[frinxor]

I think pg's whole point (and Taleb's) of a black swan is that they are not identifiable. Over the long term and after many investments, pg is pointing out that its the presence of these black swans that made YC successful (dropbox, airbnb), and extending that to other VCs, its random luck that will make or break them. You can't avoid investing in black swans, because you don't know which ones they are/will be.

[davemthrowaway]

That wasn't how I read the black swan farming piece. From what I remember he pointed out how total returns for the fund are dominated by the black swans. Then he discussed what yc would (theoretically) look like if they were going to try to optimize for black swans but conceded that they wouldn't really want to run yc that way. In a way I think the piece flouted the concept of black swans by saying they were something you could optimize for (which I think taleb would disagree with).

Regardless, the unpredictability of black swans and their potential impact on a fund's returns show the ridiculousness of Dave trying to frame this discussion as "Barry bonds vs ichiro".

[btilly]

Taleb not only believes that you can optimize for black swans, but he runs a fund called Universa that does exactly that. Most years it loses money. But in 2008 it had over 100% returns.

[tatsuke95]

Universa is run by Mark Spitznagel; Taleb is only loosely associated. Because it's a closed, private fund that isn't required to produce audited financials, nobody on the outside really knows what kind of returns it has produced.

Taleb and his theory make for a good story in the wake of the financial meltdown, but the tangible application is sometimes exaggerated [1].

[1]http://www.businessinsider.com/wait-before-you-invest-in-nas...

[btilly]

That's what I get for skimming to remind myself of facts instead of reading.

I got Universa from the Google search blurb from Wikipedia. I got the over 100% figure from http://www.bloomberg.com/news/2011-10-06/black-swan-money-ma... which quoted "a person familiar with the matter" as saying that the fund brought in 115% during 2008.

However if I read either closely it was obvious that Taleb does not run Universa.

[tatsuke95]

I wouldn't fret; Spitznagel is open about the fact that the firm's strategy is based off of economic "Black Swans" and Taleb's work, and there's definitely an association.

As far as the returns, the 100% figure wouldn't surprise me, really; more than a few bearish funds crushed it during 2008. It just seems like that number is the financial media echo chamber in action. It's repeated over and over in article after article (just look at the citations for it on the Universa wikipedia page). And yet, it always comes from an unnamed source or someone "familiar" with the fund. If there's anything I've learned from the financial crisis, it's that hedge fund managers aren't particularly trustworthy; even less so when they're talking their own book!

[davemthrowaway]

This is sort of tangential to the main discussion. That said, I thought that strategy wasnt "optimizing" for black swans (as in trying to find specific investments where a black swan event is likely) so much as accepting that the market underestimates their probability and holding a wide portfolio of securities which will payoff in a dramatic downturn. That may be a bad definition of optimizing, or just an incorrect understanding of their strategy though.

[btilly]

I would call it a somewhat incorrect understanding.

They do a lot of their investing in out of the money options. Those are investments that cost very little to make, which will make a lot of money if something very unlikely happens. (You can make these investments on both sides - both up and down.) Lots of people are willing to take the opposing side of the bet based on standard models. If those standard models are significantly underestimating the risk of unusual things happening, then in the long run the black swan strategy will make money.

Since by definition you don't know where a black swan will happen, you also need to spread your investments out across many places. But you don't just diversify, you diversify across very specific classes of investments that are very cheap to make and have a very low probability of paying off.

In the startup analogy you'd want to diversify across as many startups as possible, and invest in them as early as you can.

[ggggx]

Seems to me none of you guys have the correct understanding of what Spitznagel does at his Universa. He actually uses these out of the money options payoffs not to bet on random black swans, but rather to exploit really big moves that he is actually counting on (apparently based on bs understanding of Austrian economics). All this black swan talk was BS afterall. He says... "black swan events have been largely insignificant in at least the last century of capital investment in the U.S., including the current crisis. Investors have indeed encountered surprising and pernicious events, but the fact is those who were surprised have essentially been those (in the extreme majority) with a brazen disregard for the central concepts of Austrian capital theory and monetary credit expansions." in his paper "The Austrians and the Swan: Birds of a Different Feather" here http://www.universa.net/UniversaSpitznagel_research_201205.p...