Source Code Capital | Guangmi Li: Choosing Entrepreneurship Was My Leap of Faith

Source Code Capital

Guangmi Li, formerly an investor at HSG, is now CEO of Shixiang Tech. He interned at HSG as a junior in college, and after graduating in 2015, joined the firm's TMT team full-time, launching his career as a young investor.

As a perceptive young investor, Li witnessed the rise of multiple star companies during the era of "tech's great voyage." At the end of 2019, he leaped from being an observer to becoming a "sailor" himself — founding Shixiang Tech to actively push more Chinese entrepreneurs to join this global expedition.

Livestream Excerpts

Xingchen Zhang: As a former investor turned entrepreneur, why did you leave HSG to start your own company?

Guangmi Li: HSG was my first job. The emotional attachment ran deep — I once assumed I'd work there until retirement. I left because I hit a personal growth ceiling. I felt powerless.

When I started out, I was green. I could source good deals, but lacked the judgment and perspective to identify great founders. I missed out on some excellent investments. Despite working hard, my growth never satisfied me.

I always felt there was a restless "seed" inside me — an entrepreneurial ideal buried in my heart.

Zhang: Why late 2019 specifically? Did you spot an opportunity, or were you responding to external changes?

Li: Personal reasons came first. A mentor of mine gave me four words: "Start a company early." That ignited my resolve and made the seed truly begin to grow — it gave me permission to pursue something different.

External factors mattered too. I'd focused on software investing, but I felt opportunities in my area were shrinking. I hadn't figured out what the next big thing would be. I was in a fog.

Meanwhile, regulators kept repeating "housing is for living, not speculation." I started thinking: where would the hundreds of trillions in real estate capital go? My answer was asset management.

Globally, finance and asset management remain the largest sectors. In China, no trillion-dollar company had emerged in this space yet. There was room to maneuver. For the next two and a half years, I kept at it, iterating and trial-and-erroring along the way.

Zhang: What does Shixiang Tech do today? And why "Shixiang" — what's the meaning behind the name?

Li: Shixiang Tech was founded to provide Chinese entrepreneurs with the highest-quality investment research services, aiming to build the next-generation global investment platform.

We're a research-sharing platform built by a globally-minded, top-tier investment team, focused on studying the world's most innovative companies. Our current product lines include alternative investments, Haiwai Dujiaoshou, Shixiang Talk, and curated fund portfolios — helping Chinese LPs with optimal allocation, becoming infrastructure for the asset management industry. We never wanted to run traditional funds; we're more about innovation.

The name "Shixiang" — we chose the elephant as our mascot, the largest land animal. "Shi" (拾) means to gather or increase, but also carries a youthful, ambitious energy.

Zhang: From research to execution, what did that journey look like?

Li: Initially we wanted to build a Fund of Hedge Funds, but realized my core strengths weren't there. Then an entrepreneur LP asked us to help buy secondary shares — he paid us a fee, which became our first deal and first revenue. Later, that same mentor who said "start a company early" nudged me further, helping me see that my real edge was in private markets. We iterated constantly. We started in China; about a year ago we moved into the US market. Now our whole team is based there — I spent much of the past year on team-building.

Beyond team changes, there were market shifts. Entrepreneurship means iterating with the market and your evolving understanding. Shixiang Talk, for example, is essentially a decentralized investment committee — LPs can join, other opinionated investors can share. We'll launch indices in the future too.

Anyway, every step of a first startup is searingly memorable — not just business exploration, but first revenue, first hire, first firing.

Zhang: Did you have role models during this process — an entrepreneur or company?

Li: I'm always looking for benchmarks. I think it's critical for founders to actively study the world's best companies. But we need to learn their thinking and operating methods, not just copy business models. We've learned excellent habits and practices from ByteDance, KE Holdings, Amazon, and Google.

Our team habitually does deep research on leading global financial services, investment services, and investment innovation companies — Blackstone, Futu Holdings, S&P Global, and others.

We're also doing content output, pushing the industry to a higher level. The act of outputting deepens our research and understanding, while helping us think through how to redefine investment services, how to better develop our team, how to study innovative business models.

Zhang: Being able to identify who to learn from is itself a sophisticated skill. How should entrepreneurs figure out their learning targets?

Li: First, genuine admiration. When someone does something well, you truly appreciate it deep down — that's the starting point. Second, stage fit. Your阶段性 goals and capabilities need to match. However well others perform, if you can't execute at your current stage, or market conditions don't allow it, you need to define your stage and timing, and find learning targets suited to where you are. Third, values alignment. This overlaps with the first point — you need genuine conviction.

Zhang: How do you understand innovation in asset management — using scientific methods to reorganize the paradigm?

Li: Overseas and Chinese LPs have fundamentally different needs. US institutional LPs are relatively passive. Pensions, endowments, family offices — they've been through multiple generations, they have a professional-manager culture, they focus on long-term stable returns. Chinese LPs want high involvement; their demands for information and value are very high.

LP services will fundamentally change through scientific methods. Efficiency, innovation, client experience — all will differ. For example, using blockchain smart contracts to replace lawyer-drafted contract terms.

I believe traditional fund models are ultimately measured by brand and returns, but building a new fund brand takes at least a decade. Many institutions won't survive ten years; even if they do, they may not build a good brand. A more efficient approach is building products and brands. Take "Cathie Wood" — she bypassed traditional distribution channels and built her own system. That's innovation too.

From an investor's perspective, as a researcher you might spend a month studying a company, present to the IC, then the project gets passed — that month's work gets shelved. Shixiang wants to open-source these research findings. For now we can only open-source our own, but we'll do more in the future, forming a decentralized investment committee product.

Using scientific methods and technology to organize financial assets is a somewhat idealistic vision that may fail, but we still want to try and explore.

Zhang: In this exploration, you've developed a product matrix — Shixiang Talk, the Haiwai Dujiaoshou account, etc. How do these relate, and how do they help you find that "pearl" you're seeking?

Li: These products are our process of turning team capabilities into products. Alternative investments, for example, are like our commercial product. Haiwai Dujiaoshou is like our Xiaohongshu-style discovery and brand product. Shixiang Talk is like our livestream platform — a decentralized investment committee where we discuss with LPs. Because we believe in open-source, our business model is currently weak; besides alternative investments, most products are still investment-heavy.

Long-term, we want to become a SaaS company. We think Wind's model is excellent. We may launch a unicorn index in the future — indices might not make money, but they're useful for the industry, a weathervane, and could potentially become an ETF.

We're still in the phase of trying new products, unsure which will become truly important. But we'll keep expanding based on market intuition and client needs.

Zhang: Haiwai Dujiaoshou has become an IP — the industry even gave it a nickname, "Haidu." The brand is well-built. How do you produce high-quality content? Any tips?

Li: I think the core is investment taste and methodology. My experience at a large fund taught me the value of global perspective and the finest investment taste. So from Haiwai Dujiaoshou's launch, we studied fields where overseas markets were ahead of China, sharing that experience and advantage with Chinese entrepreneurs and investors.

Second, we only study the best companies globally. This produces high research ROI, rapidly elevates young team's perspective, and builds good investment taste.

Third, and most importantly, an open-source and altruistic mindset. Entrepreneurs need commercialization ability, but we should approach this with a public-service mentality — serving the masses, building a "public good." That's what gives it meaning.

Zhang: You've been in the US for a while, seen many companies. Will Silicon Valley continue to lead future innovation?

Li: Silicon Valley has given the world computers, semiconductors, the internet, Tesla's electric vehicles. We came with high expectations. But once here, we felt global information has flattened — most things are findable online.

Though we haven't figured out what the next great revolution will be, I can share some observations. For 20 years, Silicon Valley VCs only invested in "soft" things — internet and software, mainly for high gross margins. Perhaps because so much US manufacturing moved overseas; besides Tesla, there aren't many factories.

The past decade continued along the "digital revolution" — cloud computing, digitization, SaaS — with strong scale effects. SaaS here resembles enterprise consumer goods, with powerful brand effects. The digital revolution is what Silicon Valley deeply values now, the "soft" stuff. Whether it can keep leading innovation, everyone is waiting for the next digital revolution.

China is currently going "hard" — manufacturing and the "energy revolution." Energy revolution drives both technology and consumption, and is a massive future growth engine. Our feeling is global complementarity: Silicon Valley contributes internet software, technology dividends spilling globally; China focuses on manufacturing and the energy revolution.

Zhang: What changes or innovations do you see in US asset management, including VC?

Li: Silicon Valley has abundant VC capital, and many investors are actively sharing — their investments, sectors, market views. These perspectives drive interaction and connection, helping entrepreneurs better understand investors.

Also, early-stage investing is quite decentralized. Silicon Valley angel check sizes are smaller than I imagined, but participation is broad — many people writing $100K-$300K checks, then helping as advisors. New-style funds keep emerging in Silicon Valley, though they're still rare in China.

Third, Silicon Valley investors have more reasonable valuation expectations; competition among peers isn't as intense.

Zhang: Do you think this offers real lessons for Chinese VC? Where do you see domestic VC heading?

Li: I think the biggest lesson is that capital and policy are two crucial levers. In Silicon Valley, they idealize spotting waves and timing, focusing on bigger changes. For China, that means paying attention to policy changes.

Silicon Valley is more willing to support idealistic things, to define new trends. Selection beats effort — we believe big things emerge from big trends. So domestically, we need to closely follow policy direction, understand market trends, to find that big thing.

• MaHui | Yungang Huang, Source Code Capital: How Does AI4S Land Across Different Stages, Industries, and Markets?

• MaHui | Yungang Huang, Source Code Capital: "Cross-Disciplinary Intersection" Drives Healthcare Innovation

• MaHui | Source Code Capital: Investing in Carbon Neutrality Requires Multi-Level Cognition

• MaHui | Source Code Capital: Opportunities and Advantages — China's New Energy in a New Cycle

• MaHui | "Carbon Neutrality" Technology Innovation in Sustainable Development

• MaHui | Is 41 the Optimal Age for Innovation?

• MaHui | Source Code Capital: Focusing on Energy Cleanization and End-Use Electrification

• MaHui | Yungang Huang, Source Code Capital: Maintaining a Steady, Long-Term Tech Investment Rhythm Through Peaks and Troughs

• MaHui | Source Code Capital: Technological Change Reshapes Moats, Tech Entrepreneurship Breaks Boundaries

• MaHui | Xingchen Zhang, Source Code Capital: Tech-Driven Innovation from 0 to 1

For more MaHui members, click "Read More"