Did You Pass Monday’s Test?  | BY KEITH KAPLAN
CEO, TRADESMITH |
Like you and every other investor reading this, I opened my brokerage account on Monday morning and saw blood red. Stocks were in the dumps. Especially those mega-cap tech names that have made us all so wealthy over the past two years. The reason? You think I’m about to say the new Chinese AI app DeepSeek. I’m not. The reason investors fell over themselves to sell stocks on Monday was the same thing we see in every correction. (Yes, this was a correction… NOT the start of a crash.) Emotions got the better of us. Imaginations went awry. The pain of losing money urged investors to sell… out of fear for more pain to come. DeepSeek does raise some questions about the past two years of tremendous gains in AI stocks. Frankly, I don’t have all the answers to those questions… It will take some time (more than a couple days!) to see how this plays out. But the one thing I understand better than anything is investor emotions. And because I do, I can say with confidence that the bull market isn’t over. Everything I’m seeing under the hood says this week was a shock test of investors’ emotions. Under the surface, stocks still have the green light, as I’ll show you in a moment. And odds are good that Monday’s few brave buyers will be rewarded in due time. Before I show you all the data, though, let’s touch on what DeepSeek is and why it’s scaring the AI world… ChatGPT’s “Cheap Knockoff” We should start this conversation with a simple fact: Anything out of China should be taken with a spoonful of salt. The country has a history of embellishing its accounting to attract investment. And that goes for both the public and private sectors. So, when we hear about DeepSeek’s revolutionary new large language model (LLM)… which is said to be comparable to the reigning king, ChatGPT, but for a fraction of the cost… We should keep our skeptic hat on. And there’s more reason to do so than just intuition. The numbers, if authentic, are impressive. The media is reporting the cost at under $6 million to train the new R1 model. As for OpenAI’s comparable model, o1? When CEO Sam Altman was asked if it cost, say, $100 million, he simply said, “It’s more than that.” Sounds impressive… But there’s some selective reporting here. $6 million accounts for the single successful attempt at creating R1. It doesn’t count all the failed attempts, costs for accessing data, or engineer salaries. The true number is probably much higher, though also lower than OpenAI’s. Another way to measure costs in AI is dollars per million “tokens.” We can think of tokens as each word or piece of punctuation that an LLM outputs to a user. DeepSeek’s R1 model reportedly costs $2.19 per million output tokens, against OpenAI’s o1 costing $60 per million. That, if it holds up to scrutiny, is an impressive cost savings – about 27 times cheaper. Even so, R1 is supposedly on par with o1 in terms of its text inference and writing capabilities. In coding and mathematics, it actually scores slightly better. Overall, though, we should keep in mind that OpenAI’s o1 model is capable of a lot more right now. DeepSeek’s R1 doesn’t have any image inference capabilities or audio output – only text. Those are key features that drive costs higher. These are quick facts here. I still have to do a deep dive on the tech behind this new model. But it’s enough for me to conclude that the market’s reaction was overblown. Especially when you consider the positives of this new breakthrough. More competition is always good – it’s what makes a market. And if more companies can get their hands on cheaper AI, they’ll be able to more easily apply it to their businesses and unlock productivity and growth. Chinese engineers are really good at the second-mover advantage. They take something that works, strip it down, and make it cheaper. Some would call that a knockoff, but it’s really just a boost in efficiency, and it’s where all technology eventually leads. Now, let’s talk about what we’re all thinking – how to treat this week’s correction. Why This Correction Is a Gift The smart thing to do on Monday was to buy stocks. How do I know? Two things I can show you in a chart… And one thing I can only tell you from experience. That being the reality of how large financial institutions trade and how that can quickly impact other big players. Hedge funds, investment banks, pension funds, and other large institutions use algorithmic trading strategies, which automatically act based on preset conditions. If a stock falls some amount in a short period of time, it gets sold. Some strategies even analyze media feeds and use trigger words as catalysts to act. Meanwhile, lots of investors, big and small, trade on margin. That is, they borrow money to trade stocks. That can be very dangerous in an algorithmic “flash crash”: When your trades move against you and you owe money to the broker, you’re forced to sell. That forced selling drives prices even lower. We should also think about timing. Last week, the new Trump administration announced a massive plan called Project Stargate. With it, the goal is to build $500 billion worth of AI infrastructure involving the country’s biggest tech firms. Now, this week, word has come out that a new Chinese AI model can match the best U.S. model for a fraction of the cost. And, as Jason Bodner pointed out in an urgent update to his paid subscribers, the founder of DeepSeek happens to also be the founder of a hedge fund specializing in quantitative trading. (Subscribers can access that here.) Plus, if you thought the United States was going to push for this Stargate “AI superhighway” before… You can bet it just got fast-tracked. The U.S. economy has an existential threat in China, and it won’t back down, only double down after this week’s news. But what matters so much more here is the fact that this sell-off was hardly as painful as it seemed. On the contrary, if you’re well diversified, it may have not even felt like a sell-off to you. Here’s a chart of the Invesco S&P 500 Equal Weight ETF (RSP) – that’s the green line below – along with the SPDR S&P 500 ETF (SPY), which is the blue line. Notice anything about the very far right of this chart? 
Yes – when you weight all stocks in the S&P 500 equally, the market was actually UP on Monday even though the benchmark S&P 500 was down more than 1%. And when you look at our TradeSmith Sector view, you’ll find 7 out of the 11 sectors we track closed up on Monday. 
Defensive sectors like Healthcare and Consumer Defensive led the charge higher… But Consumer Cyclical (also known as consumer discretionary) stocks were also up… along with Financials and Communications. All of those, and especially cyclicals, are more risk-on areas of the market. If this were a true bear-market breakdown moment, we’d see a bigger flight into defensive sectors. Finally, just looking at the chart of the Invesco QQQ Trust (QQQ), which tracks the Nasdaq 100, we can see that we’re still well above the TradeStops Yellow Zone at $475.80. That’s at least an 8% drop before we’d even really consider shifting to cautious. And we’re miles above the Red Zone – where we’d be confident a bear market was underway. That’s more than 16% down, at $434.29: 
This chart brings me back to the idea of investor emotions… TradeSmith’s software is like no other because it harnesses investor emotions and turns them into data with our Volatility Quotient (VQ%). You can think of VQ% as a measurement of when to worry about investors’ emotional reactions… for any of the thousands upon thousands of assets we track. For the Nasdaq 100, that level is about 16% lower than where we are today. We developed that number by looking at investor behavior over the last 26 years QQQ has been in existence. Nvidia, which has been around for just as long, has even more emotional turbulence… with its Red Zone down about 43% from its highs. That comes into play at $84.79, and NVDA didn’t even come close to that on Monday: 
(I should note, that’s based on a TradeSmith Entry Signal that would’ve seen you buy NVDA when it was down at a split-adjusted $20.64 per share!) So, listen… I know that when big red days hit, the emotional pain is severe. You want to do something to lessen that pain. You want to take risk off the table. What you’re feeling is entirely normal. But my message to you today is simple: Never let your emotions dictate your actions. You need to take a breath, look at our data, and realize that what we’re seeing is entirely within the normal bounds of volatility. A year from now, I believe we’ll look back on this day with the hindsight that it was an incredible buying opportunity. It doesn’t have to be hindsight, though… My analysts are already serving up great trades and investments. You can trust what the data is telling you – and take action. And you know what else? This will hardly be the last surprising day of price action we see this year. The AI race is accelerating. You’re going to see a lot more knee-jerk moves like this, and you need to be ready for them. So now, I ask you the question that popped into my head as I sat down to write this… Did you pass this week’s emotion test, holding firm and even buying great stocks on sale? Or did you fail, panicking along with the crowd and letting go of your quality tech companies? Let me know at feedback@TradeSmithDaily.com. And make sure you stay tuned right to these pages for all the analysis you’ve come to love. All the best, 
Keith Kaplan
CEO, TradeSmith P.S. I’ve made this promise to you many times before if you’re reading this as one of our esteemed Platinum members with access to everything we have… or if you’re subscribed to one or two of our software tools or newsletters… But if this is the first time I’ve ever written to you – then let me be crystal clear… If those market health indicators ever change – if TradeSmith signals it’s actually not a healthy correction, it’s time to sell and prepare for another bear market – we won’t keep that information to ourselves. We’ll make it perfectly clear to all of our readers. When that time comes, we’ll be just as loud about it here in TradeSmith Daily as we are about bull-market opportunities. In the meantime, I can’t wait to see what incredible trades our analysts and tools turn up – so stay tuned. | 
No comments:
Post a Comment