Transcending Hobbyism (Angel Investing 4)

A company I really liked was out raising money. They went to a very well known early stage VC and had a great first meeting. A couple of days later the partner got back to them with a no. He said "We really like you and your company, but we have a portfolio construction problem." The entrepreneur called me. "What the hell," he asked, "does that mean?"

As an investor you don't just pick the good ones and walk away from the bad ones. Were it so easy. You are managing a finite pool of money and are looking for the best return on the pool, not just on each individual investment. That means that sometimes you have to walk away from a good one. The easiest way to separate the serious angel investors from the hobbyists is to ask them about their approach to building a portfolio. Serious angels have a plan for what they are trying to build. Hobbyists are just piling up stones.

Here are some questions. They're analogous to the questions you ask entrepreneurs when you're trying to figure out if they've thought their plan through: How much money do you need? What will your burn be? How long until you're profitable? Startup venture funds are not like the startup ventures they fund. But, like startups, they need to have a viable business model. And, like startups, that includes not running out of money halfway through.

First, how much money can you invest in startups?

I'm not going to give you advice on this because I'd be a hypocrite. I long ago blew through whatever reasonable percentage of my cash I should have put into this incredibly risky, close-to-zero liquidity asset class. Do what someone else says, not what I do. Most commenters say you should only invest what you are comfortable losing. That's solid advice*. Big institutional pools of capital often reserve 2-10% of their assets for private equity, of which venture capital is one fifth or less. I'll leave it at that.

Second, how many investments do you plan to make?; third, how much work do you plan to do with each company post-investment?; and fourth, will you have an investment thesis?

We all know it's the right thing to do to invest in a portfolio of companies to avoid idiosyncratic risk--the kind of risk that is specific to a single investment. If you have a large enough portfolio--common wisdom is some 20-25 companies--of uncorrelated investments then you have pretty much minimized the unsystematic risk.

Related is the idea that since venture investing is a hit-based business, you need to have a lot of investments to have a good chance that one is a hit and returns your entire pool of capital several times over***. The number of investments to make it probable for this to happen has been estimated at between 20 and 150.

These numbers don't take a few things into account, and you should.

Getting rid of diversifiable risk makes sense: the risk that a portfolio CTO quits after six months, that Google unexpectedly launches the exact same product, that your platform decides to ban the way you make money. Sometimes you should know this is going to happen before you invest (I'll talk about due diligence in another post) but sometimes there's no way to know. Having many companies means that if a couple hit one of these deadends, you're not finished. But keep in mind that diversifiable risk presupposes a lack of correlation between investments. The risk that a CTO leaves one company is uncorrelated with the risk that the CTO of another company also leaves. On the other end of the spectrum, the risk that there's a financial crisis and every venture investor simultaneously decides to stop doing Series As is 100% correlated across all seed-stage companies: it's a market risk and you can't get rid of it through diversification.

But in the middle are the the kind of semi-correlated risks that are the bread and butter of every good venture investor, the risks that go with a particular sub-industry. The risk of Netscape changing third party cookie policy from opt-out to opt-in is correlated across most ad-tech companies but not many other internet companies. There are similar risks to any investor's thesis, be it Big Data, Social, Consumer Internet, whatever. But trying to entirely diversify away these risks means you can't really have any thesis at all. And that means you can't have any expertise. The market risk in finance is called β and is the best you can do in an efficient market--one where all information is available to all participants. But if you feel like you know something that not everyone knows, that you're an expert at something, then you can generate α, return in excess of risk taken. β is what asset allocators brag about. α is what investors brag about. We like α.

There are two strategies: create an index-fund like portfolio with 100+ relatively uncorrelated investments and try to get β, or create a 20+ portfolio of companies you actually know something about--less diversification but more α. The latter strategy is riskier: if what you believe about the market turns out to be untrue, that whole part of the portfolio goes bad. But if it's right, you do better than the market.

There's also, of course, an impact on how much money you spend and how you spend your time. Many founders don't want investors who are putting less than $25k or $50k into the company. You may find that you don't have enough money to get to a fully-diversified portfolio. You should think about this carefully. If you're not somewhat diversified, you are handicapping yourself relative to other investors and you face a pretty decent probability of actually losing all your money****.

And then there's how you spend your time. If you hope to help the companies you invest in grow by being active, you can't make 100 investments every few years unless you have some infrastructure to help you.

Fifth, what stage will you invest at?; sixth, swing for the fences or go for earners?

I assume that most angels invest at the seed stage. It's the only time when you can get an appreciable piece of the company for an angel-sized stake. But there's seed and there's seed. If you put in first money like I try to do, when it's two people and a pony, you're taking more risk and will need to hold the investment longer before exit. But you get to invest in the best people and the best ideas. If you invest after there's a minimum viable product and the company is looking for some money to start getting customers, then you have less risk--you can assess the strength of the team as a team and look at the MVP as a product and talk to potential customers about how burning their need for it is--but by then you're probably competing with many other investors to get a piece of the deal.

In my experience, if you invest at the very start of a company, you will end up holding that investment for some six to eight years. Some companies exit sooner (although it's rare for one to exit less than three years after inception) and some take longer (I was recently talking to the founder of a company I invested in thirteen years ago--and it was two years old when I invested--about starting to look for a buyer.) In any case, you should expect a substantial period of illiquidity.

The type of company you're investing in also makes a difference to holding period. The companies that are swinging for the fences usually take longer to mature. Simple, for instance, whose vision is to disrupt retail banking--one of the largest industries in the world--by actually treating their customers like customers took three years from idea to launch. Less ambitious companies can start, launch a product, and sell in less time.

Some investors prefer the less ambitious companies, the earners, hoping to build and flip in a short period of time. I don't. I tend to follow the old adage of shooting for the moon and landing on the roof. If you're trying to build a billion dollar company and fail, you might end up with a $50 million exit. If you're trying to build  $10 million company and fail, you end up with nothing. But that's personal style; there are prominent counterexamples.

Seventh, will you follow on?

Related is whether you follow-on or not. That is, having invested in the seed round, do you also invest in future rounds?***** I'll talk about whether this is a good idea or not in a later post, but if you decide you want to be able to follow on, you need to reserve money for it.

In my portfolio, the average amount raised has been about

First round	$1.3 million
Second round	$4.7 million
Third round	$7.7 million

If you invested $50,000 in the first round at a $5 million post-money you would own 1%. Then you could expect that your pro-rata of the second round would be $47,000. Your pro-rata of the third round would be $77,000. Your $50,000 is now close to $180,000. I usually stop investing after the B. I reserve twice my original investment for follow-ons, assuming that some companies won't raise the A or the B.

Putting it Together

I ask my founders to build a financial model, even though it's inevitably wrong--"prediction is hard, especially about the future." But the act of building it means they need to think through tradeoffs, milestones, capital needed, etc. While the decisions they make because of this thought experiment should evolve as actual facts come in, the exercise of modeling allows them to know what their envelope of viability is. Answering the above questions should allow you to do the same.


NB: The undisputed master explainer of these types of issues is Roger Ehrenberg on his Information Arbitrage blog. His post on Portfolio Construction goes way more into depth than I have. Also read at least The Right Fund for the Mission.

Next: Modeling your fund [edit: not done yet, skip to The Signal and the Noise for now.]

Previous posts in this series

1. Intro: Why I'm Not an Angel
2. How to spend your time: The Work-Work Balance 
3. Positioning: How to be Different When What you Sell is a Commodity

-----
* Long ago a girlfriend brought me to a gathering of her friends. We joined a couple of other couples playing penny-ante poker. After I goaded the others into betting BIG (literally HUNDREDS of pennies) the gf walked away from the game. She chided me afterward: "Why were you being such a dick? We were just trying to have fun." I didn't get it. If it isn't painful when you lose, why even play?**

** Also, why I don't blog relationship advice.

*** The difference here is a bit confusing. Diversifying away unsystematic risk means that you actually increase the expected value of your portfolio for any given amount of risk. Making lots of bets in a hits-based business doesn't increase your expected value, it decreases the variability of your outcome.

Think about it this way. There are ten upside-down bowls. Under one of them is a $20 bill. You can 'buy' any bowl for $1. Since each bowl has a one in ten chance of having $20, your expected value of each bowl is $2. The smart thing to do is to buy all the bowls, spending $10 to get $20. Your expected value does not change (it's still 2:1), but the variability of your outcome has gone to zero: you always get the payout.

Of course, if you bet on a single bowl and won, you'd get a 20:1 payout. It's because you know nothing about the bowls that you're better off buying them all. If you have an inkling about which bowl is more likely to have the money under it then you're better off just buying that one. The idea that you should make 150 investments to lock in the average angel investing return presupposes that you have no idea which companies are better than the others. I've never met an investor who truly believed this. I consider myself pretty damn humbled by experience, but even I'm not that humble.

**** The astute reader will observe that if you follow the rule from question 1 and only invest what you're comfortable losing, you may put yourself in a position where you're more likely to lose it. You then have the choice of one of two mantras: 1) "The rich get richer, not me", or 2) "I am large, I contain multitudes."

***** If you decide you want to be able to follow-on, you need to negotiate pro-rata rights into your deal. I'll talk about this and other deal terms in a later post.