1/ First, I’m geeking out right now because I’m a huge fan. Let's put that aside. (btw- we have 5 copies of your book in our office and it’s a small office). Love how you describe complicated subjects in easy to understand ways @annieduke
2/ Q1: Defining success- really great pt and I was a bit flippant because @Zachkanter and I were having fun.
Ultimately measured by returns (i.e IRR or MOIC) but here the time frame is too short for that (I’m assuming Zach agrees)
Ultimately measured by returns (i.e IRR or MOIC) but here the time frame is too short for that (I’m assuming Zach agrees)
3/ In 2 yrs success would be:
a) demonstrate that models/factors, upon which the fund is based, are able to predict objective measure of success that correlates “strongly” with IRR.
b) raising a multi hundred million dollar fund
c) begin to deploy fund at projected rate
a) demonstrate that models/factors, upon which the fund is based, are able to predict objective measure of success that correlates “strongly” with IRR.
b) raising a multi hundred million dollar fund
c) begin to deploy fund at projected rate
4/ Obviously 3a devil is in the details, so let's say that if the Information Advantage (IA) must be great enough that across a 150 co. portfolio (N=150), expected return is >3x. Again backed up by robust back tests (out of time, out of sample testing etc)
5/ To (pseudo) reference your book, we’re thinking in probabilities (not certainties) and just trying to increase our probabilities through an Information Advantage.
6/ Not perfect but its best I can do in a Tweetstorm at 11:30pm est :)
7/ Q2: Define quant fund-
The systematic VC equity fund will find and evaluate co's algorithmically. Humans will be involved in calling co's that are identified by the tech and in doing confirmatory dili (i.e. tech “Thinks” the co. is $5m in rev, just verify that # within +/-).
The systematic VC equity fund will find and evaluate co's algorithmically. Humans will be involved in calling co's that are identified by the tech and in doing confirmatory dili (i.e. tech “Thinks” the co. is $5m in rev, just verify that # within +/-).
8/ Human judgement will not be used to identify companies to go after or to make final discretionary investment decisions. i.e. Just because an investor may not like that CEO, not a reason to not invest.
9/ So it’s not pushing a button to make a trade, but it is entirely rules based, even when humans are involved.
10/ So where would this work? I think this works in industries where the business models are very similar and there is a ton of data to evaluate companies. Consumer (CPG) is one example.
11/ In consumer (food, pet, personal care, household products) the biz models all basically the same (different margins). Same game of chess over and over again. Success has many prior examples unlike an industry like tech (before Uber nothing like Uber)
12/ The industry must also have a ton of data. In CPG - can see where a co sells products, how many SKUs, price pt of SKUs, what end users think. And if tracking then can see how those things change over time and how they compare to competitors.
More here: circleup.com/helio/
More here: circleup.com/helio/
13/ So to be fair on #2 it isn’t purely quant. But our view is that even in public markets no quant fund is literally only quant. Question to me is just what the humans do. So the 1,000 people that work at @cliffordasness fund AQR- there is some manual work to be done.
14/ [Note: I'm going to go beyond the Q and share more of my vision for this.]
Someone will think it is crazy. But most likely that person also runs a discretionary vc fund that believes everything else can be systematized except for their job :)
Someone will think it is crazy. But most likely that person also runs a discretionary vc fund that believes everything else can be systematized except for their job :)
15/ Granted this systematic VC fund will have far more manual work than a quant public fund, but that isn’t our bar. Our bar is to be far more effective/efficient than a typical discretionary vc fund (i.e. every vc fund currently out there)
16/ So to make this work, we need an industry that has similar business models and a ton of data. Some tailwinds would help too. I would argue consumer has that:
17/ Then a tech is required that will create Info Advantage. Helps find/evaluate companies. Not perfect but doesn’t have to be. It's like increasing odds at blackjack. If only do 1-3 hands, it doesn’t work. But spread out over a large enough N then it looks really interesting.
18/ You could imagine Growth Signals (i.e. brand, distribution, category, product) and Risk Factors (i.e. retailer exposure, product pricing, channel, econ cycle, theme exposure). Etc. Test signals & risk factors (validate w/ robust back testing), build target portfolio.
19/ Implementing the technology to create a systematic vc fund then is also extremely difficult. Starts with systematizing the sourcing of deals. The tech helps but would also need a b2b calling engine. Summit and TA Associates are comps here.
20/ Now you have sourcing engine and IA in finding co’s. But not everyone is raising and not every co. wants $. Need reason for the co's to take your $. In tech I don’t think that’s realistic because there is massive competition. But in consumer? Not much comp.
21/ But let's assume I'm wrong and there is a lot of competition.
To “win the deal” the systematic vc fund needs to win on key “purchase criteria” that the entrepreneur cares about. I think the fund can be better, faster, cheaper.
To “win the deal” the systematic vc fund needs to win on key “purchase criteria” that the entrepreneur cares about. I think the fund can be better, faster, cheaper.
22/ Better – because it can drive insights via it’s tech. Because instead of hiring generalist discretionary GPs that say they are good at everything it can target specific backgrounds that drive value for those entrepreneurs.
23/ Faster/Cheaper- a systematic fund would get rid of the months of diligence a discretionary investor does.
Close deal from first outreach in 14 days not 140.
Close deal from first outreach in 14 days not 140.
24/ In public markets discretionary investors do months of diligence to go a mile deep (but inch wide). We think same concept in privates. No need for a 50 pg investment committee memo. Saves CEO and funds time.
25/ To close deals quickly - need largely standardized docs and efficient way to price the deal. In an industry where there are agreed upon valuation metrics (i.e. revenue), thats a lot easier.
26/ In public markets discretionary investors do months of diligence to go a mile deep (but inch wide). We think same concept in privates. No need for a 50 pg investment committee memo. Saves CEO and funds time (and $$).
27/ But importantly the systematic vc isn’t trying to source by going to cocktail parties. It isn’t passing on a CEO because he wasn’t in the GPs fraternity. This should eliminate much of (not all) of the human bias found in discretionary vc. Decisions & process made by rules.
28/ It would have the potential to be a scalable, repeatable private investment platform.
I think that would be revolutionary.
I think that would be revolutionary.
