Profile picture
Ryan Caldbeck @ryan_caldbeck
, 37 tweets, 8 min read Read on Twitter
1/ Been talking about my vision for a systematic VC fund. Think Renaissance Technologies in the private markets. Here is initial strawman of how it could work. Flow her is industry->tech->outreach->validation-> negotiation->close->post close.

Don’t have all the answers.
2/ Starts w/ stage & industry. Cant have too much competition (not tech b/c 1k vc firms), must have large mkt, enough data (not seed stage), biz models largely standardized so you can generalize training data.
CPG perfect. $15T industry few active VCs.
3/ Need small Information Advantage (IA) + Large Sample (N). IA won't be perfect so must spread out over large population. Said differently- predictions won't be guaranteed, so need to make a lot of bets to capture the value of the IA.

4/ In publics you often see tiny advantage so portfolio for quant fund often X,000 companies. IA in privates can be larger (data isn’t commoditized) so I’m envisioning 150 co. portfolio here. Large for privates - certainly- but there is precedent (GV, a16z, First Round, etc).
5/ There must be data directly connected to co. performance that can be gathered. Data must be on entire industry- not just on co's the firm interacts w/. Also need data on bad cos to classify good cos. Think of industries where there is a ton of data about co's already out there
6/ Data must be combined (Entity Resolution) and algos built on top of data. Algos cant be blackbox only. Machine Learning isn’t interpretable by humans, also ML decisions degrade over time as exogenous factors come in. Models based on economic intuition are important as well.
7/ Algos predict “success” of the co’s (can demonstrate w/ robust back testing). But what does success mean? In publics stock trades constantly- easy to have objective success measure.
8/ Must look for clear objective measures in privates that have very high causation to successful exit and have enough data to act as training data for models (hint: we’ve found them).
9/ OK now we’ve got data/algos that can create an IA. We have a massive market (100s of thousands of cos) and our target N (call it 150). Let’s go hunt.
10/ Outreach. Envision a TA/Summit style b2b sales team. Rather than phone books & trade rags think target lists from tech-driven IA mentioned above.

Step one is portfolio construction.
11/ Tech creates theoretical portfolio (based on growth & risk signals). Team reaches out. Some co's wont take $. Tech must adjust portfolio. Iterative process but based on rules not gut.
Every VC fund in the world does this -each has co's they want to invest in but cant
12/ Risk of comp for best deals higher in some industries (tech) than others.

CPG with $1m-15m in revenue? Massive underserved market. And really hard to play it even though everyone wants in because very inefficient to find the companies.

13/ Outreach can be done effectively through mass customization. Tools like @outreach_io make this pretty easy provided you have data to find right co and customize something compelling. I.e. “We saw you just launched in Whole Food”, etc
14/ IF IA is strong enough, first outreach could be literally a term sheet. There are perhaps branding reasons you wouldn’t want to (i.e. might scare off entrepreneur) but you could. That’s the key- this fund CAN move that quickly.
15/ Validation- When entrepreneur engages, must verify #s. The IA was created from afar. Maybe you think the co. is $5m in rev growing at 100% a year. Must verify within a certain band (i.e. 10% +/-). Confirm margins in line w/ category averages.
16/ As part of Validation step, must review corp docs, run background checks. Relatively standard diligence that takes days not weeks.

More friction than public quant fund for sure, but that's not the benchmark.

Will be much more efficient than VC.
17/ Term Sheet Negotiation- this one trips up the thinking of people in later stage privates (i.e. PE). They envision PE style negotiating over docs for months. In VC, the money just really isn’t made in the docs. In CPG as example, we’ve found standardized docs work really well
18/ Kudos to NVCA for this btw:…
19/ Valuation. In CPG valuation is relatively consistent - doesnt get influenced by hype bubbles like tech does. Also valuation in CPG driven by clear metrics (i.e. Revenue). Firm should know what mkt multiples are, share that data with entrepreneur. Feels fair.
20/ Legal Processing (Closing)- If docs are standardized, legal process should be streamlined. If the fund has enough data on what “market” is for legal terms & valuation, I think that fund should just make the data public. We will.
21/ Analogy here is Kelley Blue Book that allows car buyers to just know what fair is. Founders just want a fair deal. Today few founders walk away from a negotiation feeling like they got fair deal. We want to introduce transparency, and ensure a fair deal gets done.
22/ Post Close Govn/Value Add- A really phenomenal consumer investor said: “We did a study on all of our minority investments and looked at those companies where we had board seats v. those we didn’t. Turns out the performance was no different.” Awesome intellectual honesty.
23/ I tend to be skeptical that VCs add much value post close but let’s assume I’m wrong. Or at least that some LPs and entrepreneurs will think VCs add value post-close.
24/ I think there are several sources of value this systematic VC fund can supply post close. It starts with the technology used to create the Information Advantage discussed here. Community, expert resources, partnerships are also examples.
25/ [This systematic vc fund should win deals at a higher rate than the typical discretionary vc. Even though CPG doesn’t have as many vc funds as tech, this fund will have to be able to “win” the deals and get into the ones it wants at a high enough rate for model to work.]
26/ Co's want access to that data advantage. Let them make a case to get on the shelf of a retailer, help to inform new product launches or pricing strategy, etc. This is also why insights need to be interpretable by humans. Biased examples found here:
27/ Community- First Round, USV, YC have done this well. There are network effects to this group. They are able to support and add value to each other in meaningful ways. IMO this community is extremely valuable. Here is an ex:
28/ Expert resources - think board partners, or a collection of potential board partners, that companies can elect into. Either for “Free” or for nominal fee to make sure they really care. Manual and not scalable. But everything about this doesn't have to be infinitely scalable.
29/ In CPG there are four key levers to success. Team, Brand, Distribution, Supply Chain. It may make sense to put board partners or advisors with these specific backgrounds within the fund and make them available to the companies that select in.
30/ The post-close value add I’m most excited about is providing insights via the technology directly to the entrepreneurs. But I’m sure many VCs will say that these companies “need a lot of handholding”. Really? @RXBAR, @HaloTopCreamery, @abhcosmetics all didnt need that.
31/ Partnerships - imagine a portfolio of the 150 most interesting high growth CPG brands. Would retailers find that interesting? Any recruiting firms give discounts to help that portfolio (knowing the job would be easier)? Pt is - a lot of potential interesting partnerships.
32/ Exit- I’ve always been in the camp- as an entrepreneur and investor- that the entrepreneur leads the exit. That’s effectively what happens with most large Seed Funds.
33/ But let’s assume I’m wrong. If the process and tech mentioned here is working (selecting great companies), the fund is building a reputation for investing into the [50-100] most interesting CPG companies each year. I’m going to bet getting intro to buyers wont be tough
34/ Worst case? Hire 3 full time investment bankers to help shepard the companies when necessary.
35/ So what have we built here? A scalable, repeatable VC fund. One that can help thousands of entrepreneurs to thrive over time by giving them the capital and resources they need. Also one that will have incredible barriers to entry (unlike most public quant funds).
36/ But......gathering, normalizing, linking the data, building algorithms on top of it. Then building a finely tuned operations machine to take advantage of that IA and invest into a large N of companies.

Will be hard.
37/ If you don’t buy it, congrats.. lots of people have been in your position throughout history. There is no doubt it will be hard. But everything worth doing is hard. And in this case it can completely remake private investing while fueling innovation for a massive industry.
Missing some Tweet in this thread?
You can try to force a refresh.

Like this thread? Get email updates or save it to PDF!

Subscribe to Ryan Caldbeck
Profile picture

Get real-time email alerts when new unrolls are available from this author!

This content may be removed anytime!

Twitter may remove this content at anytime, convert it as a PDF, save and print for later use!

Try unrolling a thread yourself!

how to unroll video

1) Follow Thread Reader App on Twitter so you can easily mention us!

2) Go to a Twitter thread (series of Tweets by the same owner) and mention us with a keyword "unroll" @threadreaderapp unroll

You can practice here first or read more on our help page!

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just three indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member and get exclusive features!

Premium member ($30.00/year)

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!