“Butterfly effect” is a scientific term used to describe the observation that a small change in one state of a deterministic nonlinear system can result in large differences in a later state. In a classical metaphorical example, “the flapping of wings of a butterfly in Brazil, through a chain reaction, can cause hurricane in Texas, USA”. Finding this term quite interesting, I borrowed it to describe my own investment approach. Let me explain. In United States, five thousand companies are listed on major stock exchanges and many more private entities, large or small, operate in different industries. In total, United States boasts roughly 30 million companies, or one company for every 10 people. Around the world, more than 45 thousand companies are listed on stock exchanges, and the total number of companies including private entities exceeds 100 million. That number is posed to grow even larger, as world population grows and economies in developing countries continue to advance. What does that mean for investment community? For one thing, picking winners from the vast number of companies seems like treasure hunt in the great wilderness. The process increasingly appears more like something being described by the Chaos Theory. Randomness may very well dominate, prompting questions about the usefulness of any investment research effort. How can we conduct efficient and productive research? Diving deeper into the fundamentals, one can certainly see that these businesses (private or public) don’t operate independently. Instead, through different interactions, they form a vast network. This network is comprised of distinct clusters, and within each cluster, companies’ fate are closely linked together through relationships such as supplier-customer, partners, competitors and cross company investment/ownership. Companies in each cluster do not have to be in the same industries, or even in the same countries. They come together just because of their naturel business relationships, and their financial performance (or stock performance for public companies) tend to correlate. Besides this company network concept, we should also understand the external factors affecting this network. A sudden change in external factors (Japan central bank lowering interest rate, or report on China housing slowdown) could directly affect certain companies, and then through these clusters and network, indirectly affect other companies. This, in my mind, is similar to what the Butterfly Effect describes. So to quickly understand what the consequence a little butterfly can cause, one should do two things: first, mapping the company relationship network and second, understand a group of common external factors that could directly affect companies. With this knowledge in mind, we can discover investment opportunities in areas that might be ignored by the majority in a timely fashion.