Key Learnings

What we got right and wrong building an emerging venture firm.

As a firm we are committed to principled, life-long learning. This comes with honest reflections upon our actions and the conclusions we draw from them. This post is an attempt to do just that.

https://unsplash.com/photos/1A26jv_gmZk

Under-estimating geopolitics

Since 2020, it has become readily apparent that the global landscape is undergoing a profound transformation. Conclusively, our industry is poised for significant evolution. We find ourselves in a transitional phase, bidding adieu to the once-dominant U.S.-led world order and transitioning toward a more multilateral one. This paradigm shift carries a multitude of implications, not only for the interplay between sovereign nation-states but also for the intricate dynamics within those states and their citizenry.

As the era of U.S. hegemony slowly wanes, we observe the ascendance of new superpowers, most notably, China and India. The underlying cycles governing these geopolitical shifts are apt to span multiple decades before fully manifesting. Transient setbacks, such as fluctuations in economic growth or advances in technology, may exert nominal influence on the grander trajectory, unless they precipitate systemic collapses akin to the events of 1989. The ongoing hotbeds of conflict in regions like Ukraine, Armenia, and Israel serve as poignant manifestations of this evolving world order, where the United States finds itself embroiled in numerous proxy wars concurrently, as it strives to uphold its leadership position. The duration and terms of this leadership's sustenance remain shrouded in uncertainty, with the pivotal 2024 election being a critical juncture. As conflicts proliferate and Western powers become increasingly entangled, the specter of new confrontations looms ominously. We turned a watchful eye on regions like Taiwan and the Middle East.

These external dynamics yield an intricate and nuanced relationship with the internal cycles affecting governments and their constituents. Given the interplay between these two (external and internal) dimensions, it has become challenging to distinguish real cause-effect relationships from mere coincidences. However, one discernible trend amidst this complexity is the clear shift toward re-nationalization and de-globalization. Governments across the globe have embarked on strong intervention in market affairs. Initiatives like the CHIPS Acts in the European Union and the United States, the Infrastructure Bill in the U.S., the Inflation Reduction Act in the U.S., Mica and the AI Act in the EU, and the Securities and Exchange Commission's approach to regulating the crypto industry through enforcement measures collectively underscore this interventionist fervor. Subsidization of domestic economies, the politicization of technology regulation, and the imposition of export constraints have now become the prevailing norm. Escalating debt-to-GDP ratios have precipitated the emergence of high inflation and other forms of financial repression, compelling Western nations to pursue these measures as fiscal imperatives.

These external and internal transformations are inexorably reshaping the role of governments, as articulated by Russel Napier, a market strategist and historian in late 2022:

In the past four decades, we have become used to the idea that our economies are guided by free markets. But we are in the process of moving to a system where a large part of the allocation of resources is not left to markets anymore. Mind you, I’m not talking about a command economy or about Marxism, but about an economy where the government plays a significant role in the allocation of capital. The French would call this system «dirigiste». This is nothing new, as it was the system that prevailed from 1939 to 1979. We have just forgotten how it works, because most economists are trained in free market economics, not in history.

The realization of the aforementioned dynamics has prompted us to engage in introspection concerning our societal role as venture investors. In the bygone era, I characterized the venture industry as the "gatekeepers of the future," implying that our collaborative efforts with entrepreneurs were pivotal in bringing about transformative change. However, this assertion may no longer hold true, given that governments are back to the center stage. Notably, the monumental accomplishments of our time, such as the development of Covid-19 vaccines, the development of the internet or the first moon landing, owed their realization to decisive government intervention.

In light of this shifting landscape, some pressing questions emerge:

How does a venture investor make informed decisions when markets sway in response to the unpredictable actions of politicians?

Do these kinds of developments disrupt the foundations of our bottom-up investment frameworks?

How can we effectively incorporate and weigh macroeconomic factors into our investment theses and portfolio construction moving forward?

What implications do geo-political events hold for our limited partners our global sources of patient capital with whom we collaborate?

For any venture investor or entrepreneur navigating the tumultuous markets of the 2020s and 2030s, these inquiries merit careful consideration. Regrettably, our systematic contemplation of these matters commenced only subsequent to the invasion of Ukraine, and in retrospect, the writings have been on the wall for much longer.

The last 25 years was a great time to be an investor. With Geo Politics and politics on auto-pilot, running a portfolio became a mathematical routine (quants). (…) The Goldilocks era is over. We live in a fundamentally different world now.

Marko Papić, Macro Strategist (I cannot find the exact source anymore, sorry!)

We do highly recommend Marko’s book Geopolitical Alpha where he applies a constraint-based framework to assess geo-political risks.

Adapting to changing technology cycles

As early stage investors we are exposed to market timing risks driven by technology adoption cycles. In some cases we have been right, in others we have been over-estimating the speed of technology dissemination.

(1) With the introduction of LLMs working at scale the world of software development and investing changed. The long term implications are hard to predict and it is not an easy task to reflect such developments within a bottom up investment framework either.

Similarly to the generation of content, everyone is empowered to create software by leveraging conceptual knowledge and natural language. This empowerment sets in motion a chain reaction, wherein the demand for software experiences exponential growth (rebound effects) while the market readily accommodates it, courtesy of near-zero entry barriers and immense scalability. Consequently, the next critical link in the value chain centers on the distribution of such software. As per current trends, it appears that this distribution will remain firmly in the grasp of entities overseeing consumer hardware and interfaces—think giants like Google, Apple, Microsoft, or those presiding over business infrastructure, such as Amazon. This continuity aligns with the substantial investments these companies are directing toward AI innovation.

If the creation and distribution of AI infused software is commoditised, what will emerge as the new bottlenecks of computational innovation?

“Software has been eating the world” and is now becoming marginal. The new frontier of computational innovation will be hardware led - access to raw compute sustaining Moore's Law; elastic compute infra underpinned by strong cryptography and the free flow of information between the worlds of atoms and bits.

Our take might be contrarian and counter-intuitive to some but we are quite confident in such conclusions. We will write about related topics in more depth soon.

(2) In other areas our market timing has been off as we have been overly optimistic about the adoption of crypto technologies to coordinate labour at scale as opposed to just capital. A concrete example is the emergence of DAOs (dencetralised, autonomous organisations) as new forms of organisations orchestrating labour and resources programmatically over the Internet. Our collective vision was to create global organisations which are run on self executing code as opposed to pen and paper. We broke it down into composable, enabling products (e.g. voting, governance, treasury or identity modules) to solve coordination problems without intermediaries. Unfortunately, none of such modules reached product market fit and are unlikely to do so in the near future. The reasons are obvious in hindsight. First, people need to prove their employee status, pay taxes, enforce titles against their offline counterparties or pay local merchants at some point and without proper interoperability between the old and the new systems this is nearly impossible to achieve. Second, there is early evidence of the DAO framework being effective to coordinate labour in the form of running specific algorithms (Bitcoin’s proof of work) or providing standardised services towards a network (bandwidth, compute, storage). However, those mechanisms cannot be abstracted and applied towards more complex, nuanced services or products currently. Running the conventional legal system and the internet-based one in parallel doesn’t resonate as one is prevailing the other. We are long term believers in on-chain organisations being fundamentally superior to pen and paper based ones but the time needed to make this transition practicable is likely 5-10 years out and startups don’t have that kind of time. We shared our concerns publicly at Inflection's DAOday in September 2022.

Constructing a portfolio and pacing investments

The disciplined execution of our portfolio construction plan and investment pacing have been key for us. As we closed our Inflection Mercury Fund in 3 closings over the course of 2021 our capital base grew from $4M in the first towards $40M in the final closing. To stay conservative we always defined the amounts closed as our final capital base and sliced our allocations accordingly. In hindsight that created some outsized (too small), asymmetric positions in our portfolio. Yet, we believe that staying conservative was the right thing to do as markets could have moved into the opposite direction as we've witnessed during the pandemic - when we lost many handshake commitments from one day to another. We would have been terribly over exposed if we had assumed a larger capital base before closing out.

Further, we managed to invest quite broadly within our mandate what prevented us from being overly exposed to single niche-categories further adding to concentration risks. Specialisation works well in times of market expansions but it works poorly in times of market contractions (NFT only funds, anyone?) or large scale resets as it changes the risk / reward profile of a fund. We prefer to operate under a broader mandate and sticking to first principles without being opportunistic - to be adaptive enough to changing markets.

Related to portfolio construction is the aspect of investment pace. The most important driver of venture capital fund returns is not the ability to pick winners but the vintage year. Hence, stretching out the investment period over 4 years was the right thing to do. We are grateful for our collective discipline despite the euphoria and hype around us especially in 2021 and 2022. That said, we did try to time markets when we built up some of our liquid positions over the course of nine months. In hindsight we should have picked an even larger time frame to dollar-cost-average into such positions as we intend to hold on to them for about a decade.

Diversifying service providers and counter parties

Venture funds are known to take incredibly concentrated positions coming with high risks. The concentration of single positions is a function of portfolio construction (how many positions and how they are weighted). It’s not unusual to see early stage funds holding 50% of their fund in a single positions after a large uplift in one company.

With that context it becomes existential to at least diversify all kinds of operational risks and in particular that of failing counter parties in the form of limited partners (yes they can default) and banks (yes, they can default) including custodians (yes, they can lose assets and default).

Our setup has been tested during the collapse of Silicon Valley and Signature Bank as well as FTX and its fallout. We established a war room protocol to deal with such situations and provided our ecosystem with a guide to navigating a banking crisis. Ever since operational risk management has been deeply engrained in our firm and we can only recommend every emerging manager and founder to follow through.

---

Where do you disagree and why?