저자: ETFSTUDIO · 한국인
요약 보기
반도체·AI 레버리지 투자는 스토리가 클수록 변동성이 커서 심리적으로 무너지기 쉽다. 이를 막기 위해 표준편차(시그마) 기반 기계적 매수 기준을 쓴다. 가격이 1시그마(약 68%) 밖으로 하락하면 분할 매수하고, 2시그마 기준도 제시한다. 2025년 1월 이후 백테스트상 1시그마 전략은 약 24회 매수로 74.86% 수익이라고 주장한다. 평균단가를 낮춰 큰 상승장에서 수익을 노리며, 레버리지는 자산의 일부(약 20%)만 배분하라고 권한다.
다듬은 글
This is an analysis of institutional stocks. With the market fluctuating so much again, I'm sure many of you are having various thoughts . I will summarize the content of the December "Standard Deviation" for you, covering everything from investing three times in semiconductors—a prime example—investing in large industries with minimal investment, to the importance of leverage, how to invest, and even how to overcome challenges using this system.
Now, let me explain why we cannot make money with the first semiconductor leverage, even if the direction is right. Semiconductors are the heart of AI. And this SOX contains the companies you all like, such as Broadcom, AMD, Mvidia, and the semiconductors used in mobile phones, Qualcomm . Although it is limited to hardware, considering that software comes next only after the hardware is properly established, I believe that the heart of AI is ultimately the hardware .
And ultimately, it is true that AI is at the center of the story of changing the world . But now, leverage is one of the investment tools that triples the performance of that story itself . Now, even if you know all this, the reason it is difficult to convert it into profit is that AI with a big story has high expectations.
And high expectations basically mean high volatility. The SMRs, space, aviation, and satellites that we are currently looking forward to have such big stories, don't they? So, because you have to look far ahead, it produces an effect where even a slight movement inevitably leads to a large fluctuation range .
In conclusion, it is necessary to understand that the bigger the story, the more unstable and volatile it becomes . Ultimately, it is a time difference issue. When we usually make facility investments, we don't expect to break even in just a day or two . The same goes for AI semiconductors. When building a data center, buying Nvidia semiconductors, or setting up servers, most people invest with a view to capital expenditures (CAPEX) spanning five years or more, but the stock price moves day by day.
And as for our mental state, well, I described it as one second, but it is very difficult. Since stock prices move due to various factors such as earnings, interest rate issues, or President Trump's unpredictable actions, even though we know that the performance itself is actually better when compared to the SP 500, it is difficult to demonstrate that in terms of actual performance .
In other words, industry is a long-term industry. However, since stocks move every day and leverage is even more volatile, you need to understand that even if the direction is correct, the process of moving forward is ultimately very difficult. Now, I think this slide is important. The process of most people losing money is likely similar.
For example, this one is missing $7 . In other words, it suddenly dropped sharply while I was right on the verge of winning $50 . So, since most people now have confidence in AI, they think the correction is an opportunity and some are falling by 3 to 4 percent . So, since it is dropping more than expected, I am starting to feel a bit uneasy.
Then, if it drops for another day or two, you sell. In that case, well, I guess the original owner is right. Ah, I tell myself I should invest long-term in the stock, but if the price rises again for another day or two, I lose my reason again and keep buying and selling, only to end up going completely off track.
In the end, you will collapse in the psychological battle. Then, you end up selling other stocks that are doing well to invest here. Now, I have to prevent that, but I've never learned it from anyone. And the same goes for experts. Ultimately, it's investing mechanically. While the direction of the stock market is important, ultimately, what matters more is the battle of psychology and mental fortitude .
Since we have never learned this mindset from anyone, we are investing in triple leverage even though we know that the stock is more advantageous than leverage, simply because our seed money is limited . So, in order to minimize such psychological factors, institutions typically set up quantitative indicators known as "quants." When the price deviates from or falls below these indicators, they often operate as a set by adding purchases when the price drops by 10%, selling additionally or strategically selling when it drops by 20%, or taking opposing assets known as hedge assets as a spread .
That is precisely why we should exclude the psychological aspect; you must participate in the market mechanically to ultimately invest like an institution or an expert. You can only grow by being unconditionally different . Therefore, we need to re-examine the aspect that having standards is essential for survival.
Ultimately, standards are very important to at least avoid collapsing due to psychological factors . I don't see a single loss in knowing this. Okay , we are approaching this statistically, and usually there are certain sectoral ranges, you know. Even runners who finish within 10 seconds will likely fluctuate between 10 seconds, 10.5 seconds, 9.5 seconds, and 5 seconds, right?
Then, if there is a problem with Munsu, it means going outside this intermediate section . m is the probability that it occurs most frequently. It's 10 seconds. The standard deviation is the variability compared to the median . If you input this data or other stocks over the course of a year, statistically speaking, you will see that it moves within a certain range, won't you?
If you add one standard deviation from the median, it means that it mostly moves within this interval, and this probability is 68.3%, which is the majority probability. So, the point is to buy when the price moves out of the zone . I don't buy when things are going up . The idea is to buy when the price drops below this level, using this standard deviation decline as the criterion.
If you want to increase the probability distribution, you just need to double the sigma of the standard deviation . So, since this is 95.4% and 4%, unless a massive macroscopic issue arises, everything will move within this area . So, what happened as a result? If we look at the period from January 2025 to the present, 2025, there were sharp rises and falls, it went up to $7 and down to $40—it must have been absolute chaos, right?
Then, the probability that the 1 standard deviation value moves within the 68% range you saw earlier is currently -7.12. You can consider the probability of it dropping 7.12 1 2 from the median to be almost 68%, and you buy whenever it drops like that . Then, the result yields a profit of 74.86%, and there are a total of about 24 instances .
So, if I anticipate that I will miss out about 20 times in a year, and I have a seed of 100 million won, I end up investing 5 million won each time . As a result, if you take your time, you lose 20 minutes. So, since it moves three times, shouldn't I change my criteria and buy in a different way than others ?
Assuming the standard deviation is 1 sigma, the idea is to buy when it deviates from 1 sigma. It's not about timing. It approaches probability. It is about seizing the probability. So, if you try it about 20 times, I do n't think the conditions will deviate much. It is not just about predicting stock prices, but ultimately, understanding compound interest is also important.
If you just look at it, the fact is that if it drops by 14% two or three times, it gets cut in half . However, if you listen carefully to this, it sounds really good when it goes up. When it drops, it falls threefold, so this is a prime example of a mentally unstable factor that makes you lose your mind, but the account balance is high, isn't it?
But if you lower your average purchase price by buying more, it won't drop threefold. So, it has dropped this much and dropped this much more, but if you buy here and buy here, since the average price is here, it won't drop by three times. But when it goes up, it goes up three times . So, the key is to keep lowering the average price and make more profit during a major bull market .
Ultimately, I think there is no need to rush. This is the standard deviation over the same period . What did we say this standard deviation was called? It means buying whenever it deviates from the 95% range . Naturally, the number of purchases is less than 1 standard deviation . However, since you buy at a lower price, the final return shows a more outperformed trend.
Also, because you buy when the price has dropped the most without needing to waver in the middle, I think there are few factors that would cause psychological instability. I also mainly trade using this standard deviation, and in reality, winning later leads to winning bigger . My goal is to accumulate a large amount of seed capital at a low price and then make a profit when a major bull market comes once or twice .
I think it would be more helpful to set individual target rates of return, but I am looking at a 70% return with the mindset of not selling . Ultimately, the point is that if you understand the laws of survival rather than speed and stay until the end, you can at least exclude the cases where you get shaken by negative compound interest or various mental factors.
To summarize, the story is ultimately volatility. The bigger the story gets, the greater the volatility becomes. You can think of it as a scale. But what's important? It's about mental strength. It is almost impossible to overcome mental issues, and even experts do not overcome their mental state on their own.
It is overcome through structure. This sets the standard for a decline, and well, I mean, there are n't many people who hold 60% of the common stock, right? It is enough to allocate just about 20% of your assets. Leverage ultimately requires winning with technology rather than predicting stock price movements .
If semiconductors are missing, everyone will die. I think it would be better to systematize it and observe the profits for a year, rather than resorting to such extreme methods . I hope this was helpful. If there are any specific stocks you need, I will organize them and post them for you .
Comments 0