sales nearly doubled, up 97% to a record
- 6:20$8.5 billion.
- 6:22And AMD guided next quarter to about $17
- 6:24billion in total sales, up about
- 6:26two-thirds. But AMD has been moving
- 6:28almost 5% on an average day lately, more
- 6:31than twice Amazon. So by step two, I
- 6:34stop adding. AMD's stopped fired, so I'm
- 6:38already smaller here. I also hedged my
- 6:40whole portfolio with puts on Nasdaq 100
- 6:42futures, which track the biggest tech
- 6:44stocks. If the AI trade keeps dropping
- 6:46together, those puts go up and cushion
- 6:49the fall. On the chart, it bounced back
- 6:51above its 200-day EMA this week. The
- 6:53Fibonacci level is about 19% lower.
- 6:56That's where I'd start looking again.
- 6:58And I'm still bullish on AMD for the
- 7:00long run, even though I think the price
- 7:02already bakes in a lot of good news.
- 7:04AI agents are growing fast, and agents
- 7:07lean hard on regular processors, CPUs,
- 7:09as well as AI chips. AMD makes both.
- 7:12On November 2nd, Intel CEO Pat Gelsinger
- 7:15bought almost $10 million of Intel stock
- 7:18with his own money right after the
- 7:20sell-off. The person running a CPU
- 7:22company is putting his own cash on CPUs.
- 7:25The last two are at my price. And I 'm
- 7:28treating them very differently. Nvidia.
- 7:31Why I own it. CUDA. It's the software
- 7:34developers build AI on, and it runs on
- 7:36Nvidia chips. Once a team has years of
- 7:39code on it, switching is expensive. Did
- 7:42the business change? China. Nvidia's
- 7:44guidance now assumes zero data center
- 7:46sales there. Is it still growing? Data
- 7:48center revenue more than doubled, up
- 7:50117% to $89 billion in one quarter. And
- 7:55it pays its own way. It made $21 billion
- 7:58of free cash flow in that same quarter.
- 8:01On the chart, Nvidia is below its
- 8:03200-day EMA, sitting right on its 61.8%
- 8:06Fibonacci level. By my own rule, that's
- 8:09the starting point.
- 8:10My stop sold a small part of it. It's my
- 8:13biggest winner. So, why not buy that
- 8:15back now? Nvidia reports on November
- 8:1717th. When a stock hits my level with
- 8:20earnings that close, I let the report
- 8:22answer step one.
- 8:24If the story holds, that's when I add,
- 8:27which brings me to the one I teased at
- 8:29the start. Down more than half since
- 8:32June. Marvell. On my chart, it's far
- 8:35below both lines. If step four were the
- 8:37only step, I'd be adding with both
- 8:39hands. Why I owned it.
- 8:41Marvell designs custom AI chips for the
- 8:44big cloud companies and the optical
- 8:46parts that move data between data
- 8:48centers. Did the business change? Yes.
- 8:50On November 2nd, Marvell agreed to buy
- 8:53Lumentum in an all-stock deal worth
- 8:55about $115 billion.
- 8:57Lumentum's shareholders will own about a
- 8:59third of the combined company.
- 9:01And it isn't expected to close until the
- 9:04first half of 2028. That's a different
- 9:06company from the one I bought.
- 9:08Is it still growing? Revenue grew 37%
- 9:11last quarter, and data center grew 46%.
- 9:14Marvell reports again on December 1st.
- 9:17Question two changed its answer. So, I'm
- 9:20redoing my thesis from scratch. A cheap
- 9:22chart only answers step four. Step one
- 9:25comes first. This is the prompt I
- 9:26actually used to run all four steps. It
- 9:29even checks whether my stops should have
- 9:31fired. You'll find it free at
- 9:32resetandinvest.com/free.
- 9:35Where did five landed? Amazon and
- 9:37Alphabet. I hold all of both.
- 9:40AMD, my stop sold part, and I'm holding
- 9:42the rest. Nvidia, same thing. My stop
- 9:45sold part of it, and I'm waiting for
- 9:46November 17th. Finally, Marvell, I'm
- 9:48starting the thesis over. Five stocks
- 9:50that fell together, five different
- 9:52answers. That's the whole point of a
- 9:54checklist. The plan you write on a calm
- 9:56day knows more than you do on a red one.
- 9:59My newsletter is where I show you how I
- 10:01use AI to build a game plan for my
- 10:03portfolio and stick to it. If you want
- 10:06that, the links in the description.
About this transcript
This page contains the full transcript of My Stop Losses Got Hit. Here's What I'm Doing Next. by Koh Kim | AI + Wealth, generated from the public captions YouTube serves with the video. The transcript has 1,767 words across 258 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.
What you can do with it
Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.
Free YouTube transcript tool
YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.