Regarding structural VC investments

I'm a very structural person. I was born that way and despite the rather suffocating educational environment in Taiwan, I managed to build my own thought system along the way. I've used the same principles to govern my insatious appetite for deep knowledges in every sector that interests me: classical music, literature, philosophy, economics, history, art and all of my past and current professional realms.

I brought the same mentality to the MBA program which I did in a part-time fashion, slashing through all courses in a systematic way that earned me high recognition for even the non-hardcore courses such as Marketing or High Fashion Brand Management.

For the most hardcore ones I went full length. For example, not only did I complete all finance classes at ultra high marks, I passed the CFA Level 1 and 2 exams in 6 months alongside my day job and the on-going MBA program — I am now prepping for Level 3 in June which, if passed, would put me at finishing all 3 levels of exam in 18 months, the shortest possible duration.

It didn't stop there. Upon the 2nd opportunity to serve as a judge for the internal PE competition at HEC Paris. I decided to design an LBO case on a large public firm using all public information. The case won overwhelming praises from the finance professors involved in the competition as well as the MBA students tortured during the 72 hours. So much so that I've worked with a prof. to improve it to be used as an official case for the MBA program. And I'm invited by the prof. to come to the Structured Finance class to teach the case.

All the above is not about bragging, but just to show how structural and systematic I am as a human being so that we can continue this article on a common ground.

When I started out my (still-young) VC career, I was keen on applying the same structural mentality to this new-found passion. However, somewhere in the back of my mind there has always been a question: structural methods are best applied on things that are sort of already known, whether it's scientific research, public market equity research or strategy consulting. On the other hand, startups are supposed to be about things that nobody knows for sure yet. Wouldn't structural approach contradict such nature?

As I joined board meetings of startups and heard pitches from entrepreneurs of all sorts, I quickly realized how structural methods could be a big liability in this business.

But first let's clarify one thing: there are still a lot of structures in VC investments. The fund itself has more regulation rules than most people care to know. The managing firms have a certain structure that's best aligned with such investments. The contracts between the fund and the invested startups take up certain structures to reduce informational asymmetry and align long-term interests. And although the valuation could seem to come out of thin air due to the high uncertainty regarding the future 10-year down the road of a young startup, proper understanding of all valuation theories, such as how discount cash flow is linked with multiples, are still mandatory for a VC.

However, the fact that "the venture capital business is a 100% game of outliers" implies that the success of venture investments really hinges on a very unstructural part — how to find that outlier that has the chance to change the world and help it achieve that.

In other words, deal sourcing.

A VC can be equipped with wide networks as well as deep sectorial knowledges. He or she can be super wise and know how to manage founders' egos. However, if he or she could not get access to the startups with potential, he or she wouldn't even be able to make use of those networks and knowledges.

While all the other elements of a VC firm could more or less be constructed in a systematic way, access to deals remains highly illusive to many.

That's not to say there's no existing structures in terms of deal sourcing. An inherent structure in deal sourcing lies in the traditional funnel that I briefly reminded readers of this blog last year. A good VC firm fields about 5,000~6,000 inbound pitches per year. Young associates apply criteria defined by partners to help narrow those into 600~800 for review. Then maybe 100 are invited to come into the office to pitch. And maybe 0~2 investments would be made. The structure here lies in having pre-defined criteria for screening, such as sectors, KPIs and objective backgrounds of the co-founders.

At the same time, most VCs will tell you that the best deals they've got seldom came through this traditional funnel which ironically consumes so much of the firm's day-to-day resource. Instead, the best deals usually came from referrals or scouting by VCs themselves.

Then there's Y Combinator that enjoys a fundamental symbiosis with the lean startup movement. It functions in an entirely structured way with its batches, programs, mentors, demo days, etc. The current CEO of YC, Sam Altman, even runs the famous CS183b class at Stanford, which gathers some of the best entrepreneurs and VCs to teach young students on How to Start a Startup. By the way the on-line archive for this class is a wonderful resource for those who want to get up to speed quickly — for free no less.

The success of YC seems to be validated in the three following ways:

  • The total market cap of Y Combinator companies is reported by Sam to be over $30 billion, with portfolio companies having raised over $3 billion
  • The average Series A round size of YC graduates trends up year after year
  • The fact that tens of thousands of accelerators have emerged around the world in direct copy of the YC model and that terms like "demo day" have become household phrases

With all this success YC attracts more and more young entrepreneurs scrambling to get into its programs, while VCs join it on Demo Days to look for deals. All of this seems to suggest great success based on a highly structural model — until one is reminded that none of the YC companies has gone IPO yet.

For those of you that've been to YC Demo Days or any similar events around the world, you might have noticed how the startups look and sound very much alike — not in their ideas, but in the way they communicate their products, visions and tractions. They are in fact applying a common methodology that's designed to push all the buttons of VCs of this lean era. As a result, everything seems to be more and more mechanical.

It's not an overstatement to say that YC today resembles a machine that produces well-rounded startups, apt at raising funds to grow their businesses. However, will the real homeruns (exits via big IPOs) really come out of this system consistently? Or even more, what would happen after YC ringleaders such as Airbnb finally IPOs? What's down the pipeline?

500 Startups are another innovative example of structural VC investments. Dave McClure successfully created a new VC investment model that's based more on KPIs and less on entrepreneurs. The sheer number in investments (500!) and the properly selected geographies are supposed to smooth out the granularity due to the lack of human approach. However, even Dave recently complained about entrepreneurs getting out too early:

I don't think that the problem lies with entrepreneurs wanting to get rich fast, but rather that the approach 500 Startups pioneered naturally attracts "bean-counter" — I am trying not to use the term "rent-seeking" — type of entrepreneurs who try to optimize their own perceived risk-return relationships rather than go headstrong for a distant but potentially much more rewarding future, for both the founders and the investors, both financially and psychologically.

I guess my point is: if the VC investment is 100% a game of outliers, then trying to find such outliers using a structure, however flexible and loose it is, seems to go against such nature.

I hope I'm wrong and I wish the best outcome for the VCs that use more structural approaches for deal sourcing and their startups. On my side, as a super structural person I've been working very hard to "destructure myself" when it comes to deal sourcing.

Especially given the nature of hardware investments — we have to see the working prototype with our own eyes in addition to meeting the team. My partners and I travel efficiently to every potential city to seek out the outliers and evaluate the potentials, instead of sitting in the office and waiting for pitches. We spend time with the team to learn more about them. We try to pick up signals whether the founders might be too stubborn for their own good or too compliant for our own good. We make friends with those that we come to respect and like even if they don't need our investments. At the Hardware Club everything starts with our COMMUNITY and the fantastic entrepreneurs that we invite to join us.

This is obviously very tiresome and another VC might prefer the traditional funnel way so that he or she could enjoy full access to the air conditioning office and the same food delivery services — of which he or she might even has a stake — everyday. However, my partners and I see no other way in our hunt for the next hardware Unicorn.



Risk and uncertainty — a tale of a VC

The math behind jaw-dropping mega-round valuations