This isn't as insane as people might believe. Let's look at the larger of the two funds: $1B. It's a digital growth fund. Raising that $1B doesn't mean that people wrote a total of $1B in cheques: that money is just "committed" to be called (ie: a capital call).
A simplified fund's capital call schedule might be something like 25%/25%/25%/25%. So each year, for the first four years, they "call" a quarter of the capital, or a total of $250M. Years 5-10 are for harvesting, though they might have one or two more extension years beyond that if the portfolio isn't ready to reap.
Since this is a growth fund, we can see them deploying it over something like 5-10 companies per year for the first two years, then following on in years two and three in later rounds pro-rata with some of the winners. That means that we're only looking at 5-10 companies, $25M-50M each in a series B or C in each of the first two years.
That seems at odds with what I assume is hundreds or thousands of companies in the digital space that could absorb a few million.
If you are limiting yourself to picking ten companies, it seems no one has learnt the lessons of YC - or is this something I would know if I was cleverer?
Keep in mind that their bottleneck is likely not capital but available partners. One partner can only handle a couple of companies/board seats. Andreessen calls this having only so many tickets to punch.
Another big raise. It says to me that the VC's have a seller's market right now, in that a lot of limiteds are throwing money at the VC's expecting the Unicorn Ranch to keep producing. I'm not sure where $4B-$8B in recent raises is going to go. They are going to be hard pressed to find enough rat-holes to stuff that much money into. Of course, the Unicorn bubble could continue, but at some point I'm thinking valuations are going to reconnect with revenues, and then what?
They probably have 10 years to deploy the capital though with any fund they have to invest some each year to keep up with the IRR they aim to provide for their LP's.
"they have to invest some each year to keep up with the IRR they aim to provide for their LP's"
Assuming that VCs measure IRR in the same way as other Private Equity (i.e. based on capital calls and capital distributions), then you don't need to deploy capital early to maximise IRR.
For example, let's say a fund had a 10 year life and commitments of $10bn:
- Years 1 to 9: no capital calls, and no investments
- Year 10, 1st Jan: 10MM called and invested
- Year 11, 1st Jan: 15MM received from liquidating the investment, and immediately returned to LPs. Fund closed.
The cash flows are {0,0,0,0,0,0,0,0,0,-10MM,+15MM}.
The IRR of those flows is 50%, which sounds pretty decent, except that:
- The LPs capital was deployed for only a year
- Even for that year, only 1/1000th of the committed capital was called and deployed
So, the VC fund's performance (measured by IRR alone) looks good, but the LPs paid fees for nothing. They had to manage 99.9% of the cash themselves for the whole time.
> They probably have 10 years to deploy the capital
They probably have about that much time to return the capital. 10-year fund doesn't mean you get to keep 10% of the capital in a box for a decade (apart from reserving for management fees).
There's heterogeneity in the "they" we're referring to. As a rule, no, funds do not reserve capital for follow-on rounds. A venture fund's cost of capital (i.e. expected return) is high, making the compounding cost of holding cash cumbersome.
Instead, a firm will generally raise one fund for early-stage (e.g. seed) and another, either later or if they have a strong pipeline concurrently, for later investments (e.g. A), where preference is given by the latter to companies invested in by the former.
1. The funds don't "hold cash" for follow ons. They "reserve capital" which is subsequently called. There's no dead weight uninvested cash drag on irr (or properly, very little of such).
2. "Crossover" investments between funds are generally frowned upon, although the pendulum swings on that practice and more recently it seems they are in favor again. Still, it is the exception rather than the rule that the same manager (VC X) is allowed / supposed to put money from two different funds (X fund I and X fund II) into subsequent series of the same company's stock. If you think about it, that makes a ton of sense because in downside cases, VC X may then have to play King Solomon and split the baby between its Fund I and Fund II investors, both of whom it has a fiduciary duty to, hence a conflict it's best to avoid altogether. Now, that is a problem that comes up in downside cases and recent times have been good, so once again people are overlooking the conflicts in the name of keeping the punchbowl full and spiked.
3. Funds do most definitely reserve for and participate in pro rata follow ons in subsequent rounds. 100% of market early stage A/B/C term sheets will ask for pro rata rights (for follow ons). If you don't have follow on reserves you risk facing a "pay to play" or other punitive term in a future round (again only in the downside cases which people start to neglect in the up cycle times like the last 2-3 yrs).
Then they failed. In the same way as a startup that raised $x in VC and after burning through 50% decided that it didn't work and give the money back would be considered as failed.
Investors invest because they expect a certain return. Doesn't matter if it's VCs investing in startups or LPs investing in VCs.
In case you're wondering if they would have to give the money back: Technically, they don't have all the money. VCs 'request' money over time from their LPs (they do 'capital calls').
I wouldn't be surprised if this results in more investments outside of Silicon Valley, in new rising American cities and countries in Asia mainly and some in South America, Europe, and Africa.
These funds live for a decade. The most recent funds are almost a decade away from the moment of truth.
To the parenthetical question: You invest something in a few of these funds, and a decade or so later you learn whether you own 1% of Google, or whether all of your VC funds picked the likes of Excite. People invest because Google does exist.
Which reminds me of a classic ah-investment strategy. Say you have €1m of white money and €10m of black money. Invest in 11 high-risk ventures (coffee or tomato futures, say). Later (when the new harvest is in) when you see which one won, go back and fiddle your paperwork so the white money paid for the winner and the black money paid for all the losers. I've heard about this done with volatile foodstuffs, but never startups. Perhaps ten-year funds simply are too slow? Comments, anyone?
The white/black money thing is not in vc land but is super common in "strategy" focused hedge and even mutual funds. You "incubate" 10 funds with very small assets and then after a few years pick the best one and trot it out to the market and pitch it based on its true but misleading performance.
It's p-hacking for funds.
For various reasons this particular shenanigan doesn't happen in venture, though. (hint: it's super easy to scale down a public markets strategy to micro or pilot scale of say 1/10,000)
I am always curious about who are the LPs are and the slide deck KP used to raise more money. What was the deciding factor in the LP committing more money. Is it because there are no other options for LPs to make a decent return (close to zero interest rates).
Top 5 in Cachet in no particular are probably
Benchmark, Andreessen, Sequoia, Greylock, Accel. KP comes after them. It used to be number one. Not anymore. There are so many angels, good old/new VCs (YC, LightSpeed, FirstRound, Union Square, SocialCapital etc) that calling out KP for prestige doesn't make sense anymore.
Getting funded by KP now means => You raised money from a good VC firm. Nothing more. Years ago, it was more like - "Wow, you guys raised money from KP? You must me special"
A simplified fund's capital call schedule might be something like 25%/25%/25%/25%. So each year, for the first four years, they "call" a quarter of the capital, or a total of $250M. Years 5-10 are for harvesting, though they might have one or two more extension years beyond that if the portfolio isn't ready to reap.
Since this is a growth fund, we can see them deploying it over something like 5-10 companies per year for the first two years, then following on in years two and three in later rounds pro-rata with some of the winners. That means that we're only looking at 5-10 companies, $25M-50M each in a series B or C in each of the first two years.