the people who work at tech companies in SV are like super brilliant and incredibly capable, and the last thing a person like that wants is to not to be able to get in the game and like help fix what is wrong.

...and let me tell you: every company in SV like including Facebook, including Square, including Twitter, including all of them, has got shit horribly wrong, horribly horribly broken and wrong and everybody in the company knows it and the CEO knows it, so it is better to talk about it. Because that’s how companies are, like stuff just breaks. If you have like 500 or 1000 people trying to do something, it is not all going to be perfect, it is gonna be ugly and you get that out there and you let people work on it. And that is probably the number 1 thing I would say about motivating people when stuff goes wrong.

- At http://www.youtube.com/watch?feature=player_detailpage&v=sqI...:

If you don’t have winning product, it doesn’t matter how well your company is managed, you are done

- At http://www.youtube.com/watch?feature=player_detailpage&v=sqI... :

Why even bother with management if that is the case? But the truth is: there are several things that are very important, one is if you get into trouble, and if you have bad management, your company will probably die. Like the people will just quit, they are not bonded to it, the don’t like working there, they never liked working there, and that is that, it is a wrap. ... if you succeed at building a company that everybody just hates working at, what have you done? You just made a whole lot of people a whole lot more miserable in their lives.

[1] https://www.youtube.com/watch?v=CqODNbFz6EY

I listened to Marc A on this week's Freakonomics podcast, giving a very plausible rationale for why we should still care about crypto-currency. (Go listen - its good)

I am trying to salve my conscience and turn open source software and public sector into a mission (#) - but despite loving his pitch and wanting to believe in the Golden future, I must have too-small-a-mindset.

How on earth can you spend 1.5 billion on startups in the next couple of years? I could take 5 million right now and do some amazing stuff in message queueing in local UK government - the future is bright, but cheaper. Are they wanting to put in for 300 small startups that will change our infrastructure but essentially make things cheaper, reducing the total cake? Or is this 100M for a few "build the next rocket" startups?

(#) oss4gov.org/manifesto - I still think its the best I have yet expressed about how angry I get seeing proprietary code running government.

According to Fortune, there are management fees required on the full $1.5B. It's a huge fund (same article pits the average American venture fund at $107mm), but given the fees and the assumed ease of the fundraising process, it's easy to understand why they went this far.

For all the tension between Andreessen and Icahn, they share one quality that few investors can claim. When Icahn buys AAPL and let's people know, the lemming effect allows him to profit merely from announcing his own optimism (note he doesn't sell off of this, but his shares are still more valuable). At this point, a16z essentially has the same star power. Their blessing can give a serious push to companies pursuing an IPO or acquisition, to the point that the boosted exit value overweighs any "overpaying" a16z might do on the entrance.

You are thinking small, seed or maybe round A. Rounds B and C funding is what you need to scale up a company to Dropbox or Twitter level of infrastructure. That can easily be 150-450m.

Granted not all of companies would need that kind of investment, but every once in a while...

Thanks - just thinking on the message handling front:

432 councils in UK, handling 22m households with avg 120 interactions per year : ~2bn transactions or 40m per council per year. As a shared infrastructure that's on the order of 40m per day. It's big - yes I could see how you could spend 100M usd, but that shared infrastructure approach is a big leap of trust - put it this way, Kent the council I live in is spending 40m on shared Citrix and shared firewalling for police, councils, fire etc in the area.

I just think we could do the next layer up a lot cheaper

You could and you should pitch it as Silicon Roundabout.

It is still expensive to build big companies, and $1.5B is tiny by financial markets standards. Bond firms like Pimco and Blackrock measure things by $100Bs and trillions.

Let's say you invest in 50 startups per year for 5 years. That is 250 firms, averaging $6M/firm.

1.5 is small compared to the war chests Apollo, KKR and the like amass, but it's huge for a VC accustomed to writing smaller checks. Average US VC find is 107mm.

They're either going to be doing way more deals or shifting some of the fund focus to later stage. I expect the former, given that investors seem to like the granularity of separate venture and growth funds to pick from.

You're forgetting money that they can leverage from banks with the money they raised.

I've never heard of a VC leveraging up with borrowed money. Who would be on the hook if the fund loses money?

VCs' portfolio companies often raise debt. This is called venture debt and the lenders are banks like Silicon Valley Bank.

VCs tend to have a love/hate relationship with venture debt. On the one hand, it means they suffer less dilution as the company grows; on the other hand, it means their investment gets riskier just at the point where they are finally de-risking the company.

It's not out of line of their current rate. In 2013 they had 56 new deals and 81 total investments[1]. Over a few years that could cover 1.5B.

[1]http://blogs.wsj.com/venturecapital/2014/02/04/500-startups-...

How many of their funds were active in 2013?

Pays the rent at least... Rents are going up!

Software is eating the world, and a16z is eating the venture capital world. It's no surprise LPs are throwing money at them.

They've been around for less than five years and are already setting the standard by which other VCs are measured.

Case in point: they made something like a 5-10x multiple on their investment in Oculus-- in 3 months.

They are doing quite a good job as VCs, but lets not start painting a picture of skill based on one of their companies.

Investing in startups is gambling, and A16Z is basically rolling up to the roulette wheel and placing a bet on every square. The only difference is the payout of one success is (maybe) large enough to offset the loss of betting on everything else.

I can't look up the numbers now, but I'm pretty sure when the investments are pooled, the returns are modest, relatively in line with returns on other types of investments. (Would love for someone to back this up or correct me if I'm wrong based on the numbers).

Fred Wilson gave a nice overview.[1] Indeed not much better than other investments.

[1]: http://avc.com/2013/02/venture-capital-returns/

The returns for top notch VC firms like a16z and the returns for VC as a whole are like a difference in kind.

Sure, but they still aren't abnormally outsized, probably even before adjusting for risk. If they were they would do like the very top hedge funds and charge more than 2/20.

If you know specific rates of return please share.

A set of example return information (albeit old and only from the first fund) is here:

http://finance.fortune.cnn.com/2012/07/23/nice-ira-andreesse...

I would anticipate there would be diminishing returns over time as the "bets are spread" if you will... that being said they're doing a damned amazing job so far.

Besides being only a single data point the returns for one $300M fund aren't a good indicator of the returns for 2-3 concurrent $1+ billion dollar funds.

I concur...that's why I said: "only from the first fund" and "I would anticipate there would be diminishing returns over time as the "bets are spread" if you will".

That being said it is counter to your statement "Sure, but they still aren't abnormally outsized, probably even before adjusting for risk." That particular data point is quite higher than both the median and the means of VC returns that I've seen (and would probably lend more credence towards DaniFong's statement). I look forward to seeing how their other funds progress so that we have more data points.

note: edited in a paren that was missing.

Well statistically speaking (IIRC) that the top 10 VC firms incorporate 86% of the overall gains in VC, and the top 15 make up 93%.

> Case in point: they made something like a 5-10x multiple on their investment in Oculus-- in 3 months.

But no one ever talks about the numerous businesses that fail in their portfolio, which ultimately is why such large investments are required in order for them to sustain profitability. VCs know this won't last forever, so they have to milk it for all it's worth today.

> Case in point: they made something like a 5-10x multiple on their investment in Oculus-- in 3 months.

Oculus is an exceptional case, but that's how VC funds work.

A minority of investments pay off big time and make up for often catastrophic losses on the remaining majority of investments.

Honest question, as I've never really understood the industry: where does this money come from and why? And where does the money go that acquisitions and IPOs earn them?

Venture funds are "General Partners" that manage a pool of capital on behalf of "Limited Partners" or LPs. Limited partners in a fund this size are usually pension funds (CalPERS), university and not-for-profit endowments (Harvard, Brown, MacArthur Foundation), and maybe high-net-worth individuals.

So A16Z raises commitments of 1.5B from all of these people who have a lot of money that they want to grow. Those people deploy their capital to VC, to Hedge/PE funds, to timber, to public markets, to bonds, etc.

A16Z aren't actually given a pile of cash though, instead A16Z does capital calls on an as-needed basis. Then A16Z deploys that capital to startups and other companies over the life of the fund (often ten years with some extensions).

Most venture funds operate on something like a "two and twenty model" meaning that every year they get 2% of the funds value for operating expenses. In A16Z they have a huge staff of designers, recruiters, BD experts, etc, that help their startups. The 2% goes toward things like that.

The 20% is carry, which is earned as a percentage of net income. So let's say that A16Z invests 5M in a company on a 10M pre-money valuation. A16Z owns 1/3 of the company. The company sells for $30M the next day to Facebook. A16Z now has $10M dollars. They take $2M of that for themselves and then they distribute back to the LPs the other $8M.

(This is lacking a lot of nuance like side funds, but I hope it's helpful).

It's worth saying that at a number of firms that 2% goes into the pockets of the partners.

One thing to like about a16z (full disclosure, I worked there over the summer), imho, is that by using this fee to scale their organization to help portfolio companies, they are far less out of alignment with LPs than most firms who have success with a smaller fund and go on to raise a larger one.

The old algorithm of do ok with a small fund, raise a huge one and live off the management fee is under threat -- 5x or die.

I'm pretty sure carry is profit (http://en.wikipedia.org/wiki/Carried_interest), so in your example A16Z would pocket $1M, not $2M (because the $5M A16Z invested needs to be repaid to the LPs first, and then A16Z gets 20% of the remaining $5M). Please correct me if I've misunderstood.

Oh ya, that's my mistake. I was right when I called it "net income" but then didn't do the net in the actual math! I'll leave it unedited so your comment makes sense though.

Thanks a lot for responding to me again Zac, helpful as always!

(Edit: I realised you usually go by Zac rather than Zach :)

It comes from large investors: often university endowments or public pension funds. These groups are called the "Limited Partners", or "LPs". It's usually a small percentage of those investors' overall investment portfolio.

When there is an exit--a sale of stock--most of the money goes back to these LPs. The money is (more or less) split like this: the LPs get their investment back first then, of the gain, 80% goes to the LPs and the rest goes to the company managing the venture fund to split up as it pleases (this last 20% is called the "carry".)

Thanks a lot!

VC's are typically raising money for their funds from even larger funds -- ie hedge funds, private equity firms, endowments, pensions, etc. Any funds with a lot of capital that are looking to diversify. These larger funds become LP's (http://en.wikipedia.org/wiki/Limited_partnership) and generate a return typically at the end of some time period -- something like 5 years. So any money from acquisitions/IPOs etc will be returned at that point.

Thanks, these replies are very helpful!

It's interesting to see a mention of "full stack" startups. I hadn't realized that investors have identified such companies as being members of a distinct category.

It's a specific reference to a blog post by Chris Dixon: http://www.cdixon.org/2014/03/15/full-stack-startups/

Does anyone know what kind of IRRs a16z is seeing in previous funds?

The amount of cash available to startups right now is breathtaking.

The biggest returns in this fund will be in the cryptofinance space.

How does one successfully apply to a fund like this?

Warm intros only, right?

Man Andreessen has come along way since Mosaic.

Reply on news.ycombinator.com