Starting a VC Fund, Episode 4: Built to Last: Constructing Your Portfolio

Hi there. I'm Patrick Evers. I'm the founder and CEO of MindMaven, a executive coaching firm that focuses on helping leaders and venture capitalists free up a whole work day every week and then utilize that time to invest into relationships to help these people reach their fullest potential.

I'm Charles Hudson, managing partner of Precursor Ventures. We're an early stage venture capital fund that invest in companies pre everything, pre traction, pre data, pre launch. We want to be the first investor of record on your cap table.

Awesome. So in one of our previous videos, we talked about getting the story really well developed of your new fund. What we want to do today is take it a notch deeper. If you haven't watched the other one, I would highly recommend watching that first because today we're going to talk about portfolio construction. And.

I always like asking you, like, hey, what do people get wrong about portfolio construction? Because it's often something that people think of in a very simple way and then end up missing some key points.

I think portfolio construction is really critical. Once you've got your basic story down, that's kind of the guidepost of what you're going to do. But your portfolio construction really dives into the nuts and bolts of how you're going to run your fund. It covers all of the really important strategy elements in terms of what stage you're going to invest in, what kinds of companies are in spec for the model, what you expect in terms of returns and outcomes. In a lot of ways, it is the living strategy doc that you use to run your fund, and it's the thing you'll probably spend the most time working on with prospective limited partners when you're going to raise your fund.

All right, and.

What do people usually do? They end up spending more time around.

What things.

I think most people love to talk about how good they are at sourcing, how special their strategy is, how unique they are as a venture capitalist. But at the end of the day, most limited partners really want to engage you with the portfolio construction model because in a lot of ways, in the same way that a business plan or a financial model is what governs how a startup will behave, a portfolio construction model is kind of what governs how a venture fund is going to behave. And so diving into that portfolio construction model really gives LPs a sense of how you're going to actually make decisions about what to invest in and why. And so I think every hour you spend refining that, understanding the implications of what's in model and out of model for you is really time well spent.

All right. And so if I'm a first time fund manager, or at least aspiring one, what's one of the most important things I have to realize about this process?

I think one of the most important things to realize about this whole process is that it is process in the same way that building any model is a process. You'll start with the core set of assumptions. You'll build a model that reflects those assumptions. You'll look at the output of what it says in terms of returns and outcomes, and then you have to make some decisions about, hey, does this fund as constructed actually produce the kind of outcomes that I want to market to limited partners?

And then.

I think in one of the conversations you and I had, you actually pointed out something that I thought was absolutely fascinating. That is that you felt there was a pretty strong asymmetry of information when you actually go out with your portfolio model to those LPs. What did you mean with that?

I think this is actually the place where most emerging managers get tripped up the most. If you think about your typical limited partner, they've invested in tons of funds, so they have access to a lot of information you don't. They have access to lots of other decks that are in market today with people making claims about what they can generate in terms of returns. They also have their own lived track record as a limited partner, seeing what the actual returns were from the funds that they invested in. So they know a lot about both what other people are promising in the ecosystem and also historically, what's been possible for venture capital funds with different strategies and sizes. And if you as an emerging manager come in to present a fund and you don't have access to that same information, it's very easy to make some naive or ill informed assumptions that that will immediately diminish your credibility in the eyes of the limited partner, and that.

Basically will diminish your ability to actually get the money from those people quite dramatically. It's probably one of the worst things you can do.

Absolutely it is.

Yes. Right, okay. And so it becomes very important to.

Show up, to show up with a model that you actually have quite a bit of confidence in that it works. So what advice you have in terms of how that model should look like?

You know, I think for people who are building portfolio construction models for the first time, I would say in some ways it's a fairly straightforward process. There's only a few key inputs that you have to make decisions on. That's usually what stage you're going to invest in or stages. If you're a multi stage fund, how many companies you're going to put in the portfolio, what your average check size is going to be for that first check, what percentage of companies you think you'll be able to purchase with that first check, and how you plan to deal with follow on investments and reserves. Those are basically all the inputs you have to analyze and put into the model. The challenge is going back to this information asymmetry point. Limited partners who've been in the business for a while have a sense of the ranges of what's feasible for each of those things. So if you were to say, for example, we're going to have a portfolio of 200 companies and I'm going to take a board seat for every one, most limited partners would say, well that's impossible. It's too many board seats, not enough people. Like that's not a workable strategy. Even if it's something you could actually put into a spreadsheet.

That makes sense. So what you're saying is you need to make sure that your model falls into some sort of band of consensus of what people LPs think is possible.

And from there basically build it out. So that is what I take away from your conversation here.

That's right. And I think that's one of those areas where if you don't have access to some of that private information or you don't have access to other people who can give you tips, it's an easy, it's an easy area to get really off track.

All right, that's fascinating. So basically, as a new fund manager, you have the same kind of problem that a founder might have convincing a vc. You just got to have a very compelling story and the fundamentals of that story need to align with the worldview of the person you're trying to convince.

Are there ways in which you can tread new paths and if so, how.

So? I do think the good news about a portfolio construction model is it does allow you to flex some of those assumptions and see what happens. And what I would say is I think for lots of these things, lots of the elements of the model, LPs have in mind some distribution of outcomes. So for example, one of the key things you have to decide in your portfolio construction model is what percentage of companies that you invest in at time zero will get to another round of financing.

Let's think about my world, which is really precede to seed. We modeled 60% graduation in the original version of the model and that felt on the high end of good When I talked to LPs, they said, look, we're used to seeing really good seed funds with graduation rates in the high 50s, low 60s%. So if you're in that range, it's credible. If you were to tell me 95% of your portfolio companies are going to graduate from pre seed to seed, you'd either be the world's greatest company picker or you have a model that has overly optimistic rosy assumptions.

Got it. And so basically, if you are outside of those bands, let's say it's something like 70%, you better have a really good story of why you think you can achieve that. And that story should be relatively convincing. It might be that you have a new idea or a new insight into the future of venture funds or something, and it might actually be plausible. But you will have to do a lot of hard groundwork to convince people that you can actually achieve that.

I think the hardest thing is when one of the core assumptions for your fund returns model is out of bounds. So if your whole model, keeping with the 70% graduation model, if the whole way that your fund gets to superior returns is by having 70% of those companies graduate and turning that down to something more like 50% makes the fund ho hum, that's a really critical assumption that you've baked into the model and something I, I would expect limited partners to push on really hard because they're going to say, hey, you're forecasting 70. But if you end up with something that feels a lot more like the mean, is this fund interesting? And if the answer is no, you've got your work cut out for you.

And so many ways, LPs will spend a lot of time talking to you about this portfolio construction model that you have, right? And so the more time you invest into getting that right, the better you'll be office. Is that basically what you're saying?

I think that's right. And I think it's really important that you're fully in command of your own model. And if you think about it, models are built so that they can be flexed. And the reason you, you build these portfolio construction models is so that an LP can say, well, what happens if your graduation rate isn't 65% but it's 50, okay, what does that do to the returns for the fund? Or what if instead of being able to buy 10% ownership upfront, you're only able to buy 8% ownership upfront? What does that mean for your ownership on exit? And is the fund so interesting? And so when you build the Model. Think of it as a dynamic model. It's a model where you're going to want to change the inputs, whether it's the number of companies or ownership amount or entry price, and be able to say, hey, this model can tolerate a 20% increase in average entry price. Or to say, hey, this model really requires that we buy 10% ownership in these companies. And the more you understand your model by playing with it and flexing it, the more prepared you'll be to have those conversations with LPs.

Perfect. I love that. So the obvious question that people would probably ask is, let's say I'm experienced member and let's say the industry, but first time person is trying to raise a fund for the first time. How do I best go about figuring out what the current consensus view is?

I think understanding the consensus is super important if you're doing a fund model at the portfolio construction stage. Because I think you want to know whether your assumptions are inbound or out of bound. So I would say if you're fortunate enough to be starting your own fund, coming from the industry, I would say lean on your peers who are fund managers themselves and ask them to take a look at your model. Ask them to validate your inputs and make sure that you're not way, way out there. Because I think if you can do that, you've got a real advantage relative to the person starting from scratch who has no relationships. Most of this information, it's kind of like tribal knowledge. The people who've built portfolio constructions and raised money typically know what LPs are looking for.

All right, so pragmatically speaking, what are we saying? We're saying go find 10 people that are in the industry either as LPs or as VCs, preferably hopefully people you already know, and.

Go have conversations with them, interview them.

Is that the very pragmatic advice you would have to them?

I think so. I think if you want the biggest bang for your buck though, what I'd encourage you to do is build out your own portfolio construction model.

Mm.

Take that model to the meeting and see if you can get 45 minutes to an hour of their time to review the assumptions. They might even be able to do it for you asynchronously. If you send them the model, they might be able to provide you comments over email or something like that. But I think if you've got the model built out, a trained eye could scan that thing, give you some feedback in 30 to 45 minutes our tops. That would really help you figure out where is the model maybe too optimistic? Or are there things that are embedded in your assumptions that are not in keeping with current market sentiments?

Okay.

So two things basically. One, go talk to people that are in the industry so that you can get at least a little bit of a sense of what's the consensus. And then as soon as you have a little bit of that, go build out the model.

Play around with it, flex it enough, but then also send it out to possibly similar people to see what they in terms of specific feedback for you have. What's another piece of advice you would have for people that are like literally really new, might even want to learn the lingo of the vc, LP language world. What else can one do even before those two steps?

I think before those steps there's increasingly more good content out there that limited partners themselves are creating. And so I'd recommend two podcasts. One is Origins, which is a podcast made by the team at Notation Capital where they interview really top notch limited partners about the way that they think about backing new fund managers. And the other one is Venture Unlocked by Samir Kaji, which I think is also an excellent podcast that is listened to by many folks in the industry and has lots of good insights for emerging managers trying to get off the ground.

Okay, awesome. That's really, really helpful stuff. So we'll try to link that in the comments for anybody who's interested in that. The other thing I want to acknowledge is most of those tips kind of require you having some form of relationships already in place, and if you don't have that, it might feel not as helpful. So I think, Charles, what do you think of the idea of us doing another video at one point where we say, hey, what if you basically are relatively new to that and you don't have a lot of people to tap into to learn what the consensus model is?

For example, that definitely warrants its own video.

Yes, I agree. Absolutely. All right. I think that's all there is to say in terms of what the basics what do most people get wrong about.

Constructing their portfolio model? And I'd say let's spend some time talking about.

What makes a good model in our next video.

Awesome.