E203: How Elite Endowments Invest w/John Felix

This episode features John Felix, General Partner & Head of Research at Pattern Ventures, a specialist fund-of-funds focused on backing small venture managers in the $5–50M range. We talk about the endowment principles that shaped John’s investing mindset, how to separate true specialists from résumé-driven narratives, why access and selection are two very different games, and the traps LPs face in co-investments. John also shares lessons on reserves strategy, portfolio construction, and what allocators consistently overlook when evaluating emerging managers.

Episode Transcript

(00:00):
So right out of school, you spent two years atWashington, St.

(00:05):
Louis, one of the most storied investmentoffices in the country.
What did you learn during those two years atWashington?
So those are my first two years in theprofessional world.
And I had interned throughout college at ahedge fund, but, you know, everything was still
kind of very theoretical at that point fromwhat I learned in undergraduate business school
and WashU was really drinking from the firehose for two years.

(00:29):
So I'd say the first thing I learned was theart and the science of investing.
I started covering public markets when I was atWashU, and we invested with long only public
equity managers.
We tended to favor very concentrated managersthat did deep fundamental research, you know,
very long term time horizon, kind of like thatprivate equity style investing in public

(00:49):
markets.
And so I got a great training and just likereally deep fundamental research.
My managers there, my bosses started a bookclub.
This was a great crash course in learning fromsome of the greats.
And so we read Ben Graham and Philip Fisher andthings like margin of safety and zero to one.
The other thing that I learned was abouttransition.

(01:09):
And I saw three different chief investmentofficers during my first two years at WashU's
endowment.
I joined during the kind of tail end of thelongtime tenure of Kim Walker, who had been
there for almost a decade.
And then when she retired, Eric Eupin, who hadactually run Stanford's endowment for a period
of time and then later went on to run McKennaCapital Management, he was on our board.

(01:32):
He stepped in to kind of steward the endowmentwhile we looked for a full time CIO.
And Eric was there for about a year, kind ofjust steady as she goes.
You know, we didn't do a ton of crazy stuff.
And then we ended up hiring Scott Wilson, whocame on and is still running the show now and
has done an amazing job, but three differentchapters during those short two years.
And so I learned a lot about just resiliencyand change and, like, just being in a workplace

(01:55):
where people come and go.
But it was amazing and really, like, learned alot during the different kind of three chapters
over those two years.
And you're around some of the best fundamentalpublic investors at WashU.
Give me a sense for, do you believe that themarket is efficient, whether in the long term,
in the short term, and how do you look at thepricing of public assets?

(02:19):
That's a good question.
And this is something that, you know, inundergraduate like business where you learn
about this efficient market hypothesis and youlearn about all of these like theoretical
things.
And you're taught in school that a lot of thesethings are like how the real world works and,
you know, you're kind of taught them as gospel,but then you get into the real world and you
realize that nothing is black or white.
And I think one of the biggest things I learnedis that the market is often not efficient in

(02:43):
the short term, or at least that's my belief.
And I think that lack of efficiency in theshort term creates longer term opportunities
for investment.
So the way a lot of our managers thought aboutit, the way I think about it is in the short
term, the market can be inefficient and presentwhether it's pricing irregularities or just
areas for really good long term investmentopportunities.

(03:06):
And I think in the long term, at least myphilosophy is that over the long term, the
market is efficient, but, you know, it mightnot be efficient at any given time.
But if you buy something that's mispriced or,you know, compounding over the long term,
eventually that price should, know, eventuallyreflect the underlying fundamentals of the
business.
It could take a year or it could take fiveyears.
But over a long enough period of time, thosethings should converge.

(03:30):
Eventually the trading catches up to intrinsicvalue of the asset.
Yeah, at least that was how we thought aboutit.
And that's how I think I still think about ittoday.
But again, it's hard to predict that timeframeand it's hard to predict like all of the, you
know, incremental, like inefficiencies thatmight happen in the interim.
You spent some time under Scott Wilson, who'sconsidered one of the top CIOs in the country.

(03:51):
What makes Scott such a great CIO?
Scott is very different than I'd say a lot ofthe traditional archetypes for endowment CAOs.
So, you know, the first thing is Scott had avery nontraditional background.
He was not trained by David Swenson or hedidn't kind of grow up in the endowment world.
He was actually a rates trader and spent a lotof time in Tokyo sitting on rates, you know,

(04:12):
desks, and he had kind of a quad background.
And I think this was actually a really goodthing for him because it allowed him to come in
and take kind of a first principle approach toendowment investing, right?
He wasn't influenced by the way things had beendone for many decades beforehand, And he was
able to kind of build his own approach.
And so the first thing I would say is like, hedisregarded convention completely.

(04:33):
You know, there was kind of a standard way,this endowment model of managing a portfolio
that most other endowment heads had embraced.
And by the way, that that that had workedreally well for a lot of endowments.
But Scott didn't really care what other peoplesaid or did and kind of built his own approach.
One of the the tenants of, I think, like Scottas a CIO was he is just he's completely

(04:57):
ruthless when it comes to making managerdecisions.
And I don't mean that in a bad way at all.
I think that's actually like absolutelycritical to being an elite investor, but it's
really hard for a lot of people to do.
There's a lot of psychological headwinds inthis industry.
If you're invested in a manager over the courseof many years or even a decade, you become
friends with the manager.
There's a lot of close relationships.
And I think it gets really hard to, you know,get off the train.

(05:19):
And I think a lot of institutional investorsare bad at putting in redemptions, are bad at,
you know, not re upping in funds that theyprobably should get off the train.
And Scott was really good at that.
And so the first thing we did when Scott joinedwas we went through every asset class in our
portfolio and we force ranked every singlemanager in every single asset class.

(05:40):
There could not be a tie.
There had to be a one and there had to be, youknow, whatever the last number was, right?
And what that did is, is it forced you toacknowledge and to admit like, who are your
really high conviction managers, maybe weshould give them more money, but also, who are
the low conviction managers?
And like, why are they still in the portfolio?
And I think it was kind of close to, you know,Jack Welch, who is a longtime CEO of GE, every

(06:03):
year, they would look at their workforce and,and they would fire the bottom 10% of their
employees, right?
I think Scott took like a very similar approachthat if something was in the bottom, you know,
quartile or 10% or whatever was of ourportfolio, why is it there?
And why do we still have capital with thatmanager?
And every day we continue to have capital withthat manager, was almost making a conscious
decision that we wanted to be invested withthat manager.

(06:24):
And so he really, I'd say, forced us to look inthe mirror and take a hard look at our
portfolio and make those tough decisions and doit in a really
way.
It accounted for this consistency bias where bydefault, you wanted to re up in a manager.
But here, his system was more an opt in whereyou had to force rank and you had to

(06:45):
consciously make a decision who to cut.
You didn't need a reason to cut somebody.
You need a reason to keep somebody in.
Exactly.
And I think that's something that a lot of LPsstruggle with.
Look, it's it's it's really hard.
Right?
Like the default every morning when you wake upis you have an existing portfolio already.
And so like, let that existing portfoliocontinue.
And I think Scott's view is kind of theconverse of that, which was, you know, like I

(07:09):
said, every day you continue to have capitalwith that manager, you you should view that as
like you're making a conscious decision to reunderwrite that manager and have capital with
them today.
And if you would not invest with the managertoday, then why would you still have capital
with them even if you made that decision?
The zero based budgeting type of framework.
It's the same with employees.
If you wouldn't hire them again, why are theystill at your firm?

(07:31):
Exactly.
Lack of friction.
Yeah.
And it's it's such a simple concept to tounderstand, but I think it's it's it's really
hard to actually do it in the right way.
And I think a lot of LPs still get this wrongpretty consistently.
How do you balance being very opt in consciousabout your portfolio versus what people call

(07:52):
relationship alpha or the alpha that comes fromhaving deep relationships with managers?
After Scott joined Washington Dammit, he did apodcast actually a couple of years later, and
he talked about how they ruffled a lot offeathers in those first few years.
And, know, I think a of managers were unhappywith Scott, but, you know, that was him coming
in with a clean slate and wanted to rebuild theportfolio from scratch.

(08:14):
When, you know, Scott decided to commit to amanager and re underwrite a manager, he would
set expectations with them upfront, like verymuch like, hey, this is why we're underwriting
you.
This is what you said you're going to do.
And if you do it and you do it well, we willcontinue to be a great partner with you for a
while.
But if you don't, you know, like this is abusiness and we're going to move our capital
somewhere else.
And I think just being really objective aboutit and not being personal about it and setting

(08:38):
expectations upfront is really important.
And if you're a rational investor, a rationalhuman, rational business person, like you
should understand that those decisions, whiletough to make, like, are often the right
decisions.
And I think, like, the right managers did nottake it personally.
And from my understanding, Scott is veryfocused on finding the best in class athlete,

(09:00):
the best managers versus kind of trying to fitsomebody in a box.
Talk to me about that strategy.
Yeah.
So there's a few components of that.
So the first thing I would say is Scott veryheavily favored concentration, both in as we
thought about the number of managers in thetotal endowment as a whole, but also as we

(09:20):
thought about positions within our underlyingmanager portfolios.
And so, you know, we called our manager rostera ton when Scott joined and really tried to,
you know, concentrate capital and like our topmanager ideas.
But then going a layer beyond that, Scott wouldalso want to concentrate in underlying
positions within those managers.
And so the way like Scott's general philosophyas a whole was he didn't really care so much if

(09:47):
an opportunity came from the public equitysleeve, the private equity sleeve, the real
estate sleeve.
Like he took effectively this opportunisticapproach where at any given time he wanted to
concentrate capital in the best ideas,regardless of the asset class, regardless of
kind of the end exposure.
And so we would actually like spend time withunderlying managers.

(10:08):
And they started to do this a lot more after Ileft.
I mean, Scott kind of joined at, you know, thetail end of my two year analyst program there.
But we started to do this And we would meetwith our managers, we would go through their
portfolio, we would talk to the managers abouttheir highest conviction names within their
portfolios.
And then oftentimes we would do our ownresearch on those names.

(10:28):
And if we agreed with the manager where we hadoutsized conviction in a certain name, we would
often increase our exposure to individual nameswithin our portfolio.
This was something that a lot of otherinstitutional investors were not doing at the
time, right?
Like they're, you know, the traditionalendowment model is identify best in class
managers who are great at their domain, givethem capital and effectively like outsource

(10:52):
your decision making to them.
Scott had this perspective where we could doour own really good research.
Not that we thought we could do research betterthan our managers, but we focused on like
incremental research, incremental insights,leveraging their research and conviction and
perspectives to maybe build a little more of anincremental perspective that would give us the

(11:13):
conviction to, you know, to increase exposureto those individual names.
And that concentration, you know, has hasreally been huge tailwind to to Waukeshi's
portfolio since since Scott joined.
You could think about it almost as acounterbalancing factor to growth of AUM of
some managers.
So if you think about it from first principles,the reason that returns go down as fund sizes

(11:38):
go up is because deal flow does not scale inproportional proportion with the increase in
AUM.
So let's say that the AUM goes up by two x, butthe quality of deal flow goes up by 50%.
So that incremental, you know, fourth of theirportfolio is lower quality than it would have
been in the previous fund.

(11:59):
And said another way is GPs will the GPs willnever admit this, but GPs have a grading system
for every single opportunity.
Now it might be a plus plus, a plus, a, aminus, b plus, and and may maybe no opportunity
ever goes below b plus.
Let's just let's just give the benefit out.

(12:19):
But there is a grading system.
So by actually going double clicking, havingthe relationship and having the FaceTime with
the managers, you could actually figure out,you could double down on those A and A plus
opportunities.
Even forget about co invest and lowering offees that obviously moves the needle a lot as

(12:40):
well.
But just by re concentrating your portfolio,you could actually get higher returns than the
underlying fund if you do it right, which ishuge asterisk.
Yeah.
Yeah, exactly.
And like, when I think about the traditional coinvestment model, I think it's more like
reflexive and reactive and a lot of LPs, youknow, they'll wait for a manager to come to

(13:01):
them with an idea and the manager might saylike, Hey, like I've got excess allocation, you
know, in this company that's raising a newround.
It's we have really high conviction.
Do you want to participate?
And LPs might often participate in that.
Scott was very proactive about this, right?
Like we would identify companies, right?
And we would spend a lot of time buildingthose, building our own conviction in those

(13:23):
companies.
And, you know, to your point, like, yeah, if wedo a good job of that and WashU has done a good
job of that, like that adds a lot of alpha tothe portfolio.
And, you know, you could also like take a stepback and say, well, like if managers had such
conviction in their top ideas, why not justmake, you know, size those bigger in their
portfolio?
We were already investing with veryconcentrated I mean, you know, the managers in

(13:47):
our long only public equity portfolio, theywould sometimes already have 10%, 15%, 20%, you
know, sometimes even more than that, 25% intheir top name and often 50% to 75% of their
portfolio in their top five or 10 names.
Right.
And so they were kind of maxed out onconcentration.
But we really thought like, if you have suchconviction in these names, why shouldn't we

(14:09):
have more exposure?
And actually one of the projects, one of thelast projects I worked on at WashU before I
left is we looked back at all of our publicequity managers going back, I think ten years.
We looked at the performance of theirportfolios as a whole.
And then we compared each of their, you know,ten year performance periods to the performance

(14:31):
of their top one, two and three single names intheir portfolio.
And almost always those top names outperformed.
It wasn't always like the top one name thatoutperformed the broader portfolio, but it was
usually either a combination of like the topone, two or three or, you know, some
combination therein.
And that really gave us the conviction thatlike managers are generally good at sizing, you

(14:52):
know, higher conviction bets better.
And that I think also gave us the conviction toexecute on the strategy where we would buy more
their top names.
Said another way, you could be very confidenton investment, but would you stake your entire
career on it if you went over the ledge?
Or would you stake Lil Johnny's college tuitionon it?
So, you know, asking managers to take a careerrisk is is such a high standard, and there's

(15:17):
somewhere between that and the averageinvestment in the portfolio, there's this
alpha.
So you might be very convinced on it, but notwilling to stake your career on it, and there's
still that conviction is still valuable, avaluable LP insight.
Yeah.
Exactly.
And like, you know, look, if we, you know, weyes.
We had the conviction in in, you know, everyposition that we put more money in, but we

(15:39):
would sometimes get it wrong.
And and I think the way to mitigate that riskwas to still have exposure to the manager's
broader portfolio, right?
It wasn't like we wouldn't invest in their, youknow, number two to 10 idea.
We would still have that through our core LPcheck-in their fund.
We would just increase exposure to some oftheir best ideas.
It's like a complement that strategy.
Yeah.
I'm still struggling to understand how do youoperationalize a best idea strategy within the

(16:03):
context of a endowment portfolio?
So how do you go about taking that strategy andturning into practice?
It's a really good question.
So the first thing I would say is, and thefirst thing Scott did was prior to Scott
joining, we were siloed by asset class.
And so I worked on the public equity team.
We had a hedge fund team and we had a privatemarkets team.
Right.
And anybody working within one of those teams,all you did was work on that asset class.

(16:28):
The first thing that Scott did when he joinedwas he got rid of that structure and everybody
became a generalist and almost retrained andrewired us to think about not only finding the
best ideas within, you know, a single assetclass, but at any given time trying to find the
best ideas in general, right?
And so we thought about this idea of like,everything should be competing for capital

(16:51):
against everything.
And at any given time, let's be opportunisticand flexible to put capital in the best idea
regardless of asset class.
And so I think like decoupling, you know, anyindividual from one asset class just gave us
like the broader mental space to not onlythink, you know, again about like the best
ideas within a constraint, but like the bestideas in general.

(17:14):
The other thing we did, and they started to dothis a lot more after I left, but the WashU
team is on the road traveling all the time.
They go meet with not only managers, butthey're meeting with management of portfolio
companies like constantly.
They're touring factories, they're talking tocustomers, they're doing like that really deep
primary research that a lot of other endowmentsdon't do.

(17:37):
And I think to execute this strategy well ofreally forming an opinion on individual assets,
like you can't do that without traveling orwithout doing this level of primary deep
research that the WashU team does.
Double click on that.
So let's say I'm backing Bill Ackman inactivist investing, and he's I'm sitting with
him and he's saying, is my one or two bestideas.

(17:59):
WashU would also go and essentially visit thesecompanies.
Let's say they're private companies forsimplicity.
What are you trying to do?
How are you trying to get better informationthan a Bill Ackman or choose your top quartile
manager?
So we didn't necessarily think about it astrying to get better information than the
underlying manager.
Right?
A lot of it was we started from a place of onevalidating the underlying manager because as an

(18:23):
LP, you're really two degrees removed from theunderlying asset typically, right?
You're one degree removed from the managerwho's then another degree removed from the
underlying asset.
So a lot of LPs take whatever a manager says atface value and you have to believe them.
But WashU did not ascribe to that.
And I think we were like very much we startedwith, let's re underwrite the asset and build

(18:44):
conviction in the manager's thesis, right?
So almost like testing what the manager toldus.
And then from there, seeing if there was anyincremental value or insights that we could
add.
But we definitely did not think that we coulddo a better job of underwriting the asset than
the underlying manager.
It was like doing the work to kind of reunderwrite it, understand their thesis, borrow

(19:06):
a lot of conviction from them, and then see ifwe could add anything incremental.
Another way
to look at it is co investments.
There's a lot of pressure to give coinvestments to large institutional investors.
If you think about that from a first principlesperspective, that pressure could lead to bad
incentives and offering the incremental coinvestment that maybe is the average investment
in the fund or maybe even sub sub average, butthere's so much pressure to offer that that the

(19:31):
GP may have no choice but to give kind of a subaverage opportunity to LPs.
So you have to re underwrite that from your ownperspective just to make sure that truly is a
great opportunity.
Yeah, exactly.
And I think to your point, like in the coinvestment world, the biggest problem that LPs
deal with is adverse selection, right?
You have to understand, well, if this is such agreat deal, like why am I seeing this?

(19:53):
Right?
And I think that problem is more prevalent inthe private markets where things are capacity
constrained and less prevalent in the publicmarkets where anybody can go and buy a security
in the open market.
But we did a lot of public markets, you know,almost co investing as well, where we would go
and buy more of a stock, typically through anSPV that the manager set up to align incentives

(20:14):
where they would manage the asset.
But it was, again, it was much more of thatproactive approach where we weren't waiting for
managers to send us an idea where, you know,who knew what the incentives were like, we were
really going to them and saying, hey, let'slet's talk about your best ideas today.
And even if you weren't planning on likeactioning on them, let's create a structure
where we can increase our exposure and alignincentives at kind of the appropriate

(20:37):
So after WashU St.
Louis, you went to Bowdoin College, soconsiderably smaller endowment, but still
today, couple billion dollars.
What was the difference in Bowdoin versus WashUSt.
Louis?
There were a lot of differences in somesimilarities, but both had done exceptionally
well over a long period of time.
And, you know, the first thing I observed atBowdoin was just excellence.

(21:02):
I joined at the end of Paula Valens tenure.
She had spent twenty years at Bowdoin and grewthe endowment from something like 400,000,000
when she joined to over $2,000,000,000 that wasnet of spending, net of paying out
distributions to the college.
So just an incredible track record.
And I learned a few things from Paula.
The first was of going back to this Buffettadage of like, be greedy when others are

(21:25):
fearful.
And she aggressively leaned into venturecapital in the early 2000s after the dot com
bubble when I think a lot of LPs had still hadscar tissue and trauma.
You know, she wasn't at Bowdoin before that.
And so in a way, was like fortuitous that shejoined after this and had this amazing
opportunity to invest, like without the legacybaggage of, you know, major markdowns and that

(21:47):
trauma from that bubble popping.
But she was really greedy during that time andit worked out really well.
And she was early in a lot of amazing Sequoiafunds and other great managers.
So like definitely learned to be greedy whenothers are fearful.
One big difference between Bowdoin and WashUwas Bowdoin was a lot more diversified than
than WashU under Scott.

(22:08):
So, you know, Scott had this reallyconcentrated approach.
Bowdoin had a lot of line items, a lot ofmanager line items, and Paula wasn't afraid to
kind of try things out.
And if it didn't work, you know, redeem capitaland kind of move on.
And so you know, what that resulted in was amuch more diversified portfolio, but still a
portfolio that generated exceptional returns.

(22:29):
And so this, you know, kind of made me questionconvention a little better, but I learned at
Scott, and I think it made me realize thatthere are more than one ways to win in
investing and, you know, it worked for somebodywho might not work for somebody else.
And another kind of good example of that was,you know, Scott was really allergic to macro

(22:49):
managers, right?
Managers that did not do fundamental research,but took more of a tops down approach.
And Bowdoin, especially under Paula, reallyleaned into that type of manager.
Stan Druckenmiller, who's a Bowdoin alum, was,you know, on the board and investment committee
and obviously famously ran Duquesne, which haddone really well and was probably one of the
best macro shops ever.
And Paula leaned into that, you know, reallyleaned into his conviction.

(23:13):
A lot of his network folks that had worked ateither for him or at Soros and spun out.
And those managers had done really, really wellfor Bowdoin over a long period of time.
And so again, challenge like some of myconvention that, you know, those didn't have a
place in a portfolio and made me realize thatactually they can't, those type of managers can
do very well.

(23:33):
And again, goes back to this like one size doesnot always fit all kind of approach.
What would you call Bowdoin's investmentphilosophy?
How would you explain it?
I would say very manager driven.
And so, you know, we had somewhat lessopportunistic than WashU in the sense that like
we did have asset allocation targets andparameters, and we did view our portfolio more

(23:56):
in buckets, but it was very network driven.
And so we would, you know, Bowdoin would get indeep in a network, whether it was Stan's
network or a venture network, and just continueto compound within that network.
And so being early to venture, you know, andinvesting in Sequoia's funds in the early 2000s
meant that we got really close to and this wasbefore my time, but, know, Paula got really

(24:20):
close to the Sequoia team, but really deeprelationships and venture and then would back
like people that came out of those networks.
And those networks compounded over those twentyyears.
And so we ended up being really early to a lotof great new funds that launched between
02/2015, funds like Founders Fund andAndreessen.

(24:41):
And I think like being early and like reallyleveraging those networks to compound was core
to Paula's philosophy.
Like she really took advantage of thosenetworks that she built to get really good
access.
And that was like where I first really like,Scott was much more about the individual
manager and like underwriting them and beingopportunistic.
And Paula was very much about like, we havethis amazing access to these amazing networks.

(25:05):
Let's use out for all we can and like use thatto get access to really interesting things.
That's almost like founder product fit.
They both played to their strengths and reallyleaned into their strengths and trying to kind
of fit their personality to a strategy.
I think that's definitely right.
And again, it goes back to my, this wholephilosophy that like different things work for

(25:26):
different people, right?
And what works really well for one person mightnot work for another person.
And then you went to Allocate, which just fulldisclosure, I was a seed investor in the
platform because I believe in it, obviously.
What was your role at Allocate and what aresome of the lessons that you learned there?
So I joined Allocate right after it had raisedits seed round and I effectively launched and

(25:50):
let Allocate's emerging manager platform.
So Allocate is, you know, it's a startup,raised venture capital and it's building an
alternative investment platform with a coupledifferent parts.
On one end, you have this asset managementplatform where we would, you know, we had fund
of funds that we manage, but we also would getaccess to really good venture funds and
effectively syndicate out that access.

(26:11):
And then on another end of the platform, webuilt software for the private markets.
And, you know, it's funny, the reason I joinedAllocate in the first place was by this point
in my career, had been I'd kind of become aventure specialist.
I moved away from being a generalist and Ireally wanted to concentrate venture.
But I thought that, you know, still being an LPand venture, you're always two degrees removed

(26:31):
from like the real action.
And I thought that if I worked in a startup andsaw that zero to one phase, it would give me a
lot more empathy for my managers, but it wouldalso just make me a better investor and make me
able to ask better questions.
And so, you know, we were, even though I kindof managed the fund of funds and acted like an
institutional investor, we were still doing itwithin the context of a startup and we were

(26:55):
breaking conventions and really like doingthings from a first principle standpoint.
We didn't have to like do things the waythey've always been done, which is like the
worst thing you could ever mutter within thecontext of a startup.
And so we backed a ton of emerging managers,you know, when I was there.
We also backed a lot of non emerging, a lot ofbig platform firms.

(27:15):
And I learned a lot about selling and salesand, you know, things that you didn't have to
to do within the context of like a a a non forprofit endowment.
I often think about with how I invest, I thinkabout how do people become better investors.
And one of the things that's really importantis to get ground truth in different industries.

(27:36):
How could you be a great endowment investorwithout understanding without understanding AI
and what's driving that market?
How do you become a good venture investorwithout understanding what drives startup
returns?
I think it's really important, especially ifyou're focused on one asset class to go as
close to ground truth as humanly possible inorder to be better and be able to go back

(27:59):
10,000 feet up and be a better investor.
Yeah, absolutely.
And like, I had never seen how the sausage getsmade within a startup.
I didn't know what it took to scope outbuilding a software product, working with an
engineering team and a product team andactually building and shipping something, you
know, these were things that were like verymuch in the abstract for me and still, I think

(28:20):
very much in the abstract for a lot ofinstitutional investors.
But being an allocate and like seeing that zeroto one, seeing what it takes to hire a team and
go from 10 people to 50 people, seeing what ittakes to raise a series A.
You know, these were all things that happenedwhen I was there.
And just seeing it upfront, I think gives me,it just gave me a lot better understanding and
appreciation for like, how hard these thingsare, like how hard it is for founders to build

(28:46):
companies, how hard it is for investors to pickcompanies and understand companies.
And it just gave me a lot more empathy, Ithink, for what like all of these players in
the ecosystem have to go through.
And and I'm I'm really glad that I did itbecause I think it's made me a better investor.
So then you went on to found a fund of fund,Pattern Ventures.

(29:07):
Why in the world do we need another fund offund?
It's a good question.
And my somewhat snarky response is like theworld doesn't need another fund to fund
actually.
And I tell this to managers we meet with, theworld doesn't need another venture fund.
But if you have the right to exist, then Ithink there's always room for more.

(29:27):
And I think we earned the right to exist for afew reasons.
You know, in Ventura, if you want exposure tothe big brand name firms, the Lightspeed, the
Sequoias, the Andreessens of the world, It'sreally purely an access play and you don't need
a fund of funds to do that.
You can try to get access directly or you canleverage a platform like Allocate to get

(29:51):
access.
Like there's ways to get access.
It's not really like a diligence or anunderwriting game.
It's an access game.
And I think it's becoming increasingly easierto access these bigger firms.
But in the emerging manager world, it's muchless of an access game and it's much more of
like a discovery and diligence and managerselection game.
And it just takes a lot of time.

(30:12):
There's thousands of venture funds in theemerging manager world to pick from.
You need the right network to find these funds.
You need the right skill set to diligence them.
You need to reference them.
It takes a ton of time.
You know, return dispersion in this world andthe emerging manager world is massive.
And so manager selection matters even more.

(30:35):
And that means you need to spend a lot of timemeeting with a lot of managers to pick the
right funds.
And a lot of LPs just don't have that time ordon't have the bandwidth or the skill set to do
that.
And so, you know, to do this part of themarket, right, I think a fund of funds makes a
lot of sense.
And, you know, I've been doing the emergingmanager, you know, LP investing for a while.
My two partners who started Pattern have beenfund of funds investors for multiple decades,

(31:00):
and they've been investing in venture companiesfor multiple decades.
And so I think we've earned the right to dothis, but I don't think the world again needs
another fund of funds unless you have earnedthe right like I think we have.
Another way to
look at return dispersion.
There's obviously just beta and market beta,but I would actually call it LP error.

(31:22):
I think there's significant LP error in theventure space.
Said another way, the return dispersion is onlypartially due to actually just variance or
randomness.
A lot of it is actually due to picking thewrong managers.
I caution people to really think criticallyabout their venture exposure.

(31:43):
First of all, it's completely idiosyncratic toevery other asset class on the planet.
Mhmm.
There's no learnings.
In fact, if you know a little bit or maybe evena lot about private equity, it's a recipe to be
a terrible venture investor because of powerlaws versus downside protection and and things
like that.
The second aspect is there's a right way and awrong way to enter any asset class that you'd

(32:07):
have no experience with.
So if you look at how endowments orinstitutional investors enter asset classes, at
least most of them, the humble ones, they'llstart actually with a fund to fund.
They'll start with the the most high viewedkind of 10,000 square foot view, then they'll
start to see what does look good look like.
Maybe the fund to fund will start giving directexposure into the funds.

(32:29):
So they get kind of this this sense for whatdoes good look like.
It's almost impossible in the first year toeven know what an asset what a good asset looks
like.
And then maybe five, ten years later, they'llstart to actually build out a direct platform
into the actual underlying companies.
And if you look at the behavior of seventy,eighty, 90% of high net worth investors when

(32:52):
they go into venture, they actually go theother way.
They go Yep.
Directly into startups because that's whatnaturally comes into them.
Then they go into funds if they kind of learnfrom their mistakes.
And then once in a while, they'll go into fundof funds.
But, again, the the LP error on the managerside and even more so on the startup side, that
that LP error, because there's almost nobarrier to entry start a startup, is so massive

(33:16):
that whatever investors save in fees, they pay,you know, five, ten x oftentimes.
Oftentimes, they just lose all their money.
It's like, do you wanna pay 2% management feesor do you wanna pay a 100% mistake fees?
Anytime you're going into an a new asset class,even if it's for the first year, maybe two
years, going with a fund to fund and learningthe ropes on somebody else's dime, I think, is

(33:40):
so valuable.
I completely agree, and it's it's the crawl,walk, run approach.
Right?
And it's funny, I made this mistake earlier onin my career, actually, after WashU, I joined
an OCIO where we manage endowment capital.
When I joined, we had almost no ventureexposure.
And I kind of took the lead on building out ourventure program.

(34:01):
And we were starting from zero, and this waswhen my venture network was very, very limited.
And we basically spent a year trying to justget up to speed and understand.
And then we started making a few commitmentsthe following year.
But I think in retrospect, like I should havestarted with a fund of funds to help me get to
know the ecosystem and build relationships andthen maybe co investing alongside the fund of

(34:25):
funds directly in some managers as you buildkind of conviction and then eventually going to
do it myself.
We could have jump started the program a lotbetter.
And like, we made some mistakes early onbacking managers that, you know, in the early
days, everything sounds great, right?
When you're and even now when I'm, you know,training like our analyst or intern, you in the
beginning, everything sounds great andeveryone's interesting.

(34:45):
And by the way, like almost every manager wemeet with, even if we pass, they're all very
impressive.
They have great experience.
They have great backgrounds.
They worked at a great company or they workedat a great firm.
But it's really hard to delineate between who'sjust okay or good and who's really great.
And, you know, we've had LPs come to us and saylike, hey, look, I tried to do this myself and

(35:06):
I just got really burned and I just need help.
Right.
And I don't think it has to be like an eitheror, right?
Like, you know, you can use a fund of funds anduse it as an educational experience to start
doing this yourself, which a lot of our LPs do.
And so I definitely am a big believer in crawl,walk, run approach.
I made the same mistake, which is why I sointimately understand it.

(35:29):
I think I probably did a 100 startupinvestments before even investing in a manager.
So I did not take my own advice.
Had I known it, I of course would have done it.
The I wanna double click on something that yousaid that's rather obvious for people in the
industry, but if you made your money in awidget factory and then you start investing in
venture, the average, let's call it the bottomquartile venture investor is the top 1% of

(35:54):
society or Yeah.
The top 10% of Ivy League.
So the bar to even have a venture fund or raise$20.30, 50,000,000, a 100,000,000 is so high
that even just to be in the game, there's aquality bar.
That's why you have to meet like a 100 managersbefore you actually know what good looks like.
It's because everybody actually looks prettygood.
That's a common mistake that people make.

(36:15):
Yeah, absolutely.
Like, everybody's smart in this industry,right?
Like when we talk about the managers we like,it's, you know, we're not saying things like,
oh, they're so smart, like we need to invest,right?
Like those are, that's table stakes.
And I actually think this is what some managersget wrong about their differentiation, right?
Because we, you know, we'll talk to some newmanagers and I'll ask them like, why do you

(36:36):
think you have the right to exist?
Or why do you think you have the right to win?
And they'll say something like, oh, well, youknow, we're operators and we have great
operating experience and founders want that.
And by the way, that's great.
And maybe you do have great operatingexperience.
But I've met with 200 other firms that also,you know, were operators of great companies.
And that in and of itself is not enough, right?

(36:57):
It's impressive and it's impressive when thesefolks worked at great companies and did great
things.
But it's almost table stakes and it makes it alot harder to really differentiate again, if
you haven't met with hundreds of firms and youdon't know like where that relative baseline
kind of is.
So for your fund, you're targeting sub$50,000,000 managers.
Tell me about your strategy and your portfolioconstruction.

(37:19):
Primarily what we did is $50,000,000 funds.
The reason why we settled on this is, well,there's a few reasons.
One, when my partners were setting up Pattern,they went back and looked at every fund going
back about forty five years, I think to 1980until today, and looked at every 5x net venture
fund in that period.

(37:41):
And what we noticed was the largest cohort of5x returning funds was in the sub $50,000,000
range.
This is for a few reasons.
One, smaller funds are less competitive againstthe bigger incumbent firms.
You know, you don't have to compete for thatlead spot.
Leading around is binary.
You either win the lead position or you don't.
You know, sometimes you can co lead.

(38:01):
But if you have to, you know, have a biggerfund and you have to deploy more dollars in
lead, now you're becoming competitive againstnot only the big, the other seed funds, but
also the, you know, the big incumbent kind oftier one brand name firms that occasionally do
see it as well.
And so we like that smaller funds are lesscompetitive.
They can get better access to deals.
We also like that they're purely focused on preseed and seed where we think the, you know, the

(38:23):
biggest opportunity for outlier power lawdriven exits exist.
Valuations are better.
And the fund math, most importantly, to get toa good return, it's just a lot easier, right?
You don't need to rely on the next Uber just a3x refund.
If you're a $50,000,000 fund, you know, anacquisition at a couple $100,000,000 can really
move fund level returns.

(38:44):
And, you know, if you get a single billiondollar company that could return your funds
sometimes multiple times over.
And so, you know, venture we believe is hardenough as it is.
And we just want to try to stack the odds, youknow, even more in our favor in terms of
generating great returns.
And so that's why we've focused on this sub$50,000,000
I had recently had a podcast with a guy thatruns a lean AI leaderboard, which is looking

(39:07):
for these one person led kind of he calls itseed strapping, which is you raise a seed seed
round, then you you become profitable.
I've been thinking about the second ordereffects of AI on on investing.
So if AI does bring down costs, and what I'vekind of settled on is this bifurcated model,

(39:27):
pre seed and seed, and then the mega funds.
So you need money to prove out your thesis, andthen you need capital as a moat.
So in my version of the future, you see thesehighly value added precinct investors, and then
then Dreesons and Sequoia is writing kind ofthe billion dollar checks.
What do you think about that thesis and help merefine that thesis?

(39:49):
So I agree.
I very much agree that venture is becomingbifurcated.
I think you either need to be like a small,nimble, you know, collaborative early stage
investor or you have to be really a deeppocketed multi stage, you know, big brand name.
I think in the middle, it's really hard to winfor a few reasons.
Put it this way.
I think a lot of founders at the pre seed andseed stage, they know the risk of taking

(40:12):
capital from those big incumbent Tier one multistage firms, right?
The risk is you're kind of an option check tothem at the pre seed or seed stage.
And if they don't follow on and do your A,there's material adverse signaling in that,
right?
And so a lot of founders say, I'm going toraise my pre seed or seed from a smaller
specialized pre seed seed fund where I know I'mgoing to get, you know, really good engagement

(40:33):
from the GP.
Often it might be a solo GP where like, I knowexactly what I'm getting and who I'm getting it
from, and they can be really helpful and I'mmeaningful to them as a part of their
portfolio.
But the best founders when it comes to theSeries A, I think they want to raise from a big
brand name firm because it's signaling, youknow, raising from a big, great firm in your
Series A is a great signal.

(40:54):
But more than that, those firms have deeppockets and can fund you, you know,
theoretically through IPO and beyond.
Right.
And if you're a founder Which
they want to do too.
Which they they want want to Yeah, absolutely.
Which they want to do.
And if you're a founder, the biggestdistraction to your business is fundraising.
It can take a lot of time.
And if you have to go out and fundraise forevery round, a new process, like that's a big

(41:15):
distraction and a big time commitment.
And you know, if you can just raise your SeriesA from a big brand name firm and then
effectively not have to worry so much aboutraising subsequent rounds because you can just
go back to them, like that's a reallyattractive value proposition.
And I think it makes it really tough for, youknow, the mid sized Series A specialist funds,
you know, like the $200 to $400,000,000 SeriesA funds who maybe they can lead your A, but

(41:39):
after that they can't and you're going to haveto go out and find a new C or Series B investor
and again, you know, a new investor beyondthat.
So I I completely agree with you that thisbifurcation is is definitely happening, I think
it's only gonna get more pronounced.
You're investing in these sub $50,000,000funds.
Is there a pattern in terms of are theyconcentrated?
Are they leading rounds?

(41:59):
Are they collaborative?
Distill what you see as best in classstrategies?
So we take a very GP centric approach toinvesting, meaning we focus a lot on the GP.
What I mean by this and kind of going back tosomething you and I talked about earlier, this
concept of like GP thesis fit or founder marketfit, what works for one GP might not work for

(42:21):
another GP and beyond.
And so we want to back GPs where their strategyis really playing into their strengths.
And that means sometimes we back veryconcentrated GPs where we think like they're
exceptionally good at picking or engaging withfounders and like they have to have a more
concentrated portfolio.
But sometimes we pick more diversified GPswhere maybe they have really good network,
really good deal flow, really good access.

(42:42):
Maybe they're not as strong on the pickingside.
And so having a more diversified portfolio likeallows them to like really take advantage of
their network and their deal flow withouthaving to make like even more concentrated
bets.
And we've seen both models really work out.
We have a pretty eclectic portfolio where ourmanagers look different from one another.
Like we've got so we have a lot of solar GPs inour portfolio.

(43:06):
We have one or two partnerships.
We actually tend to prefer solar GPs.
This is maybe a little counterintuitive to whata lot of other LPs think.
But I think there's more risk in a partnership,actually.
I've just seen so many partnerships blow up andI've seen so many partnerships formed out of
convenience rather than out of a real likereason to partner with someone.

(43:29):
If you're a solo GP, you don't have to worryabout dealing with someone else and the
psychology of that.
And investing is such a business of biases.
And I've seen in organizations where peoplebecome possessive about their investments and
when people want to strike down other people'sinvestments, because you know, maybe that
person that's struck down one of yourinvestments and like, I think the decision

(43:50):
making and staying power and underwrite abilityin a solo GP is actually more attractive at the
end of the day.
And so we tend to focus a lot on those.
But we're open to partnerships so long asthere's a real reason for the partnership and
we don't think there's a lot of risk there.
What's one or two mistakes that you've madeearly on that you've since corrected?
One thing that I've actually changed my mind onover the really over the course of my career,

(44:15):
but more of the over the course of the pastcouple of years is, you know, I used to think
that having meaningful reserves in yourportfolio was a really good way to, you know,
increase exposure to the best names and likejuicy returns even more.
But I think it's really hard, at least whatI've noticed and observed is a lot of pre seed
and seed managers, it's really hard for them toeffectively deploy their reserves.

(44:39):
And here's why.
If you're a seed manager and you invest in theseed and maybe you have reserves to take you
through the A, you know, Series A companiesstill have a very high failure rate.
And it's, you know, you often don't have a lotof time to gather a ton of new incremental
information between a seed and a series A oreven, you know, pre seed or a seed, right?
And so by the time you have to deploy thosefollow on reserves, you're often doing it only

(45:04):
with marginally incremental more informationand you're doing it at a materially higher
valuation usually.
Right.
And so I think like reserves are actuallyreally tough to be effectively deployed for pre
seed and seed funds.
And I would almost rather those funds just buyup more ownership in the early days.
And I can't tell you how many times I've talkedto managers and asked them like, if you went

(45:27):
back and by the time this seed company raisedat Series A, like, you know which companies
were going to break out or did you know whichcompanies were going to fail?
And oftentimes they say, you know what, weactually did it.
Or the companies that we thought were really,you know, strung out of the gate ended up
plateauing and stagnating.
And the companies that were slower out of thegate ended up really ramping up.
And so I do think reserves can be reallyeffective, but I think they're more effective

(45:52):
actually in the later stages.
Like Founders Fund has been like exceptionallygood at deploying reserves and doubling and
tripling down.
But that's because they can deploy massivereserves like the Series C and D and beyond,
where like there's metrics, there's traction,you know which companies are already like
really on that hockey stick trajectory.
And so I've kind of come around and actuallyprefer our managers to have lower reserves, if

(46:15):
that makes sense.
It makes total sense.
And it's very interesting that the managers aretelling you that they may not know who the
breakouts are because there's a very directincentive for them not to say the opposite, to
say, oh, yeah, we should have just let's doubleour fund size, or we could have piled into
these three.
We knew for sure that these three were, but thefact that they're saying the opposite despite

(46:36):
the incentive is a pretty strong signal.
Yeah.
And I think this is a tough thing that you needto like try to figure out.
It's very easy for managers to retrofit a storyand to say, oh, you know, we made this
investment, like five years ago, we made thatinvestment because we had this thesis on this
thing that nobody else believed in.
And we did it and we knew that this company wasgoing to break out and become a billion dollar

(46:57):
company or a $10,000,000,000 company.
The reality is venture managers think every oneof their companies could become a
$10,000,000,000 company at the time theyinvest.
Right.
But, you know, obviously not every company isgoing be successful.
Most actually aren't going to be.
And I think what people don't realize isoftentimes it's not until much later when it

(47:19):
becomes obvious that a company is going to besuccessful.
This has been an absolute masterclass on downinvesting, fund of funds.
What would you like our listeners to know aboutyou?
Pattern Ventures or anything else you'd like toshare?
The first thing I'd say is, know, Pattern,we're open for business, we're actively
deploying.
And so we'd love to chat with, you know, anymanagers that are really high quality and
raising.
And we source through our network, sourcethrough a data tool that we have, but we don't

(47:42):
have a monopoly on sourcing.
We don't believe anybody does.
And so we're always happy to meet with folksthat, you know, people really highly recommend.
And the other thing is, I would just say like,you know, when we were building Pattern and how
I've thought about being an LP over the courseof my career is I've seen too many LPs that I
think feel entitled or feel that GPs owe themsomething because like we're giving them
capital.
And I just think this business needs more liketrue partnership.

(48:06):
And I think like at the end of the day, aremarriages.
These are really long term relationships.
Two way streets.
And so I would just like encourage everybodylistening to like really think deeply about
partnership, both if you're an LP or a GP andrecognize that like none of these relationships
are are one way streets.
And and I think that's, like, something that wereally hold, near and dear to us here here at

(48:26):
Powder.
Big purpose of this podcast is to tell the LPstory in hopes of basically telling kind of
both sides of the story so that there could bemore empathy and relationship in in the
industry.
Thank you, John.
Look forward to sitting down in person verysoon.
Likewise, David.
Thank you for having me.
Thanks for listening to my conversation.
If you enjoyed this episode, please share witha friend.
This helps us grow also provides the very bestfeedback when we review the episodes analytics.

(48:50):
Thank you for your support.