In our last update (“Sourcing Resilience”), we shared how we identify which sectors to invest in using a quantitative method for assessing total vulnerability to volatility and thus opportunity size. Now, we would like to share more on what we think about portfolio construction and in particular our desire to invest in what we call “Earthshots” and “Moonshots”. We will start by defining those terms and then use our “Built Environment'' (construction) theme as a case study for how we use these terms in practice.
Earthshots & Moonshots
Fund I will comprise ~12 investments. Those investments will be made within the core themes of the Built Environment, Energy Infrastructure, Supply Chain / Logistics, and Advanced Manufacturing. We selected these themes because we identified large vulnerabilities to volatility within them. Those specific vulnerabilities form the basis of our thesis driven investing.
For each specific vulnerability to volatility that we identify within a theme, our goal is to invest in two types of companies (an Earthshot and Moonshot) whose products address the given vulnerability. We define those as follows:
To summarize, Earthshots are companies taking market risk (can you drive adoption?) not technical risk (can it be built?). Earthshots need to identify and capture scalable, defensible distribution channels. Whereas Moonshots are companies taking technical risk not market risk. These companies need to have unfair access to the talent and capital necessary for R&D.
Observing the diagram above, you can gain an intuition for the idea that Earthshots and Moonshots are complementary to one another – two sides of the same coin. What we want to propose next is that any given opportunity (large vulnerability to volatility) could be addressed by either an Earthshot or Moonshot type company. Below are a few examples:
Nassim Taleb, a statistician and options trader known for his best-selling books (Fooled By Randomness, Black Swan, Antifragile, etc), often advocates for what he calls “non-predictive decision making”. He emphasizes making decisions based on robustness and resilience, rather than on predicting specific outcomes. It encourages focusing on the potential range of outcomes and how to handle them, rather than trying to predict exactly what will happen.
So how do Earthshots and Moonshots relate to non-predictive decision making? Non-predictive decision making allows you to thrive in highly uncertain environments (especially ones with power laws at play). Early stage venture capital is certainly one of these environments. We can predict with confidence that volatility within a certain sector will occur. However, given the nature of complex systems, we cannot predict how a system will resolve as a result of that volatility – namely whether an Earthshot or Moonshot type solution will be favored by the market to address the volatility.
Using our coin analogy, we can be confident that a coin will be flipped (volatility will occur) but we cannot predict whether it will land heads or tails (whether an Earthshot or Moonshot will be favored). Given the power-law nature of returns in venture, it is therefore optimal for us to “maximize exposure to the coin” to ensure we capture as much potential upside as possible. Hence, we will try to pair Earthshots and Moonshots to drive higher returns.
In the illustrative chart above, we are also showing how the adoption curves of Earthshots and Moonshots differ. Because Earthshots tend to be software companies that have incremental adoption driven by network effects, they show smooth exponential growth (ideally). Moonshots on the other hand often show step-function changes in growth as technical leaps are made that exceed specific thresholds for performance demanded by a given market (ie price per kWh in energy, price per square foot in construction, etc). Thus another benefit of pairing is: Earthshots help to smooth the J-curve in venture, while Moonshots maximize potential upside.
Case Study: The Built Environment
Now that we have a conceptual understanding for pairing Earthshots and Moonshots in our portfolio construction, we can move on to grounding this methodology in a specific case study. For this update, we will focus on what we call the “Built Environment'', which encompasses everything from how we plan buildings (design, compliance, tracking) to what we build with (cement, timber, steel, etc) and how we build it (prefab, modular, 3D printed, etc).
The Built Environment is beset with vulnerabilities to volatility, especially in the US. In construction, volatility often takes the form of delays. Why are delays so costly with construction? It’s because of cascading dependencies. If you can’t source a material in time, contracted labor sits idle (accruing cost), future materials then need to be delayed (or take up space on site), future contracting needs to be adjusted (admin costs), and so on. The more complex the project, the more surface area there is for these cascading delays to drive the cost of construction up. Furthermore, the more something gets delayed, increasing the project timeline and thus the likelihood of future delays. As our advisor René Morkos, Ph.D., CEO of ALICE Technologies (a construction planning company) and adjunct Construction / AI professor at Stanford, says, “90% of projects are delayed and over-budget.”
One recent example of this type of shock was during the first year of the COVID-19 pandemic. Prices of processed wood products, such as softwood lumber and plywood, nearly quadrupled. Wholesale prices for plywood increased from $400 to $1500 per thousand square feet. This triggered a worker shortage in the forest products industry, which led to limited availability of wood products, such as softwood lumber and structural panels. The domino effect continued with supply chains further disrupted by a lack of truckers to move materials.
Worker shortages then led to an increase in average industry wages. Higher pay, however, did not result in enough extra workers to make up for increased demand and limited supply. High-level skills are increasingly needed to work with advanced machines making labor harder to source. This problem is not limited to the timber supply chain. It extends to the construction industry more broadly. In 2022 there was an estimated shortage of 650,000 skilled construction workers. This number is expected to increase to 3M by 2026. Needless to say, these labor shortages represent a massive vulnerability to volatility at a time where we need to be building faster than ever. There is currently $1.1T of unmet demand for housing in the US alone.
Material and labor shortages are not the only challenges. It is estimated new regulations in US construction each year cause a ~1% annual cost increase in single-family home construction. Given all these factors, is it at all surprising that the only thing construction costs have done since the end of WWII is go up?
So what are we going to do about it? We see two paths … Earthshots and Moonshots. Our first investment in Hedral (highlighted in our last update) is an Earthshot tackling these problems. Building on advancements in machine learning and scalable cloud computing, Hedral is able to automate the structural design process for large structures. A process that used to take weeks, now can take a day. This allows designers to explore multiple scenarios simultaneously (ie building with timber or steel). This means that shocks like COVID can be more easily planned around.
ALICE Technologies (highlighted above), is another Earthshot tackling these problems. With advancements in constraint programming and AI, ALICE can use existing construction schedules to simulate millions of outcomes and options for a given project to optimize for cost and time. This active monitoring keeps projects on-time and on-budget, avoiding incidents where possible and mitigating the consequences via rapid response when needed.
Hedral and ALICE are both Earthshot companies because they build on existing technologies. The determining factor for their success is how quickly they will be able to penetrate the market. Distribution is key and both are de-risking it. Hedral is operating as a tech-enabled business to more quickly win contracts and drive ROI. Forsight minimizes switching costs and onboarding for developers by leveraging existing cameras on sites.
On the other side of the coin are Moonshot companies like Cuby, Sublime, and Strong By Form. Each of these are taking large technical risks to develop products that (if successful), will gain traction and provide resilience against volatility in construction.
- Cuby is developing transportable factories that function as an end-to-end building platform, mass-producing and delivering single-family and multi-family residential homes at up to 50% of the cost and 10x the productivity output (sq / ft) per unit of labor. The technical risk is building a fully manufacturable housing construction system (integrating hardware and software) that can be implemented by low-skill workers.
- Sublime Systems is developing and commercializing an alternative to the standard Portland cement that does not require carbon-intensive, high-heat kilns. Instead Sublime's concrete is synthesized using electrified processes while meeting the performance and cost requirements for modern structures. The technical risk is developing energy efficient chemical processes for producing Sublime cement at the massive scale Portland cement is produced today.
- Strong By Form (shoutout to JT!) is creating lightweight structural solutions that fuse the sustainability of wood with the performance and productivity of advanced composites. This results in a viable alternative to metals for construction. The core technical risk is producing these novel bio-composites as well as the systems to scale their utilization.
As the construction industry adopts technologies to mitigate its large vulnerabilities to volatility, we will be using the Earthshot / Moonshot framework to construct a portfolio that stands to generate outsized returns for our LPs.