E204: Going All In: The Risks and Rewards of Concentrated Investing

28m
In this episode I speak with Rafael Costa, who co-founded Across Capital to back category-leading software companies across the U.S. and Latin America. We dive deep on the Brazil tech flywheel — from why the central bank and Pix have accelerated fintech innovation, to the infrastructure winners like QI Tech that are becoming foundational rails for payments, banking and credit. Rafael walks me through Across Capital’s concentrated, high-conviction approach (a ten-company portfolio, deliberate sizing, then backing winners over time), how they underwrite downside protection in growth equity, and what AI actually changes for regulated industries. Along the way he shares practical diligence habits (the “what really matters” slide), how they build conviction over ~17 months, and one piece of advice he’d give his younger self about focusing on the present to compound relationships and learning.

Listen and follow along

Transcript

Rafael, welcome to the How I Invest Podcast.

Thank you for having me, David.

Pleasure to be here.

You started your career at some of the really top growth equity firms on the planet.

You were at Summit Partners.

You were at Vulcan, which was the investing arm of Paul Allen.

And today you have your own fund.

Tell me about what you're focusing on today.

Cross capital really started over a decade ago.

It was a culmination of who I am as an individual.

I was born in Brazil, been in the U.S.

for the last last 25 plus years.

I'm always investing in technology throughout my career.

So I went down the traditional path.

And as you alluded to, I had the great opportunity to kind of cut my teeth into the world investing initially at Summit and subsequently at Vulcan.

And now within Across,

we bring with us a lot of the learnings from those times.

And we're focused on backing software businesses across the Americas.

We really started out a decade ago, the thesis for Across, which was, you know, predicated on the fact that I i was brazilian in silicon valley i saw a lot of um entrepreneurs come through here particularly from latin america trying to raise money silicon in in sent hill um and i was helping them think through how to position their businesses um for you know the the silicon valley vcs um and at that time i started personally investing in what went on to become the first crop of unicorns in lat m um that gave me conviction for time in lat am make us made us a little bit of a unique animal um in the U.S.

where our portfolio company was seeking us out to help them with Latin America, whether that was on the talent front, whether that was in the business development side.

And that was the genesis for Across.

Tell me about the Brazil tech ecosystem.

How's that evolved in the last decade?

Great question.

Look,

I think Brazil has delivered one of the most successful returns in all emerging markets when it comes to tech.

I'll say it's Latin's Silicon Valley, right?

Whether you want to look at the Mercury Libre of the world, which is 100 billion plus market gap company, New Bank, which is a 60 billion plus market gap business, perhaps one of the most successful fintechs globally,

or others such as Pog Ceburo or XP.

We have had some MA also happen down there, but it's a fertile ground for innovation.

I think the backdrop is a big market measure, both in terms of population and GDP, that is extremely tech-savvy, both on the consumer side, but also on the enterprise side.

And we're going reaching a point where the talent density is there, right?

We're taking it to this flywheel where we have had exits.

We have had folks building 10 billion plus companies, and they're on to their second or third go-around.

So again, it's pretty fertile ground for innovation tech in general.

One of the undertold stories of Brazil is its fintech and financial infrastructure.

Why is Brazil have such high market share when it comes to these fintech startups?

If you take a step back and you understand kind of how the banking system in Brazil was built, it was really built out of chaos, right?

If you look at the 80s and the 90s, the country suffered from hyperinflation.

That required the banking system to operate in effectively real-time with

same-day settlement.

And that kind of sets the tone for a lot of the innovation we have seen in the Brazilian fintech ecosystem, which is further propelled by a central bank that thinks like a startup, one of the most innovative central banks in the world, extremely progressive when it comes to innovation.

What they've done in the last decade is nothing short of remarkable.

We have had innovations such as PICS, which is the real-time payments rail provided by the central bank, which now has

a penetration of 80% plus of the adult population in Brazil, whether that's open finance and

providing for data, portability, embedded finance, and a lot of innovation within

the financial markets.

And then you have other more structural nuances that makes it quite compelling to perhaps disrupt these incumbents, right?

You have a very concentrated banking system, extremely high spreads when it comes to credit, still penetration to be had as it relates to just financial products in general.

And that ultimately translates into a very compelling market when it comes to innovating on the fintech side.

You have a Brazilian company, QITech, that really powers a lot of the infrastructure and the FinTech in Brazil.

Very fortunate to have had the opportunity to work with the team over at QI Tech.

It was actually the first investment we did out of the Across Capital Fund.

And we have been spending a great deal of time within financial infrastructure in Brazil specifically.

And the notion being, you know, picks and shovels plays into this massive theme which I just alluded to earlier, betting in the race, not in the horse.

And QI Tech, what it does is provides the infrastructure that enables any company, whether that's a retailer, a fintech,

a shopping mall, a big pharma company to offer financial products.

So think of it as a stripe of that part of the world, but with other elements as well.

It is a business that was extremely capital efficient when we had the opportunity to get involved,

had explosive growth, growing 200% plus year over year, with EBITDA margins in excess of 50%, something that you do not see every day, and still with plenty of runway ahead as relates to the market itself.

Tell me the difference between a LATAM fintech company and a US fintech company.

How is it different to start a company there versus in the U.S.?

The way we see it is down in Latin America, you have the opportunity to build more end-to-end platforms.

And this is just driven by just the overall market dynamics.

If you look at QITech, you really started out offering much more of a linking as a service product, whereby they were providing the infrastructure for traditional companies or fintechs to provide credit.

Now we have a banking as a service product, which is checking accounts and payments.

Now we have a fund administration business, which also helps custody the funds of clients.

So it is building much more of an end-to-end platform, which is very different from the U.S., whereby oftentimes you're trying to solve a specific pain point or going after a big, but perhaps narrow market.

In Latin America, there's an opportunity to build things that are much more platform-driven and platform-oriented.

And why is that?

Is that the barriers to entry?

Is there some technological advantage?

Why are you able to kind of verticalize these solutions?

Resources are just more scarce down there, whether that's capital, whether that's talent.

And as a result of that, the winners tend to compound and drive barriers to entry for new entrants.

Whereas capital is more abundant here in the U.S., talent definitely density is

much deeper relative to what am.

So those dynamics drive some of these forces I just mentioned.

And you have very concentrated portfolio construction, 10 companies.

Tell me about the strategy behind concentrating into less than a dozen companies.

Concentration forces clarity at the end of the day.

When you only get 10 shots, you tend to do deeper diligence.

You lean into your edge, right?

It gives you the ability to partner with companies in a meaningful and active way.

So ultimately, it is something that's foundational across our investment philosophy.

It's not a mindset of buying call options, right?

It's every single one of our partnerships are meaningful to us.

And that drives alignment with the entrepreneur.

We don't have a portfolio of 30, 40 names.

It's 10 of those.

So everything is incredibly meaningful to us.

It drives a scarcity mindset, which

for us means

we need to have

ruthless prioritization, I'll say, on how we spend our time.

And that changes the way we source.

It changes the way we engage with the companies.

It changes the way we pick them

so the punchline is look by by only investing into 10 companies that forces us

to be incredibly focused you know everything is meaningful to us and we like that alignment with with our entrepreneurs and we also like the fact that you know gives us the ability um to drive in our view very good risk adjusted returns for lps And you said it forces ruthless prioritization.

Why does having a concentrated portfolio change your sourcing capacity or your sourcing function?

Top of the funnel, we're still talking to thousands of companies, right?

But on the middle of the funnel and bottom of the funnel, that's where the ruthless priorization comes in because we are spending a great deal of time with efficiency and speed to get to the bottom of things that really matters on the investment decision.

So ultimately, we're spending

an incredible amount of time, probably around 30 or so opportunities per quarter where we're diving in and really conducting deep diligence.

On average, we get to know a company 17 17 months prior to making the investment.

So that diligence is occurring over a year and a half period.

That allows us to, one, ideally have built a relationship with that entrepreneur and have earned our seat at the table.

Two, to have the conviction to move with efficiency and speed at the time that they're looking to raise or transact, if you will.

We have these top companies that we're pursuing.

We're ultimately aiming to do two to four investments in any given year.

And we need to focus on the companies that are going to beat those two to four.

And that is where the ruthlessization comes into play it's Warren Buffett that popularized this idea when you become an investor you graduate college you should have a punch card of 20 companies and you only get to invest in 20 companies kind of throughout your career and it's this kind of ruthless conviction I would call it that you have to be so convicted I also think there's this bias towards diversification on a fund level that's unnecessary and that's a very pro-GP strategy, not a pro-LP strategy.

LPs want to have high concentration hats, want to have a basket of highly concentrated names.

Those tend to be the best performing funds.

You could argue it's not the case for the pre-seed and seed, where there's power laws where any one company, you want a lot of shots on gold.

But once you get to a certain stage, you typically want your managers to be highly concentrated.

Couldn't agree more.

And it's actually interesting because...

I can't speak for LPs, but if you think about it, at the LP portfolio level, they have ample diversification, right?

Because you're investing across multiple funds.

Each one of these funds are investing in underlying companies that provide further diversification.

So

they are diversified for all intensive purposes.

In our view, as a manager of a fund, we like concentration because of the points I alluded to earlier.

And to give you a practical example, QYTech, which we talked to about it earlier, is a 20% position in our fund.

We truly live by this high-conviction, high-concentration mindset.

Is that a fund returner?

It's looking like it.

We just actually announced yesterday a follow-on round, which marks the fourth time we're investing into the company.

We co-edited with General Atlantic.

Given where the pricing of that came out, which is close to $2 billion, we first invested in the business at $300 million, is very close to being a fund return already.

I just interviewed John Felix, who was at WashU St.

Louis and worked under the Faybuilt CIO, Scott Wilson.

And Scott Wilson not only invested into funds that are very concentrated, but he would meet and his team would meet with the managers and find out their top concentrated positions, whether there was an opportunity to deploy even more.

So you had this kind of like ruthless concentration.

Now, that being said, it probably was still hundreds of positions, but the idea that you have to have every fund be 30 to 50 companies and then you have 20 funds, the math does not justify that kind of diversification.

Makes a lot of sense to me.

We had this back and forth last time we chatted.

You believe that growth equity is not about the power laws.

It's about consistency and maybe even downside protection.

What leads you to think that growth equity is kind of a different asset class than traditional early stage venture?

There's different types of strategies within the growth equity.

I think growth equity, you know, has evolved quite a bit over the last decade.

You could argue there's what some may view as traditional growth equity, which is minority investing into high-growth businesses that tend to be capital efficient or profitable.

This is how the likes of some at TA in the mid-80s were born, right?

So you're effectively sitting between venture and buyout, but still driving returns predominantly through growth and have been a minority investor.

And obviously, I think to a large extent, there is venture growth, which is just later stage growth rounds.

And that resembles much more of the growth equity 2.0 than the growth equity 1.0.

But to your question related to consistency versus power law, particularly the way we think about it, given the concentration of our portfolio, we're much more focused on consistency of three to five X returns versus chasing something that can be binary, which is chasing the power law, something that could potentially return 10, 20, 30x, or be a zero.

We're less interested in those.

We're much more interested in finding things that we have extremely high conviction on hitting our minimum return threshold of 3x.

with the downside protection, whether that's driven by the state reinvesting, the structure reinvesting, the IP the company has, but we have high confidence that capital impairment is going to be very low.

In fact, if you look across our track record, we only lost money in one deal.

So that speaks to our mindset around downside protection and being mindful of capital impairment while still preserving

the capability of still achieving a 5x plus return.

So that's ultimately what we're looking for is this notion of consistency.

That does not mean that we won't have perhaps one or two companies in the portfolio, in our case, of 10 companies that will drive a lot of the returns in the fund.

We just alluded to QITAC being a 20% position and being a fund returner.

Obviously, the fund returning component drives by us sizing that position accordingly, which I think is something that we don't talk about enough is how sizing matters.

And that's, you know, a byproduct of concentration on the return of the overall fund and portfolio.

But I, as a manager, I cannot be chasing things that

could potentially have capital impairment because the portfolio construction does not allow for that.

I want you to double-click on the sizing component.

You were at Summit, you're at Vulcan, you saw the kind of best practices.

Obviously, you invested 20% in one company and you have a 10-portfolio position, so it was roughly double your average position.

Is that basically how you explain your sizing strategy, which is invest even more into your higher conviction bets?

Is there more math behind that or double-click on your your strategy?

On QI Tech, like I said, we invested four times in the business.

So we built this position over time.

So it was not from the get-go that we got to 20%,

but from the get-go, we dissized this position at 12.5%.

That was our initial check.

And then subsequent to that, we have increased it to 20% as the company has beaten our plan every single time we have re-underwritten the business.

And that's part of the strategy.

Part of the strategy is to back your winners, if you will, and put more money in things that are working.

And we think they're still, on a risk-adjusted basis, a compelling case to be had to allocate capital to it.

And I think QITAC is a prime example of our strategy.

One of the things that you see a lot in the venture space, growth equity space, is this retrofitting of the thesis where you fit your thesis to what the current situation is with a company.

And sometimes cynically, that's done to LPs to kind of paint a new narrative, but also it's done to fool the investors themselves.

So tell me about that.

And it seems like you've you've institutionalized memo writing and assumptions and underwriting outcomes.

And how important is that to stay on course with being a good investor?

For us, discipline is core, right?

We're not in the business of getting lucky.

We're in the business of driving consistent returns and having a mythology to do so and a framework by in which we do that.

And that translates into

system frameworks and processes that we hold internally.

So to your point, yes, every time we re-underwrote the business, we looked at the underwriting we did, whether that top line, bottom line, margins, returns, the drivers of each one of those things, how has it performed relative to plan?

Why did it go better?

Why did it go worse?

Every memo we write, we have a slide that's what we call a what really matters slide is really forcing ourselves to pick no more than five drivers of the investment case, right?

Ultimately, an investment decision, in our view, comes down to three to five factors.

And that's what really matters.

And in a very objective way, assessing that.

And then as we re-underwrite the business, assessing how those what really matter points have changed.

Um, and as in an as objective manner as possible, and as quantitative as possible, obviously, there's qualitative assessment that goes into any investment, but that's how we think about it.

The second time we invested into QI tech, I don't have the number top of my mind, but it was 20 to 30 percent above our underwriting plan that gave us conviction to put more money behind the business.

The margins actually expanded.

We underwrote margins decreasing and suffering from competition was actually the opposite that happened.

And as a result of that, we decided to put more money into the company.

So it's all about having a process which we're constantly fine-tuning.

I'll say that.

That's also an element of, for us, it's extremely important.

That's what makes investing, for me,

really fun and intellectually challenging because you need to be adapting and improving yourself every step of the way.

Tyler Sosin, who spent 17 years at Menlo and Excel, he told me that no company ever died because they saturated the market.

In other words, if you go in, you saturate a market, you execute well, there's always these tangential markets that oftentimes you don't see until you've gone to that level.

So I often think about this embedded asymmetry in companies that succeed.

A famous case is Carta.

They went in, they did the cap tables, and then they did fund admin and 49A evaluations and all these different markets.

Have you seen that over your career, this embedded asymmetry of if you do execute, sometimes a 3 to 4 X could end up being a 1020 X?

or is that an extremely rare case?

The way I think about it is the compounders keep compounding and the compounding is much more powerful than at times you go in thinking about it.

And to your point is exactly that, I think, is this notion that these businesses that are foundationally solid with phenomenal teams behind them, they consistently find ways to keep compounding.

And in this case, whether that's expanding TAM, that's expanding products organically, inorganically.

And I have seen that.

I think QITEC is something that we're living through as we speak.

The business has

expanded TEM considerably since the first time we invested, both organically and inorganically.

At times, we see geography, geographically, as well.

So, you're operating in one given market, which we underwrote the base case on them just playing that market.

All of a sudden, they are playing

regionally or globally.

So, I completely buy that.

And I think we,

at a different sort of level, we see that with the Magnificent, with the Mag 7, right?

These are businesses that just have,

they just keep compounding, right?

It's a flywheel that just keeps getting stronger, whether that's because they have phenomenally sound businesses, which I think is the case, but also talent, capital, products, is a flywheel that just has gotten stronger and stronger.

And I firmly believe that.

And why I think investing, you know, one thing I learned in my career, investing in the number one player in a given market is incredibly important because the value disproportionately accrues to that number one player.

Going full circle to these businesses that compound in Brazil once they have an advantage, but even the US, even Uber, who would have thought that Uber eats and now probably at some point they'll have driverless Ubers.

It's almost literally unknowable what the opportunities will come, either from within the company or something happens in the industry.

When Uber started out, it was just black cars and then Lyft started with their pink mustaches and the peer-to-peer and then this unlocked this whole industry.

So there is something extremely valuable to staying alive and compounding that I think is underpriced.

Absolutely.

Absolutely.

I couldn't agree more.

And I think just to thread the neo a little bit, I think as an investor,

it is almost like the blue sky case scenario that we talked about earlier, the whole notion of power law within growth and how we think about it.

Ultimately, we're we're investing in businesses that have that tail potential outcome of keeping compounding and being the parallel.

So, we want to have that.

We do not want to depend on that.

So, that's, I think, a very different way.

And in fact, when we underwrite, we underwrite cases and we assign probability to them.

And those we get to like a weighted probability set of outcomes.

And oftentimes, in the upside case scenario, if I will, it relies on this

expansion of product, temp, and future compounding.

And I think on the best best ones, we always undershoot it, always.

But you cannot rely on that.

Perhaps an obvious question, but what compounds in the market leaders?

Is there anything non-obvious that compounds in a market leader?

Look, I think the obvious ones is what's often talked about, whether that's network effects,

data, network effects.

But ultimately, if you think

in a very first principles thinking, you're ideally investing in businesses that as they scale, the value that they provide for their customers increases over time.

And perhaps the cost to serve that additional customer decreases over time.

And that drives the compounding effects, whether that's grounded on the data network effects, whether that's grounded on just the social network effects, whether that's grounded on the infrastructure that they built.

What we're seeing now with LLMs and

the likes of OpenAI, we're talking about the compounding effects.

I think that's a a business that's going to keep compounding for a long time.

Why is that?

Users drive data.

Data drives monetization.

Monetization drives capital.

As capital comes in, they can drive more users and drives that five-wheel effect and drives the stickiness.

And they're able to drive more value with the unit cost of serving that additional user decreasing over time.

So that is

ultimately what drives this compounding effect.

You mentioned AI.

How is AI affecting the growth equity space?

There's this thesis that in the future there's only going to be the pre-seed investors and the Andreessens and the Sequoias writing billion-dollar checks.

You know, there's this concept of seed strapping where you just raise a seed round and then you don't have to raise more money.

How is that affecting you today, day to day?

It's an interesting question.

Look, I think bootstrap companies have existed for a long time, so that's nothing new.

I think that's the reality of it.

We talked about the Growth Act 1.0.

100% of Growth Act 1.0 was predicated on backing bootstrap companies that had gotten to growth and scale without a single penny.

And, you know, why is that?

I think it is the best proxy that you have a sound business model in a great team, right?

If you're able to get to scale with growth and being profitable, it's you probably have a good business model and you're very good at executing behind it.

So it's just empirically a very clean way to look at it.

So I don't think the whole seed bootstrapping is anything new.

The way we think about it is,

look, there's a tremendous amount of disruption happening.

And as an investor, that creates excitement, but that also creates a very dynamic environment that you're investing into, which I tell my team with a lot of landmines, right?

Because the level of innovation is only speeding up.

So what may be unique

and innovative today may not be the case tomorrow, may be disrupted extremely quickly.

So most become increasingly important.

What we are very excited about on the advent of AI as it relates to to growth equity is regulated industries.

It's industries, you know, a la financial services, which we just discussed, which are going to benefit from AI in a massive way.

And perhaps that's driven by efficiency, automation,

and drive what was a good business to become a great business.

But they're not necessarily going to be disrupted by AI.

They're going to be

enabled by, I wouldn't call it enabled, but they'll be sort of propelled by AI.

They're going to benefit from AI.

And that's what gets us excited because just playing on the

edge of AI of innovation, I think for us

is just tricky.

It's very tricky.

And there you go to the whole notion of betting on binary outcomes and perhaps having the need to back 30 or 40 companies and two of them is going to pay for the entire portfolio.

It's just not the game that we're playing.

So we tend to shy away from those.

Wherever we're excited, it's more regulated industries that will benefit from AI in a massive way.

And teams are adapting that very quickly.

So we've known each other since my senior in college, your freshman year.

So we go way back.

If you could go back and give yourself advice when you were graduating from college and starting at Summit Partners, what would be the one piece of advice that you could give yourself that would accelerate your growth in your career and maybe avoid mistakes as well?

It's a great question.

I think, look, focus more on the now

and just make sure that you are, you know, oftentimes I think a lot of individuals, myself included, that tends to be ambitious, that tends to be a builder, if you will.

You tend to focus a lot in the future and kind of how to maximize the future, and you don't realize the future is being built now.

So, focusing more on the now, whether that's on the networking piece, which is incredibly important, and obviously, that only compounds with time.

I think now that I'm 15 years in my career, I'm still crossing paths with you, as you alluded to.

We met each other, you know, I was a freshman in college, and here we are having these conversations.

So, these relationships compound, whether that's also on the learning, learning today, that will compound.

So, say to sum it all, is focus more on the now i i think oftentimes i was so caught up with what the next step may be um whether that's in my career my learning on the investing and i lost sight at times that you know the field the future depends on what i do today and today is the most important thing that i can do um so be ruthless about today i've tried to a little bit focus on the now.

I still respect my future thinking, but sometimes it's like, let's celebrate this, like let's celebrate this milestone.

Yes, it doesn't mean that we're done.

It doesn't mean that we're at the finish line, but let's take the time to celebrate this milestone because that'll only motivate us more in the future.

Yeah, couldn't agree more.

What would you like our listeners to know about you, about across capital, or anything else you'd like to share?

Ultimately, Across is built on conviction, and I think we covered this on craft and on our cross-border insight.

We're looking to forge meaningful relationships, which I think is the best kind of relationship where you're mutually meaningful to one another.

And that translates into 10 partnerships where we can be relevant and additive partners to them.

We're builders first, investors second.

We're not in the business of chasing momentum or things.

We're in the business of building

businesses for the long haul.

That's ultimately what we stand for and who we are.

The final piece is we value a lot the personal relationships.

I think this is something that oftentimes gets lost.

This is the notion that great businesses are built by people.

And you want to be in business with people you're

enjoy being in business with.

And we value that component of it

quite a lot.

Well, I value my relationship with you.

It's 18 years and counting, probably in a couple of months.

So, I appreciate you being my friend and jumping on the podcast.

I appreciate it, David.

Thanks, Rafa.

Thanks for listening to my conversation.

If you enjoyed this episode, please share with a friend.

This Helps Us Grow also provides the very best feedback when we review the episode's analytics.

Thank you for your support.