A Venture Capitalist's Personal Stock Investments

(Boyu Hu, May 2016)

I became a venture capitalist 6 years ago after spending the previous 10 years as an engineer. I’ve been quite lucky to have the opportunity to learn closely from some great investors, such as Allen Zhu from GSR Ventures, Hurst Lin and Ruby Lu from DCM, and Lei Zhang from Hillhouse Capital. The aggregate valuation of my 10 portfolio companies grew 21 times in an average of 3 years, and my personal stock portfolio also grew 5 times in value.

The track record looks good. I must say a lot of it is luck, but probably not all of it - in this blog post I will try to identify some scientific elements behind my luckiness, and talk about the lessons I learned about thinking habits as I reviewed my investment decisions. I have to go through some theories, which could be dry and counterintuitive. Hence I was showing off some performance data above, not to brag but to keep your curiosity in reading on.

I will use my stock investments as examples due to confidentiality obligations in private investments. The chart below shows the performance of my personal stock investments. The red line is “cumulative net gain” as a multiple of original principal, the blue line is “balance of principal”, and the green line is S&P 500 (as a benchmark).

You can see that when my net gain was just about to surpass principal, the account experienced a “redemption” as large as 1x principal, followed by two other “large redemptions” that cumulatively represented over 3x the original principal. Under such tough situation, however, my net gain still reached 4x the original principal.

Here are a few more facts just to clear suspicion for data manipulation:

This is my only active stock account.
I didn’t used any insider-information.
I made all investment decisions independently.
I didn’t get lucky by holding one small-cap stock. For most of the time, I held 4 to 5 large-cap stocks almost evenly.
The chart covered most of my stock investment history. The earliest report I could export from Interactive Brokers is for 2012. I had a small loss of 20% before 2012.

How did I do it? My basic logic for stock investment is almost exactly the same as for venture investments: buying shares of great companies when they are under-valued and holding them for a very long time. A great company should have a) huge market potential, b) significant entry barriers that can grow over time, and c) a great management team.

At this point you might want to argue that this is no secret and that I might just have been very lucky. Indeed, again, a lot of it is luck. However, it’s not a good habit to attribute something to luck and stop thinking when we see things we don’t understand. The right way to do it is to “assume something is out there and observe carefully”.

So I tried to see it from a different angle. I was thinking, the stock market is actually a “zero-sum game” if we exclude profits coming from dividends and market growth. Someone has to lose money. Since I had no secret weapon, is it possible that others did something wrong with their decision making

I did some study on my “counter parties”. I found three types:

Retail investor, approximately 1/4 of the market cap in the US,
Passively managed funds, 1/2 of the market cap, and
Actively managed funds, 1/4 of the market cap.

The passively managed funds are followers. Retail investors are not only followers but also too fragmented. My real counter parties, the ones who are actually pricing the stocks, are actively managed funds, or, in other words, fund managers with the power to allocate lots of money.

After observing my counter parties, I developed a few theories of how to beat them in a scientific way. Now I will talk about these theories and some lessons I have learned from my own mistakes. Some of the concepts discussed below are explained well in the book “Thinking, Fast and Slow”. If you have read it, you’ll understand that the “human instinct” and “thinking hard” I’ll talk about are actually “System 1” and “System 2” as defined by the author Daniel Kahneman

Theory 1 - the world has changed and different skills are needed to search for and understand the next-gen great companies.

Warren Buffet is one of the best active fund managers in history. He is well-known for saving his own grace by staying away from high-tech companies in the dot com bubble and following his principle of not investing in companies that he doesn’t understand. There are two very “old” high-tech companies in his latest portfolio – IBM and AT&T – and he lost 15% on these two investments. With that said, Warren Buffet is still one of the world’s top investors. The companies he has been holding for 30+ years, such as American Express (he invested 50 years ago), Coca-Cola, Wells Fargo, and Gillette (P&G), are still great companies today.

However, many other great companies emerged in the last twenty years. Now, Apple, Alphabet (Google), and Microsoft are the three largest companies in the world (by market-cap). Facebook and Amazon also entered top 10. By market-cap, these tech companies are 55.2% of the top-10 companies. Admittedly, there could be some tech-related bubble in today’s market-caps. But if you take a deeper look you will find that the combined NET PROFIT of tech companies as a percentage of top-10 is even higher, at56.2%. This is no longer a bubble.

The last high-tech bubble and the depression following the bubble burst made fund managers realize that “user needs are not real” and “technology barriers are not long-term barriers”. Then they stopped thinking. However, the Internet never stopped changing human behavior, but just a little more slowly than expected. Moreover, high-tech companies have since built layers and layers of new barriers on top of their technology barriers, for example, the eco-system built by Microsoft, the network-effects Facebook enjoys, and the user-generated-data and Android eco-system built by Google. To understand these new barriers, you need to deeply understand people’s needs and behavior in the cyber world. You need certain new skills that most “fund managers” don’t already have

I have a bold theory –fund managers are currently allocating most of the active money in the market, but the way they analyze the world have not adapted to the great changes happening in the world

Let’s take a look at how they think about industries and companies. If you observe closely you’ll find that they rely heavily on an “information eco-system” consisting of market research institutions, corporate finance departments, sell-side analysts, and buy-side analysts. These parties typically collaborate like this:

Corporate finance departments release financial data and forecasts, using industry data and forecasts from market research institutions
Sell-side analysts issue equity research reports using public information, build financial models and issue target prices
Buy-side analysts produce research reports and give investment recommendations for internal use.
Fund managers make investment decisions.

However, a lot of things have changed in the post-internet age – connectivity infrastructure, how people get information, and how they make spending decisions, etc. As a result, the best ways to understand industries and companies have also changed.

I’ll give you an example. In May 2013, Consumer Reports issued a report on Tesla’s Model S. On the next day, I decided to do some study on this company, so I did two things. First, I visited Tesla’s store and tested the car in Menlo Park (I was traveling in US at the time). Then I spent a few hours online and reviewed all the user comments and professional reviews on Youtube, Twitter and other websites.

Then I had two conclusions – First, Model S is a good car. Demand will be way over supply for quite some time. There’s chance that it’ll become a mainstream car in US. Second, Tesla Motors’ market cap is still very small compared with the potential size of its addressable market. Therefore near-term stock price will be driven by how much attention the company gets. Using my “common sense” on media behavior and word-of-mouth, I decided that the public attention to Tesla would continue to grow for months

Then I immediately bought Tesla’s stock (at around 80 USD). The stock price grew 3x in less than a year. (My investments are usually long-term and this one is not a typical case. I sold it at around 210 USD when I believed the “social attention” had peaked, yet at the same time I didn’t feel I understood manufacturing and battery industries enough to have the conviction to hold the stock for a very long time.)

In the pre-internet age, it is almost impossible for retail investors to study a company’s product and user feedback in such ways. It’s also impossible for them to estimate changes of “social attention” on a company. But they can do these things in high quality now, and sometimes even better and faster than the analysts in big funds. This is just one example. I use different approaches to understand different companies. I have a protocol – never touch a stock unless I feel that I understand the company better than most buy-side analysts in big funds

Fund managers’ advantages in understanding industries are also diminishing. I often say to my colleagues that we venture capitalists are on the top of Mount Everest of innovation. I meant two things.

We spend most of our time looking for opportunities in structural changes led by technology innovation and behavior changes of consumers and organizations.

We are surrounded by a group of great entrepreneurs. They are probably the smartest and hardest-working people in the world, who never stop searching for disruptive opportunities and oversights of large companies. They will develop innovative products and test them in real life, and then feed the information - user behavior and data - to us venture capitalists. The quality of such information is extremely high. It’s like a window that allows us to peek into the future and understand the real world better.

Now, some of you might be wondering if most venture capitalists can make good money in stock market. I used to think the answer is yes. But after talking with some friends in the industry, I found that it’s not true – most of them had mediocre returns in public markets (unless they didn’t tell me the truth). Why? My current thesis is that, there are many pitfalls in stock investing, and most of us will be trapped by one or another, even though we have access to very high quality information. Next, I’ll talk about these pitfalls.

Theory 2 – human instinct is not good at excluding interference of emotions when making decisions

Facebook’s IPO price was $38. A lot of venture capitalists bought it. It fell under the water next day, and then dropped to the lowest point, $17.7.

Please take a few seconds to imagine how someone would feel if he/she bought the stock at IPO. The stock stayed at the bottom, then rebounded a little, and then dropped again. A lot of my friends felt relieved when the stock price reached their cost, and sold it.

The below chart shows Facebook’s stock price in the following three years.

Why did they feel “relieved” and sell the stock? The fundamental reason is that they were affected by an irrational emotion – being “under the water” gave them a mild depression, and this feeling pushed them to sell as soon as possible. But in the long run, stock price of a company is only related to its business outlook and has no logical co-relation with emotions.

I also bought some Facebook stocks at 38 USD at IPO, then I bought more when the price dropped and a lot more at 19 USD, then I held it for a very long time. I sold some when I felt the potential threat from WhatsApp, but then bought it back when Facebook acquired it. Here’s my rationale:

https://www.zhihu.com/question/22105802 (in Chinese)

I am not immune from emotions when the stock prices go up and down, but I try to make investment decisions solely based on independent observations and independent judgment on companies’ long-term profitability. When a stock price changes significantly or when I feel the impulse to trade, I will have a reflection to see if there is information I don’t already have, and review the business fundamentals (e.g. user growth, stickiness, monetization potential). If my judgment on the fundamentals remains the same and that I believe the stock price dropped only because of fluctuation of market sentiment, I will feel happy and continue to buy.

Another emotion that often clouds judgment is greed. A well-known angel investor asked me to recommend a stock. I said I have been buying XRS since 23 USD. He said he liked the company a lot, and then told me why. I then asked him if he bought it. He said he didn’t because the stock price kept rising and never corrected to give him a buying opportunity. This greed kept him irrationally waiting to take advantage of short-term fluctuations, and miss the real opportunity.

In my view of the world, rational decisions are the most efficient. Whenever emotions get in the way of decision making, the result will be compromised. However, it’s been proven again and again that human instinct is prone to mixing emotions with decision making. What’s more, humans are very “lazy” thinkers. Next, we’ll talk about how so.

Theory 3: to understand the real world you need to “think hard”, but human instinct is “lazy”

Let me share a mistake I made. A friend recommended stock QIWI to me. It’s a Russian company listed in the US. It’s the “PayPal of Russia”. I found some research reports on the company, which suggested that all of its metrics were far ahead of its closest competitor and the gaps were widening - transaction volume, market share, active users. Moreover, it owns 170 thousand offline terminals (like ATMs and POS machines), another great barrier. The stock price was 22 USD with a market cap of about 1 billion USD, and P/E ratio was only a little more than 10. I thought it was significantly undervalued, so I bought the stock. In the next 3-4 months, the stock price went up 50%. I was quite happy and thought I found another “10x stock”.

Shortly after, QIWI started to drop. I did some research and found that it’s because oil price and the Russian Rupee dropped. I thought the currency should rebound because exchange rate is pegged to PPP (purchasing power parity) in the long-term and changing of PPP is slow and painful. So I continued to buy, and Rupee rebounded as I expected. I felt happy. But shortly after, the stock started dropping again sharply and didn’t stop this time. I was worried so I did more research but found nothing. I didn’t figure out the real reason until quite a while later – 1) some of the 170k terminals that QIWI owned were to be shut down due to regulation change, and 2) young people in Russia were switching from using QIWI wallet to using bank cards (debit/credit).

Once I got the two new pieces of information, I quickly sold all of my holdings. In the end all the prior profits were erased, along with 1/3 of my principal on the stock. QIWI might still be a great company, but I didn’t know enough about it and I didn’t have the time to understand it. To be honest, I still don’t understand why these two issues happened to QIWI. Looking back, I almost knew nothing about Russia except some general data about population, e-commerce penetration, etc. But why did I feel that I understood the company when I bought and held the stock? What gave me the false sense of security?

The simple answer is that human instinct is lazy and my instinct had me fooled. Below are a few examples of how humans act lazy based on our instincts when we need to think hard.

a) Human instinct often relies on indirect information and refuses to observe and think independently.

Formality often gives people a false sense of security. The research reports I found contained a lot of nice charts, abundant data, detailed analysis, and pointed out many risk factors in a very “balanced” way. Also, they are all very “thick”. All these beautiful, perfect formalities fooled me into believing the reports’ conclusions directly and completely forgetting an important code of conduct I had for investment decision making – that is, before touching a company’s stock, I need to have confidence that I understand it better than fund managers. The conclusions of the research reports are indirect information, and so is their formality.

Since birth, we have been taught all sorts of rules and knowledge. This saved us a lot of time, because it’s more efficient to gain experience through learning from others than observing and inducing by ourselves. However, this convenience also tricks us into the habit of relying on indirect information to understand the world. For example, we often decide whether to trust someone by looking at how he/she dresses, speaks, and looks. The location and look of a bank or a company’s office often determines our interest level in doing business with it. We might get used to air pollution when we see a lot of people walking on the street unprotected. For the same reason, we might think that “this must be right as so many people think so”, and “this must be right because people have been doing it this way for so many years”.

At the beginning of this blog post, I mentioned the names of a few “great investors”. I did it purposely to let you feel that “these people are great investors so this author must be good too, and what he says must be right.” This strategy to offer indirect information usually works, but it’s not logically sound – the people I cited are great doesn’t necessarily mean I am good too, or what I say is right.

b) All forms of communication are inefficient; human instinct tends to overestimate the efficiency of language.

I later found that in the “Risk Factors” section of QIWI’s 2013 annual report, the two risk factors were disclosed, but I completely overlooked them. It’s not that I didn’t read the annual report, but that it’s impossible for me to digest the information – I wouldn’t understand what it means even if I read it, because every prospectus and annual report contains risk disclosures of change in regulations and user behaviour, just like in my case.

In fact, human language is quite inefficient in passing on information. When a person translates facts into language, he loses a lot of information and adds a lot of his own understandings. When another person translations language back into facts, he again loses a lot of information and adds a lot of his imaginations. The problem is, we are always over optimistic about language’s ability. A lot of rules and systems designed to govern the modern society and companies are based on the assumption that “language is perfect in passing on information”.

c) People make decisions based on a projection of the real world in their brains, and most parts of that projection are imaginary. However, human instinct can’t distinguish imagination from facts

I have lived in China, US, Europe and Singapore, but never lived in Russia. When I did my research on QIWI and saw numbers related to population, internet, e-commerce, my brain was automatically filling in a lot of life experiences to build a complete “Russian model”. These life experiences mostly came from my observations on consumers and companies in other parts of the world. However, I couldn’t tell which parts of the “Russian model” came from my imagination and which parts are real facts. (By the way, human memory works in a similar way. You cannot tell which part of your memory is real memory and which part is made up by your brain using other memories.

d) Human instinct tends to be over confident.

When QIWI’s stock price went up, I thought I was right about the company, when it fell, I looked for evidence of it being misjudged by the market. With this mindset, for a very long time, I selectively neglected the correlation between the exchange rate and the stock price, not to mention paying attention to the change of consumer behavior. It was not until I reviewed my mistakes that I found out that the first time the stock rose was due to change in the exchange rate too.

Research has shown that often times we human beings can only see facts we want to see while selectively neglecting facts contradicting our wish. However, it is easy for us to know what we know and very hard for us to know what we don’t know.

Human instinct is lazy in many other ways: for example, it’s often affected by “anchor effect”, and it often struggles to think statistically, etc. Many of these problems are well explained in “Thinking, Fast and Slow”.

To overcome the laziness of our instinct, we should make it a habit to always “think hard”, to see the world with a suspicious eye, to admit and accept our ignorance and mistakes, and to remind ourselves not to be affected by irrational emotions and irrelevant motivations when we make decisions.

Some of you might be discouraged to see the all those flaws of human instinct – maybe it’s better to give my money to fund managers? After all, they are institutions. Decisions made by a group of people must be more fact-driven and rational than by a single person. However, this could be a mistake. Next, we’ll talk about decision making in group setting.

Theory 4: group decision tends to rely on human instinct, and human instinct tends to rely on group decision

I found that, when “thinking hard” is needed, it is tougher for a group of people to make objective, rational decisions. The more people in the group, the harder it is. Why?

Let’s play a mind game first. Please imagine a scene where a five-person family in Beijing is making a decision to purchase an apartment. After several months’ visiting and comparing, they are now choosing between two candidates of similar total prices: apartment A: good location, but smaller and older; apartment B: newer, larger, but location is not good.

Father likes A for its better location, performance-price ratio and potential for appreciation in value. Mother likes B for the quality of neighborhood and comfort. Since they disagree, the mother proposes a vote among the whole family. Grandma looks at the price and says, A’s price per square feet is too high – I vote for B. Grandpa says I am fine with both. Mother says you can’t quit and must make a choice. Then grandpa says I’ll choose A as it is closer to my Mahjong friends. Another tie. So the father asks their only child, who is Grade 1, “last weekend we went to see two apartments - which one do you like?” The child responds “I don’t want to move.” The father insists, “Sorry, our apartment is too small. We have to move. Which one do you like better, A, or B?” The child blinks his eyes and mutters “B”.

3 v.s. 2. They end up with apartment B. Soon after they move in, grandma regrets. She has to take a one-hour bus ride to escort her grandson to school in the morning and then another hour by bus to come back. In the afternoon, another round-trip to take the kid home. 4 to 5 hours on the bus per day. The kid also feels unhappy, so he requests to move back. Then father says, “B was YOUR choice – why did you choose B?” The child answers, crying, “when we came to see the apartment, there was a cute dog around, but it never turned up since we moved here!”

A group of people making decision is actually quite similar:

The “number” of people for or against an idea and the “number” of positives or negatives usually define the resolution. But oftentimes the weighting of one or two key factors is much higher than all other factors combined, and in a group setting, most people cannot think in this way.
It takes time to “think hard”, but it is rude to pause in front of a group, forcing one to think on his feet using instinct. Moreover, some of us can only think effectively when they are alone. They start to feel a little nervous in front of another person, and might become a complete idiot in front of two or more people, especially when one or more of the participants are senior. Others may think much slower in conference calls. What’s even worse is that very few are aware of this. I don’t know how many people are like this among the population, but I myself am one. When I don’t feel I understand a subject enough, I tend to keep silent in group discussions or conference calls, and sometimes this gets me misunderstood.
Encouraging everyone to “contribute” is a well-accepted team culture. But it’s flawed. Usually, among a group of people, if a subject like an industry or company that needs “thinking hard” to understand, only a minority have deep observations, and the majority only knows “some” or “none”. However, they either tend to be unaware of the fact that they don’t understand it, or try to pretend to understand it even if they don’t. However, everyone’s opinion has the same weighting.
There is no universal definition for “good company”, “overvalued”, or “under-valued”. If the benchmark everyone uses to express their opinions are different, an inclusive “group opinion” is meaningless.
Some observations may sound like “facts”, but in fact everyone has a different definition for the perceived “fact”: for example, “the product is liked by users” and “the management team is great”. When someone tells an observation and opinion, others hear different things.
“Neutralists with no stance” always sound more rational and correct.
Most people just follow others.
It’s rude to express an opposite opinion directly.

In summary, a group of people tend to rely on human instinct to make decisions, and human instinct is lazy.

A counter argument is that a lot of mutual funds and hedge funds don’t make investment decisions by voting, but by the fund managers.

Yes, that’s true. But everyone is constrained by certain “force fields” and each “force field” entails group decision making to some extent. Let me give you a few examples.

“Reflexity Force Field”. The stock market itself is a giant voting machine. When most people believe in something, it will happen. So each participant is watching and predicting others, and at the same time, affecting others with his voting power (buy or sell). This mindset is reasonable in the short-term but completely irrational in the long-term. A company’s long-term value depends on the characteristics of its industry and the company’s business, and has almost zero logical co-relation with the number of people believing its stock price will go up or down. It’s very difficult to fully avoid this “force field”. Neither can I.

“Investors’ Handcuff Force Field”. I once had a chat with a marketable fund manager managing a 3 billion USD portfolio. He said, “I sometimes envy you venture capitalists. The money you are managing is so long-term. The 3 billion looks great, but investors will start to redeem when you under-perform for just a quarter. I have served some clients for more than a decade and helped grow their assets several times, but they almost never hesitate to redeem when they feel your best time is gone. I’m 50 years old, but I have to act with utmost caution every day. I have to be on constant watch for companies’ next quarterly reports and get ready to sell at any sign of market movement, be it substantiated or not. . I can’t simply hold a company even if I understand its long-term value.

“Professional Manager’s Force Field”. Some fund managers look very powerful, but actually they have their own boss(es), such as “Chief Investment Officer” and “Chief Risk Officer”. The bigger the fund, the more “officers” it has, and the more complicated the investment process is. Fund managers have to consider a lot of non-return oriented factors when they make investment decisions.

“Winner’s Curse Force Field”. If a fund always under-perform it will die. But if a fund always out-perform, it will get bigger. When it gets bigger, it has to invest in more companies, so the “big fund manager” will have to hire more “small fund managers” and analysts to study the companies, and the “big fund manager” will have to rely on “research reports” to understand the world. While he is smart enough not to take the reports for granted, he wouldn’t have capacity to do anything on this own. Then he gets a little nervous. Then he starts to comfort himself by asking others’ opinions, or having more group discussions…

“Culture and Human Nature Force Field”. We are in this force field since the day we were born. It’s human nature to pursue power and express oneself. Even if you completely agree with me that group decision making is not efficient, when someone in your “council” or “committee” claims dictatorship, you will still vote him down with no hesitation. In a tribe, this man might be killed. Similarly, whenever a fund adopts a culture (or even a system) of group based decision making, it can never change back.

Having read the above you might feel a little hopeless – it seems human beings are doomed for failure. A single person is inefficient in decision making because human instinct is lazy and not good at excluding interference of emotions when necessary. A group of people is also inefficient in decision making, because group decisions tend to rely on human instinct, and human instinct tends to rely on group decisions.

What can we do? How did Blue Lake Capital solve these problems? To be honest, we are still experimenting and thinking. On the one hand, we set a limit on the size of our investment team (currently 6 persons) to minimize communication cost. On the other hand, we encourage a lot of one-on-one “offline” communications. Looking back, of the 16 deals we have done, only one was approved in our group weekly meeting and the others were all approved “offline”.

I didn’t plan to write so much when I started. Hope talking about these can give you some inspirations. Hope it also gives leaders of the society and companies some ideas about structuring organizations. Writing about this helps me as well to reflect on my experience and stay curious. I enjoying honing my thinking habit, and often use “external brains” to help myself think more rationally and objectively. By the way, if you have a habit to “think hard”, you are welcome to contact me. I’ll be happy to get to know you and maybe we could serve as each other’s “external brain”.

As a tradition of my blogs, here are some advice for entrepreneurs:

Develop a habit of self-reflection and practice. Try to “think hard” and don’t rely on experience or instinct. Make “fighting instinct” an instinct.
Don’t trust any internal and external reports without doing your own work. Use data and your personal experience to understand your customers and employees.
When you design an organization, try to avoid settings that introduce group decision making, and try to eliminate the influence of irrational emotions and incentives at key management positions.
Have faith in young people, and give them responsibility when you find ones with good thinking habits.