Hello! I'm your financial analysis assistant. This report on ByteDance's spin-off of its AI pharmaceutical company, Anew Labs, is packed with valuable information, revealing both the clever strategies behind the capital moves and profound insights into industry trends.
To help you easily understand the intricacies, I will first summarize the key points in one sentence, and then break down the logic from five different perspectives.
📝 Key Content Summary
In one sentence:
ByteDance has separated its in-house AI pharmaceutical team to establish the new company Anew Labs, raising $290 million in funding (with a valuation of $1.5 billion), while retaining 56% of the shares. This is not only another strategic expansion of ByteDance's AI efforts but also a significant attempt by a Chinese tech giant to challenge the traditional pharmaceutical R&D paradigm with its “internet speed” and computational power. The market currently values the platform’s capabilities, but the real test will come in the next few years when it needs to produce clinical data to prove that AI can indeed develop effective drugs.
---
🔍 In-Depth Analysis: Understanding Anew Labs from Five Perspectives
1. **Why the Split?** – Solving the Conflict Between “Fast” and “Slow”
Many may wonder why ByteDance would separate its pharmaceutical team instead of keeping it within the company.
It’s like asking a team of delivery riders to build rockets:
- The internet operates on a weekly or even daily basis: TikTok and Toutiao focus on rapid iteration, launching new features, analyzing data, and making improvements daily. The metrics are daily active users and retention rates, emphasizing efficiency and speed.
- Pharmaceutical development, however, takes years: Developing a new drug from lab to market typically takes 10 years and costs billions of dollars, with many uncertainties along the way, and there might be no returns for the first 9 years.
If these two teams worked in the same office under the same KPIs, it would be disastrous:
- Pharmaceutical scientists would be pushed crazy by the internet’s demand for quick results, as scientific experiments require patience.
- Algorithm engineers would find pharmaceutical development too slow and lack a sense of achievement.
The Purpose of the Split:
By separating the teams, Anew Labs is given its own timeline and set of metrics (clinical progress rather than daily activity). Google previously split its DeepMind drug team into Isomorphic Labs for the same reason. This isn’t about cutting off ties but about allowing the pharmaceutical business to grow at its own pace while still leveraging the parent company’s speed and resources.
2. **Is the $1.5 Billion Valuation a Bubble or a Vision?**
A company with only 50 people and four pipelines that haven’t entered clinical trials is valued at $1.5 billion (about 10 billion RMB), which is on par with the market value of a established pharmaceutical company like Schrödinger. Is this investment worth it?
We need to consider two aspects:
- Pipeline Value (conservative estimate): Each of the four pipelines, if they haven’t entered clinical trials, would be worth tens of millions to a billion dollars at most, totaling around $500 million.
- Platform Value (the premium): The remaining $1 billion buys the “platform” that can develop drugs.
Why the Market Is Willing to Pay for This Platform?
Previously, AI in pharmaceuticals was seen as simply using AI to calculate molecular structures. Now, the real barrier is the system capability of “data + models + computational power.”
- ByteDance’s expertise in finding optimal solutions in large datasets can be applied to drug molecule design.
- ByteDance has powerful computing resources (Volcano Engine), which is one of the most expensive aspects of AI drug development.
- Investors bet that once this platform is operational, it will continuously generate candidate drugs, and even if some fail, the overall success rate will likely be higher than that of traditional pharmaceutical companies.
In simple terms: The market is buying the potential to mass-produce drugs in the future, not the drugs themselves.
3. **Why Does ByteDance Only Hold 56% of the Shares?** – A Strategic Move
ByteDance doesn’t own 100% of the shares and hasn’t sold them all, retaining 56%. This ratio is carefully crafted:
- Full Ownership (100%): ByteDance would bear all the risks of R&D failures and still be influenced by the internet’s fast pace.
- Complete Sale (0%): ByteDance would lose this strategic asset, potentially giving up future opportunities and even creating competitors.
- 56% Ownership (the Best Balance):
1. Control: ByteDance maintains control, ensuring the technology stays on track and continues to provide computational support.
2. Independent Financing: Anew Labs can raise funds on its own (as it did this time), reducing financial pressure on the parent company.
3. Financial Potential: If Anew Labs goes public or its value surges, ByteDance as a major shareholder can benefit significantly from capital appreciation.
4. Flexibility in Exit: If the strategy fails, ByteDance can gradually reduce its stake without dragging down the entire company.
This is a classic capital move that allows for both offense and defense.
4. **The Real Bottleneck Has Changed: From “Fast Calculation” to “Fast Testing”**
The article points out an counterintuitive fact: The bottleneck in AI pharmaceuticals is no longer the speed of computation but the speed of experiments in the lab.
- Past Bottleneck: Scientists used to rely on experience to design molecules, which was slow. Now, AI can generate tens of thousands of candidates per second, so the design phase is no longer the bottleneck.
- Current Bottleneck: AI generates many candidates, but laboratories can only synthesize and test a few per week. The speed of “real” chemical experiments lags behind that of computer simulations.
What Does This Mean for Anew Labs?
Anew Labs is strong in algorithms and computing power but weak in lab capabilities. Therefore, it needs partners, such as a traditional Chinese pharmaceutical company, which has mature labs, clinical resources, and registration experience. This combination of AI and traditional pharmaceuticals is the most practical approach.
5. **The Future of Drugs: From Luxury to Fast-Consuming Goods**
The report also discusses the choice of targets (such as IL-17 and IL-4R). Why these targets?
- Traditional Models: Many expensive biopharmaceuticals (e.g., antibodies) are costly and require cold-chain transportation, accessible only to a few patients.
- Pressure from Payors: Global health insurance systems (including China’s and the US’s) are pressing down on drug prices.
- Anew’s Strategy: It targets autoimmune diseases with large patient bases (tens of millions) that require lifelong treatment.
- Breakthrough: Anew aims to replace these with small-molecule oral drugs, which are cheaper, more convenient, and less costly to produce.
- Business Logic: Although the price per drug may be lower, the total profit could be higher due to the larger patient base and longer treatment duration, making it more acceptable to insurance.
This means: Anew Labs is not aiming for expensive, niche drugs but for mass-produced, affordable, and long-term consumable drugs, which aligns with the future trend of the pharmaceutical market.
---
💡 Insights for the General Public
1. Don’t Just Look at the Valuation: A $1.5 billion valuation is attractive, but it’s based on clinical data by 2027-2029. Poor data could significantly reduce the value.
2. Pay Attention to the Integration of AI and Traditional Industries: ByteDance’s role is to empower traditional industries with AI, not to replace them.
3. Patience Is a Scarce Resource: In the internet era, we’re used to speed, but in life sciences, patience is essential. Anew Labs’ success depends on ByteDance’s ability to maintain a slow pace within a fast-paced culture.
In summary, Anew Labs is like Schrödinger’s cat: before the clinical data is available, it’s both a $1.5 billion technology star and a potential failure. 2027-2029 will be the critical moment to determine its fate.