Four Steps to Investment: How to Use AI to Find Undervalued Assets?

Author: jiayi

One of the hardest things in investing is to hold back when the market is hot and to have the money and judgment to support yourself to buy when the market is bad.

Warren Buffett's saying, "Be fearful when others are greedy, and greedy when others are fearful," translates to waiting for the "bloody chips" before striking, ensuring that the risk-reward ratio is favorable enough, and then using time as an amplifier to achieve super results.

When it comes to myself, the problem arises:

Is it panic now? Are the assets that have dropped being undervalued, or is there a fundamental issue? If I buy, how much further can I tolerate the drop, and how long do I need to wait?

Recently, I watched a video using AI to monitor market sentiment, and I think this direction is worth pursuing: turning the search for "bloody chips" into a long-term process.

Let AI organize data and track changes regularly, and when worthwhile research opportunities arise, then focus attention on them.

This is especially meaningful for those who have their main business but also want to participate in Crypto and US stock investments.

Step 1: Tell AI what you want to buy.

If you only ask daily, "What’s worth buying today?" you will likely receive a list that follows the hot trends.

I prefer to first establish an observation pool that I can understand.

For example, in Crypto, I would focus on: Is there real demand? Has it achieved product-market fit? Is revenue supported by subsidies? Does the team rely on selling tokens for survival? Can token holders actually benefit? What future catalysts are there?

For US stocks, I look at: business competitiveness, cash flow, debt, valuation, and whether growth expectations are fully reflected in the price.

Think ahead: Which assets am I willing to research and buy at the right price?

Only after identifying the assets I want to buy can I know what to look for when prices drop.

Step 2: Let AI monitor if the market has started to "discount."

This step does not require AI to invent a very complex formula right away.

For US stocks, you can refer to the CNN Fear & Greed Index; for Crypto, you can refer to Alternative.me or CMC's Fear & Greed Index, combined with the observation pool's drop percentage, trading volume, and capital data.

These indices have different criteria, and their scores cannot be mixed directly. They can provide clues about market sentiment, but judging individual assets requires more information.

I hope AI can tell me:

• Is this a market-wide drop, or is it a specific sector or asset that is dropping?
• What confirmed events correspond to the drop? Which explanations are still just speculation?
• How does the current pressure compare to historical data? Is it a normal fluctuation or relatively abnormal?

If I need to score myself, I will first test it with historical data and then set fixed rules. I cannot adjust the buy line to 40 points just because "below 20 is too hard to wait for."

Step 3: When panic occurs, check if the chips are worth holding.

This is the most important step, in my opinion.

Panic means someone is eager to sell, but the next step is to judge: Why are they selling? Is the original investment logic still valid?

For each candidate, I will ask AI to provide both the buying rationale and the strongest opposing reason, and then answer four questions:

Upside: If I’m right, how much upside is there? Downside: If I’m wrong, how much could I lose? Time: How long might it take for the logic to play out? Volatility: How much volatility do I need to endure during the wait?

Especially in Crypto, having revenue does not mean the token has benefits; just because the price has dropped significantly from its peak does not mean it is cheap.

Only by looking at quality, price, and risk together can I judge whether this is the "bloody chip" I am looking for.

Step 4: Set it as a real operational task.

Choose a tool that supports online data and scheduled tasks, and set the observation list, monitoring rules, and reporting times.

For example, scan Crypto every 12 hours and review US stocks after each trading day closes. Provide briefings during normal times, and when preset conditions are triggered, list the assets that need to be focused on for research.

It is necessary to confirm that the task has been successfully created and that a report has run through once. Simply saying "remind me every day from now on" in the chat box does not mean it is running in the background.

Scanning every 12 hours also means that anomalies can only be discovered during the scan, not treated as real-time monitoring.

The following prompt can be used as a starting point:

Please establish a Crypto/US stock opportunity monitoring process based on my investment preferences and observation list. First, check data sources and historical coverage, then use public sentiment indices, prices, and trading volumes to identify abnormal pressures. Analyze the market, sectors, and individual assets separately, indicating data time, sources, and missing items. When a significant drop occurs, verify whether it is a short-term shock or a fundamental deterioration. For candidate assets, list quality, valuation, catalysts, strongest opposing reasons, as well as upside potential, downside risk, waiting time, and volatility. Do not fabricate target prices and probabilities without basis. Panic scores should only trigger further research. Buying must also meet asset screening and risk budgeting; thresholds that have not been historically tested should only serve as observation reminders. If you support scheduled tasks, please create them at the frequency I specify and confirm the next run time; if not, please clarify. The report should prioritize informing me: What changes have occurred, which assets are worth further investigation, and which situations should continue to wait.

My idea is to first set up this process and then use historical testing and actual running records to determine its usefulness. The formulas organized by AI still need to go through this step.

In investing, there are many times when there just isn’t an operation that suits you.

The value of AI is to reduce the cost of continuous research and monitoring, allowing oneself to be less swayed by emotions and to miss worthwhile opportunities due to busyness.

Ultimately, what I want to earn is the money from good assets bought at reasonable prices, followed by value recovery and long-term growth.

Regularly study diligently, set conditions for prices, set boundaries for positions, and then leave time to live your own life.

This content is provided for general informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online promotions, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets. Crypto assets are highly volatile and may result in loss. The availability of WEEX services, products, and related events may vary by region. You are responsible for ensuring that your participation is in accordance with applicable local laws and regulations.