Friday, May 26, 2006

The 5% solution

Hasty generalization

From Wikipedia, the free encyclopedia
Hasty generalization, also known as "fallacy of insufficient statistics", "fallacy of insufficient sample", "fallacy of the lonely fact", "leaping to a conclusion", "hasty induction", "law of small numbers" or "secundum quid", is the logical fallacy of reaching an inductive generalization based on too little evidence.

Examples with contradictions
"I loved the hit song, therefore I'll love the album it's on": Fallacious because the album might have one good song and lots of filler.
"This Web site looks OK to me on my computer; therefore, it will look OK on your computer, too": Fallacious because many computers present content differently.

_________________

So why did I post this seemingly random snippet from Wikipedia?

This comment was made in an earlier post, and i thought it deserved to be highlighted as a separate posting:

That's why over ninety percent of such startups fail. And that just doesn't apply to alt.space or dot.com. In the early 80s, it was personal computers. Anyone remember the Osbourn?

Of course even knowing how to run a business is no guaruntee of success. Andy Beal is a great banker. His launch company failed when the market collapsed on the late 90s.

In the great scheme of things it doesn't matter. The five or so percent of alt.space companies that succeed will change the course of history.


So here is the issue i have with statements like this - it is the same fallacy that many VCs (and everyday investors who followed the investments of those VCs) made during the dot com bubble.

In the traditional VC world (you know, the guys who were around in the 70s and 80s, long before it was "hip" to be a VC and all you needed was an MBA from the right east coast school...), the venture capitalist was someone who had built and run one or more successful businesses, made a decent pile of money, and was looking to help the next generation of entrepreneurs build more great businesses.

These "Old School VCs" (OSVCs) would do incredible due diligence in the process of deciding whom to invest in - it was about more than JUST the business, it was about the management, too. But good management and a BS market was still BS. This due diligence was NOT, unlike today's New School VC (NSVCs) model, about waiting until another VC came in and then jumping in too, pretending to do proper analysis while blindly chucking in millions of dollars into a venture because "Benchmark Capital" or (insert big name VC here) had already agreed to invest.

THIS IS NOT DUE DILIGENCE, FOLKS.

It's betting that someone ELSE has done the homework, and just copying their answers.

So, back to the OSVCs. What they found, over time, was that if they were to invest in 20 companies, ALL of which had been grilled through the harsh cold logic of the OSVC due diligence machine, the performance would be something sort of like this:

  • 10 failed

  • 6 muddled along

  • 2-3 did ok

  • 1-2 were home runs



So out of 20 seemingly sound investments, 5-10% actually became winners. The rest were dead, walking dead, or just poor performers.

This statistic was regularly abused by the NSVCs (and still is) - in that they first deluded themselves, and then the larger investment community (including individual shareholders) that they actually understood how this model works.

It's not enough to just make 20 investments. you have to make 20 SOUND investments, and even then you can only hope to get a 5-10% hit rate. But if you invest in stupid businesses like pets.com and 19 similar ventures that show a fundamental inability to grasp what the WWW actually was (a distribution channel, not a manna from heaven), then you are likely to have a 0% success rate. 20 bad bets won't get you the distribution of wins/losses that 20 good investments will.

So, in the alt.space context - in isn't simply enough that there are lots of startup companies out there all seeking funding. There have to be lots of OSVC-ready companies - and even then, if we're lucky, 5% of them will be successes.

My other little quibble with the comment from above that i used to start this blog is regarding Andy Beal. His rocket company didn't fail because the market collapsed in the late 90s. That "market" was a paper one, based on fictitious satellites that would never see the light of day and the VAST majority of which were never funded. Building a rocket company on the late 90s launch industry market forecasts for satellite demand was a recipe for failure. As he found.

8 Comments:

Blogger Shubber Ali said...

That's why you won't find any VC fund that lists "aerospace" much less "alt.space" as an portfolio focus.

Really?

I'm sure that will be a big suprise to my friends at SpaceVest Capital, which has raised and invested three funds in... um... spacey stuff (as well as more mundane internet/wireless stuff on the ground).

Saturday, May 27, 2006 4:14:00 PM  
Blogger Mark said...

Just to make things clear, since there seems to be a misreading of my previous remarks, I was not disdaining the necessity of due dilligence. I was speaking really from the point of view of history. Also, we know that the sattelite cellular market was a "paper market" only after the fact. That sort of thing has happened before and will happen again.

Monday, May 29, 2006 5:53:00 AM  
Blogger TomsRants said...

Shubber, while SpaceVest may invest in "spacey stuff", it's very old school and mainstream, as a rule. They invest institutional and pension money. They are not "VC" that I know of, and they do not get into alt.space, last time I checked. Nothing speculative or high-risk. There are good reasons for that, however. ;^)

Monday, May 29, 2006 6:15:00 AM  
Blogger Shubber Ali said...

They are not "VC" that I know of, and they do not get into alt.space, last time I checked. Nothing speculative or high-risk.

Tom, that is true NOW - but take a look at the companies they invested in with their first fund (and even to some extent larger second fund). Companies like Space.com and, yes, AstroVision. But they also invested in real businesses that were not based on future possibilities but actual existing customers, etc. AGI turned into a huge success for them - after the business model was changed, of course.

High risk is never the issue - that's the difference between VC and Bank. Believable high return within a reasonable time frame through an understood exit strategy IS the issue.

Monday, May 29, 2006 2:27:00 PM  
Blogger Shubber Ali said...

if you look at SpaceVest's current portfolio only two qualify as space related...

thus reinforcing my point - there are very few investments that qualify for a VC, even one which has tried to focus on Space.

Monday, May 29, 2006 2:28:00 PM  
Blogger Shubber Ali said...

Michael,

first - your question:

Are you claiming that if you can't meet the requirements of a VC fund that you're a bad investment?

Yes. Unless you change the definition of "good" investment to include gambling, I believe it is a "bad" investment. As you implied (or perhaps i just inferred it) from this earlier comment:

Angel investors have vastly different investment justfications: ego, dreams, glory, a simple desire to fly, a hunch that something big is up, etc.

Ego, dreams, and glory aren't the qualifications for a "sound" or "good" investment - they are exactly what they are. Don't mistake them for sound investment criteria simply because the odd one occasionally pays off. MANY don't. And i would argue that the failure percentage is much higher than that of VC investments, which i think you'd agree with.

f so then you're basically arguing against the entire private equity method of funding startups. I can't imagine you're actually suggesting that.

No - because, first off, VC *IS* private equity. It's just much more diligent private equity. Friends and family rounds are done because they are friends and family, and therefore less discriminating. VCs are responsible to other people whose money they are investing. So the bar, and the share of the company demanded, are much higher.

Monday, May 29, 2006 5:34:00 PM  
Blogger Shubber Ali said...

There isn't a single company under the "alt.space" moniker that qualifies for VC level funding and there won't be until you have several large M&A events or a significant IPO.

Why would a significant IPO or large M&A event change the investment-readiness of some other alt.space company? If i understand what you're saying, the event would lend credibility to the rest (or some portion) of the alt.space marketplace. Perhaps. But that is dangerous reasoning, because it would be like saying "hmm... Google went IPO, so i should invest in a search engine company startup". That would actually depend on the company, how much of the "market" is left, and a whole host of other normal investment factors.

But now that I at least know the definitions you're using I can understand why you say the things you do. There's nothing wrong with them. Its just been my experience that few others use them the way you do...

Our experiences differ significantly, then. But the definitions I used were taken straight from your own earlier post. You were the one who cited angel investment as often being driven by " ego, dreams, glory, a simple desire to fly, a hunch that something big is up, etc."

So I stand by what I said.

Tuesday, May 30, 2006 4:16:00 PM  
Blogger Stephen Fleming said...

I think we're seeing a classic Internet phenomenon here... two (or more) smart people talking past each other while not really disagreeing.

So let me dip my own oar into this water. As someone who made his living as a VC for ten years -- and I like to think of myself as one of Shubber's OSVCs -- I can confidently state that there are excellent companies (and excellent investment opportunities) who simply are NOT candidates for VC investment.

First off, funds are raised from large institutional investors based on a particular investment philosophy. The fund will run for 10 to 12 years. If, in the course of that 10 years, an individual VC stumbles across a fabulous investment opportunity that doesn't match the stated philosophy of the fund... too bad. The limited partners (the big institutions that committed the funds) are not interested. If you sold yourself as a biotech fund, you had better invest ONLY in biotech.

(Note that hedge funds do NOT have this limitation, and may be more interested in alt.space.)

Second, even for the rare-as-hens-teeth firm (like SpaceVest) which at least nods towards space-related investment, given that ten-year horizon, the deal needs to have a high likelihood of positive exit within whatever remains of the ten-year life of the fund. Hard for any of the current alt.space companies to make that claim, because of the lack of exits that Michael has pointed out.

Third, don't lose track of the fact that (here let's ignore the alt.space market for a moment) there are thousands of excellent companies being funded every year that are not candidates for VC investment. Service companies are the obvious category leader here... when your outputs scale linearly with your inputs, it's difficult to get the "home run" sorts of multiple that Shubber points out as being necessary to succeed as an Old-Style VC. Doesn't mean they are bad companies, or even bad investments... if, for example, you'd like to take dividend payments, you can get a handsome rate of return with no liquidity risk. But that's not what VCs do.

Venture capital is wonderful, but it's not magic. There are lots of ways to get startups financed. I've just started talking to a venture leasing company about alt.space, and they're seriously interested. Different model, different dynamics...

Thursday, June 01, 2006 7:13:00 AM  

Post a Comment

<< Home