Seven to Ten Years

I have worked in three venture capital firms over the last thirty-three years and am intimately familiar with the performance of the fifteen (ish) venture funds raised and invested by these three firms. Much of what I have written about fund management and investment performance here at AVC over the last sixteen years comes from my observations of these funds and firms.

Starting in the mid-00s, The Gotham Gal and I started investing in other venture capital funds, always limiting these investments to firms where we knew the partners well and had sat on boards with them.

And The Gotham Gal started angel investing around the same time, often writing the first check into startups. She has made something like 140 angel investments over the last dozen years, mostly into companies founded by women.

We keep good records on these personal investments and I now have another data set to observe.

Across these three sources of data (my firms, other firms, angel investments), there are well over 1000 individual angel, seed, and early stage venture capital investments over four decades.

I have no plans to publish this data. It is not in a single database and there is a ton of confidential information in it.

But I can observe things about this data and have been doing so and will continue to do so.

One of the great truths about early stage investments is that you have to be patient with them. The losses come early and the winners take longer to realize.

It takes seven to ten years to get to real liquidity in a portfolio of early stage venture investments. You can’t short cut it. It just takes time. But come years seven, eight, nine, and ten the returns will start coming in.

I am not sure why seven to ten years and not five to seven or not ten to fifteen. It’s seven to ten. That’s how it has always been and seemingly always will be.

#VC & Technology

Comments (Archived):

  1. David Pethick

    Fred – what a great dataset. I’d love to see some of the trends over the last 15 years in common metrics used by VCs. I also wonder if a dataset that large kicks up any metrics with predictive power for the outcome.

  2. jason wright

    Seven Years in Debt, starring Fred Grit. The new Heinrich Harrer.

  3. DanaJ

    Weird request with low chance of success but here it goes: would you be interested in sharing that dataset with me, to do some analysis and see what insights may be gleaned over the ~2-3 weeks I have in between my leaving consulting and joining a PE firm? If you’re still in Europe after the 23rd, happy to also meet and discuss f2f in any city with a decent sized airport!

    1. Tom Labus

      Good luck!!

  4. Michael B. Aronson

    Totally agree, when i entered the formal vc biz i had a 4 to 6 year track record on a handful of deals but its been proven to be in the range you quote for deals in philly and nyc. In philly its been 10 years plus, recently with Instamed at 15 in a great sale to Jp Morgan. Maybe there is east coast bias here? First Round capital’s turn seems to be quicker.

  5. johndodds

    Maybe it’s because neither of you are looking to flip your investments but rather are interested in real sustainable businesses solving real problems.

  6. Michael B. Aronson

    Would be interesting to know how many would have qualified for qsbs capital gain exemption , five year hold, and also 1244 ordinary losses for first million invested. Vc returns are always quoted pre tax but in real world pay taxes!

  7. Tom Labus

    Some great companies were abandoned by pressure to show early results. It’s a fine line between pushing and interference. But how do you know? I’ve moved out of stock multiple times and despaired more than once but still thought it was a great deal. And sometimes it was!!

  8. TeddyBeingTeddy

    Translation: your next fund will have a 10 year investment horizon?

  9. pointsnfigures

    Would be interesting to see if you kept track of initial valuations and then the average step up in valuation on the ones that stepped up-along with the ones that didn’t. I might add Brad Feld has said it’s “20 to build an ecosystem” which matches the 7-10. To build an ecosystem you need an initial crop of companies to pave the way for the next.

  10. William Mougayar

    It’s interesting to contrast with crypto land, where projects can get their tokens publicly listed typically 18-36 months after inception, giving them some kind of market cap (and potential returns for early investors).Then, either they tank afterwards, or take off depending on what they actually end up doing, and the scale of adoption they achieve.

    1. Our Man In NYC

      I’m not sure if crypto is the right comparison; I think crypto is more like what VC would be like if there was daily pricing/trading.By it’s nature as early stage the delta around fair value (at any point in time, let alone on a daily basis) is going to be wide and then throw in the impact of ‘sentiment’ and there’s a lot of noise around the signal in daily crypto prices. It might well be it takes 7-10 years to figure out what the winners will be and that to truly play out, but the sharp moves in prices create opportunities for those with shorter time horizons.

      1. Chris Albinson

        Tulip bulbs (-:

    2. JamesHRH

      You can also go OTC BB and ‘have a market cap’ of some sort.This is an equivalent comparison

      1. William Mougayar

        But OTC is not typically available for VC backed startups.

        1. JamesHRH

          You compared a mature startup to getting crypto token adoption in 18 months.That makes those tokens seem more like a OTC stock where you get it listed and fund it by producing press releases.OTC startups are a better equivalency to crypto. There’s no mature business in either of them.

    3. Joseph K Antony

      This is very interesting as tokens raises the possibility for far finer granularity in both funding and exits. Granularity raises complexity reducing risk. A model that deserves to be developed.

  11. T. Trent Gegax

    Fred, how has Gotham Gal tracked her portfolio over the years? Excel? Now, moved to Carta or Airtable? Something else? Appreciate it.

  12. sigmaalgebra

    On the 7-10 year interval:While I have no inside information about VC firms, from all I can see from outside VC it appears that somehow, maybe from consensus views of the more important LPs (limited partners), VC firms have surprisingly strong consensus, even uniformity, in how projects are conceived, planned, evaluated, and executed. So, there’s not a lot of variety or freedom for changes that might make the interval shorter.IMHO something that should help make the interval shorter would be better approaches to how projects are conceived, planned, evaluated, and executed.My background is only partly from VC funded projects, e,g., FedEx and there the VC funding was long after the founder Smith had invested many $millions of his personal fortune, but mostly from high technology projects for US national security.First, with the current, consensus, conventional system, VCs are not much involved in conceiving or planning.Second, shockingly, and of strong importance, VCs insist on a certain simplistic approach to project evaluation that essentially ignores any technology.For a broad view, the VC world is populated almost entirely by people who are not qualified in technology — mathematics, physical science, engineering — for US national security, either (i) planning, …, executing projects or (ii) evaluating them.E.g., at one time I was in the group that did the Navy’s version of GPS and heard the stories of how the project was conceived, planned, and proposed: Basically, before the project evaluation, funding, and execution, the core work was on the back of an envelope. More generally, US national security is awash, is heavily dependent on, projects that before funding and execution were just presentations on paper.From all I can tell, VCs just will NOT fund, or even seriously evaluate, projects presented just on paper.Instead, it appears that, in simple terms, VC funded projects (i) are proposed by some marketing guy — with at least some broad understanding of what can be done with current computer and communications hardware and routine software — who maybe has found a good application, (ii) have the technical work done by the largely self-taught hacker culture, and (iii) are evaluated for funding by VCs in terms of traction significant and growing rapidly in a large market.With that approach, the problem is that marketing guy and the hacker culture just are not very powerful for valuable projects and, in particular, have had next to nothing to do with the technology successes of US national security.So, in practice, the people good at the technology that has done so well for US national security largely stay away from the world of VCs: When such people have tried and written VCs beautifully polished proposals on paper, the results were the VCs ignored the proposals and the technology people gave up on VCs and, really, usually such entrepreneurship.Yes, such proposals should contain very clear explanations of how the project will achieve its goal: That IS done in projects for US national security. For entrepreneurship, sure, the proposals need to be equally clear on market, customers, revenue, and earnings, that is, how to achieve the goal of return on investment. The problem, the reason VCs won’t fund, is that the real promise of the ROI of the project is the power of the technology, and the VCs insist on ignoring that power.So, IMHO, those circumstances generate, essentially guarantee, the 7-10 year interval.But, in just the last few years, there has been a big change which should mean that the flip side, good side effect, of the VC approach is that for projects where the powerful technology is some mathematics, say, information technology projects, the associated computing and communications hardware and software can be so cheap that a thin personal checkbook has a good shot at getting after-tax earnings significant and growing rapidly. I.e., the crucial, core, powerful math might be only a few dozen pages of code, and with careful project selection otherwise the rest of the code, routine, might still be a one person project. Yes, the world has yet to see a lot of such projects that did an IPO, and this means a great opportunity. In simple terms, so far there are not many such mathematicians ready to take this opportunity. If they try, they might hear that VCs like “leading edge, new, disruptive technology,” etc., send in proposals that so emphasize, get back silence or nothing, and give up.There are several large, solid parts of our society quite ready, willing, able, and eager, with surprisingly good accuracy, to evaluate projects just on paper; it appears that the VC/LPs are not among those.For my project, I got held up: The work unique to the project has been fast, fun, and easy. but there have been some outside interruptions where handling those with just a thin checkbook delayed the core work. Some modest funding would have helped, but the project is doing so well now, so close to good revenue, that funding is not needed.The history of technology for US national security shows that the US military did well with technology surprisingly quickly.The main point here for making the time interval shorter is that changing how projects are evaluated should provide some additional opportunities. In particular, if what is on paper looks really good, hopefully has some proprietary powerful new technology, then maybe DO evaluate the project, AND its technology, and, if the evaluation looks good then get ON with the funding and the project.For making the interval shorter (i) get to start on the real project, the real goal, directly right away without a lot of struggling with a thin checkbook, (ii) get to consider projects that are especially valuable from technology that is especially powerful and, thus, can get to an IPO faster, and (iii) get to select projects that, really, are remarkably fast to bring to market because the real power is in the technology, which might be only the few dozen pages of code as mentioned above, with the rest routine and maybe surprisingly small.Net, starting with good work just on paper, if the work is really good, then will get a new variety of projects with more opportunities to make the interval shorter.

  13. Richard

    The early roi is psychological. With so many Investments in kids in their 20s, there is not a lot of incentive and too few life experiences to move faster. The board members also seem to enjoy the father son/daughter relationships. .

  14. JaredMermey

    Have you written about Fund Management? Number of bets you try to make per fund and how much capital you keep for follow-ons vs shift to Opportunity Fund?

  15. Ian Chapman-Banks

    Wow – I have only one data point my AI company I co founded is nine years old and only now are we showing real positive returns to our super patient investors – its true deep tech takes time – for us year eight was the real turning pointIt just takes time

  16. JamesHRH

    It’s close to a universal constant.FB started to seriously print $ in Year 8 as an example…..

    1. Lawrence Brass

      Print ads, more likely. 🙂

      1. JamesHRH

        They were CF+ in like 2006. But Sandberg showed up later and they committed to the model and buried the revenue needle in 2011/2012 – up 65%, etc.It’s amazing that it takes that long to pull all the prices together. Every founder should know this metric.

  17. Lawrence Brass

    Super interesting.What would you say is the thing that matures during this development period:The teams and people behind the project, the product or the process?Or maybe all of them at the same time?

  18. Eric Nehrlich

    Is there a correlation between the fund lifetimes and the liquidity you reference? My understanding (as a venture outsider) is that typical VC funds have a lifetime of ten years; if that’s the case, it makes sense that the liquidity events would happen towards the end of that time, as the VCs have to start extracting funds to return to their LPs. Do it too early, and the returns aren’t big enough (you’d rather have them compound for more years if the company is on track). You can’t wait until after year ten, because the fund ends and you have to return the money to investors. But that seems too obvious an explanation of why liquidity peaks in years seven to ten, so I’m probably missing something.

  19. vadimoss

    It makes me think of The Biggest Little Farm movie. He is a National Geographic style cinematographer, she is a chef living in LA. They got evicted from their rental apartment because of a barking dog they adopted. They decided to start a traditional sustainable farm near LA. Their mentor said in 7 years it will all come together and becomes sustainable. After seven seasons of struggles, perseverance and hard work it magically just happened. Seven years it is!

  20. Chris Albinson

    Building a real business that lasts, not unicorn (meaningless metric), does take time. Our data @breakawaygrowth on the median time to $1B in revenue is 12 years. Doing it with capital efficiency of growth so that all stakeholders do well, not just the VCs, or a certian class of investors with difference share classes is what we focus on. To the extent it is helpful to the conversation:* Only 0.3% of funded companies hit the $1B in revenue where all stakeholders do well * The median time to this outcome is 12 years.Twelve years is problematic for most funds (avg. fund life of 10 years) and for most human beings taking on this very stressful and lengthy building a $1B+ revenue company. Not to mention at least one sometimes two economic corrections during those 12 years. Real Business that LAST, take time to build and having the conviction and the patients & support founders during this full journey critical for all involved.Much love and admiration to: $NN, $ISLD, $JNPR, $SHOP, $DOCU, $PINS, and few on the way!