Tuesday, 9 December 2014

What is Fishburner's? Are 500 Startups better than 5?

Lately, I've taken the view with tech companies and for that matter startups in general that you need to see a lot of them. That means when I get an invite to an invent I rarely say no. Last Thursday I have to admit that I went along to Sydney incubator "Fishburners" with little more than the title of the event to describe what was on offer. Having said that I was glad I did.

Fishburners describes itself as the "largest tech co-working space in Australia." There's over 100 startups paying up to $300 a month, less if they're hot desking, for space within the facility. For that, you're encouraged to share ideas and connections with other entities. Fishburners is a not for profit, but that doesn't mean there isn't a benefit to having so many potential sources of innovation under one roof. If just a handful prove successful then they will have fulfilled their mission of fostering Australian based tech talent.

Thursday's event was a special offering for investors interested, but not yet fully embedded in the world of angel investing and venture capital funds.

Dave McClure is a founding partner of "500 Startups" a silicon valley based tech accelerator that offers to:

 "... invest $100k in exchange for 7%, and charge a $25K program fee for a net $75K investment."

They have four funds. The first is the flagship fund; the second is a follow-on fund. The other two have a geographical bias (LatAm and SE Asia). McClure wasn't presenting the funds per se; rather he was there to talk about how 500 Startups approaches building a portfolio of possible winners. Eerily for me at least there was no secret sauce, no magic formula for picking winners, nor any new trend to follow. His presentation reminded me of one of those lectures on portfolio construction. It was blunt and to the point. If Dave had been talking to potential retirees his laid back, up-front style would have been a winner; with a room full of investors of various experience levels it was a good wake up call. Quite simply when you're dealing with companies in their infancy volume of investing becomes a valid approach. If 80% of startups don't make it, then how can an old hedge fund manager like me bring the mentality of 3 - 5 core portfolio ideas to investing at this level?

According to 500 Startups you have to approach the opportunity as being extremely asymmetric. To give you an idea of how this differs from a hedge fund investor approach, you immediately draw a red line through investments that present as asymmetric. If you see a fund that has losing months much bigger than their best winning months, you walk away. As one investor said to me in 2006:

"If it shits like an elephant and eats like a bird, you're in trouble."

That guy is just not going to cope with a fund that has 80% of it's picks go under, another 5 - 15% barely pay you back and a few success stories in the end. Maybe if you bag the proverbial "unicorn" the asymmetry might be acceptable, as in:

"I invested in Google when Larry and Sergei were still at Stanford, and I doubled up at every subsequent offering."

I hate hearing from "Google-guy" at events. I do get it that if you make a billion dollars you can afford to leak out 50 million or so in tiny 100k bets in an effort to add to your unicorn stable. And that essentially is the proposition.

500 Startups is in the business of buying cheap options. Build a portfolio of tiny $25k bets across a multitude of plausible businesses and ideas, and then look for that inflection point where they're getting traction. Typically this is when you start to see validation through revenue or audience. At that stage, your 25k has bought you the right to make a decision as whether to double (triple, quadruple, etc.) up.

When Dave McClure says that these options used to be cheap, I believe him. In fact he says they are still. It reminds me of trading equity options in the early nineties. Before the '87 crash options, especially put options were insanely cheap. Afterward, options in general became extremely expensive as investors started to calibrate second level inputs into their models (skew and kurtosis as examples). You have to add to this the propensity of cyclical crashes. It's easier to understand the attractiveness of making a large number of high leverage, small cost events in the hope of another '87, '00 (dot com), '08 (housing) type event. In '93, my boss wanted me betting against this as the cost was far greater than the opportunity. We wanted to be like a casino that knows that occasionally someone on the roulette table gets lucky. 500 Startups sees it more like pre- '87. Things are still cheap, and the portfolio of long ideas properly structured looks likely to show an outperformance. Take the roulette example. Right now portfolio investing in startups is like playing roulette, except that occasionally instead of paying out 35-1 on a 37-1 event, you get a jackpot that pays out several million to one for the 37-1 event. That's why these options are still cheap.

And just to prove the point about option pricing, here's a Monte Carlo from Patterns of Successful Angel Investing by Simeon Simeonov
It's interesting to an equity guy like me that the optionality in the startup world didn't get more expensive. If crashes lead to liquidity events, that in turn make options more expensive because the risk reward ratio has changed, why hasn't this happened in the case of the tech world? Part of the answer according to McClure is the downward spiral in business costs associated with tech startups. In 1999 $25k wouldn't have got you much computer power. Now $25k gets multiple computers, some office space, and some server capacity, for a good chunk of that crucial first year. The founders biggest cost is time and sunk costs before being in a position to market the idea to an accelerator or angel. The leverage available has increased to compensate for the failure rate. Meaning the portfolio theory is still valid.

I hope none of the above sounds too disingenuous because it's not as simple as throwing darts at a list. Say, 500 Startups hears from 1,000 businesses each year and interviews 10%. Then they offer a small fraction of these some cash. That task is enormous in itself. And think about it this way; if you believe that the opportunity is in the portfolio and not trying to find unicorns then this is what you have to do. That's why as an investor you may be better off putting your cash into a fund rather than trying to build your own portfolio of options. I haven't looked at the returns from 500 Startup's various funds because that wasn't the idea of the presentation. Don't get me wrong, I wanted to ask Dave about his Sharpe Ratio, but somehow that seemed a bit rude given that he'd just sold me a very cheap option on investing in his world. Fishburners kindly brokered that transaction.

Thanks to Dave McClure of 500 Startups and Murray Hurps of Fishburners for allowing me to attend the event. Please don't hesitate to contact me directly through my website http://www.ibcyclist.com if you'd like to hear more about what I thought of this and other events I've written about in the blog.



  1. The costs of distribution and integration have significantly decreased since 1999
    - With 10 mins of work and $2 you can advertise to ~ 1000 people
    - Open "apis" are now so widespread the costs of interfacing with everything from payment systems to physical sensors are trivial.

    Thus the cost of failure has decreased significantly, the portfolio approach makes good investment sense