Our Investment Process
Our ideas typically comprise companies exposed to big-picture secular themes around which we have built conviction and/or companies with business models that we are familiar with from prior analysis. Some of these include the transition to the public cloud, the expanding need for and use of data, the continued transition of commerce/advertising online, the increasing reliance on bandwidth and internet infrastructure, the diminishing use of physical cash in society, rising disposable income and consumption in China, and increasing access to global travel. We prioritise companies where we have preliminary indications of business quality and where we think our existing work might help maximise our return-on-effort. It is a bonus if we suspect the valuation will be attractive, but this is not imperative — we are happy doing our work and then being patient for future valuation opportunities.
“I very frequently get the question: 'What's going to change in the next 10 years?' And that is a very interesting question; it's a very common one. I almost never get the question: 'What's not going to change in the next 10 years?' And I submit to you that that second question is actually the more important of the two because you can build a business strategy around the things that are stable in time.” — Jeff Bezos
Like most primary research processes, we seek to understand the core business model, the investment landscape, and competitive position. We use a range of resources including company materials, industry periodicals, academic journals, government publications, survey data, podcasts, books, and expert call transcripts. We source industry experts across the value-chain through personal networks and expert network firms but are careful to target the experts we need rather than simply relying on who is available. We take a first principles approach and prioritise the research process around the areas that are most impactful and in which we have lower conviction. However, all of this is simply “table stakes”.
What makes our process unique to Kalakau Avenue is the synthesis of this research into our business quality framework. We score all companies across this common framework to arrive at an overall business quality score that determines its cost of capital. This is not to make the process quantitative but rather to promote consistency of decision-making across the portfolio and over time.
In addition to our business quality framework, we attempt to understand why the business needs to exist and the few things that really matter. We narrow down the key risks and unknowns, including the regulatory environment, and try to understand their likelihood and impact.
All our research, investment frameworks, investment decisions, meeting notes, team priorities, and contacts etc are logged in our custom-built research management system in real time for easy access and recall. Our research management system is a key facilitator of shared learnings among the team and represents a key piece of our infrastructure.
We see valuation first and foremost as a risk mitigant rather than a return driver. That is, we are not trying to capture the reversion of mispricing in the market. Our returns mostly come from the long-term compounding of intrinsic value. We use a variety of methods to value our portfolio companies, all aimed at answering one question, “does a margin-of-safety exist?”.
Valuation as a tool is a sledgehammer not a scalpel.
In general, we model our companies to the point they are likely to stabilise (typically 5 years but this could be longer) and assess the IRR over a range of scenarios, always thinking probabilistically. We strive to build solid forecasts within each scenario and to understand the broad probability of each scenario. However, we also try hard to avoid false precision; valuation as a tool is a sledgehammer not a scalpel. We have more confidence in our views of the business quality — that is, the characteristics we see today that increase the odds of success no matter what the future holds.
We assign a valuation / risk-reward score by assessing the range of returns versus the cost of capital. Again, this is not to make the process quantitative but to help promote consistency and tracking of decision-making. If a company passes our quality filter but is too expensive or not more attractive, risk-adjusted, than the current portfolio then it simply goes on our investable list.
<aside> <img src="https://img.icons8.com/ios/250/000000/light-on.png" alt="https://img.icons8.com/ios/250/000000/light-on.png" width="40px" /> Our Valuation Principles
1. Demand a margin-of-safety always: Continuing to own a liquid asset is the same as buying it. We ensure our assets are always held with a margin-of-safety and not just at the point of purchase. Our criteria for buying an asset are the same as our criteria for continuing to hold an asset.
2. Range-based probabilistic thinking: Because predicting the future over the long-term is incredibly difficult, our valuation considers a broad range of scenarios to get a broad sense that a margin-of-safety exists. We strive to always think probabilistically.
3. Avoid false precision: Much of the future is unpredictable and even more so the longer the time horizon. With our long-term investment horizon, we work hard to understand what scenarios the future may hold. However, we must also recognise what is inherently unknowable and the subsequent inaccuracy of our forecasts. We try to avoid false precision.
A big part of our process is understanding what will remain the same. We seek to understand what we are really backing over the long-term and what we can rely on. We want to back the best economic machines that are built around things we think will stay the same and that have the flexibility to adapt to whichever future scenario unfolds.
A simple example of this concept is a weighted coin that lands 60% of the time on heads and 40% on tails. We can have strong conviction that it is a good strategy to bet on heads but we cannot have strong conviction that the coin will actually land on heads. A real life example is Meta Platforms (”META”). META’s constant is that humans are a social, community orientated species. The long-term business, therefore, needs to be orientated around providing community in any form that users prefer. Exactly how users will communicate and network in 10 years is not predictable today but we can predict the continued time-served need for community. To invest in META, one has to believe the platform has an advantaged position that can adapt over time to provide this need in the form users want. For Google search, as another example, the constant is the need to efficiently access information. Again, predicting how this will be done in 10 years is very difficult and so to invest in Google, one has to believe in Google’s advantaged position to adapt the platform to suit users over time.
</aside>
Positions are sized based on conviction and risk-reward. Conviction is developed through our research and the risk-reward is determined with our valuation score, by comparing the range of potential returns versus the cost of capital. Further, sizing can never be decided in isolation from the rest of the portfolio, there is always a relative trade-off. We maintain a sizing framework that outlines suggested sizing ranges based on our quality and valuation scores. As before, the purpose of this framework is not to make the process quantitative and in any case, the framework is only a general guide. Rather, it’s purpose is to eliminate bias and inconsistency in our decision-making across the portfolio and over time.
The portfolio is adjusted only when there are very material changes in the risk-reward profiles. Positions are sold entirely if the investment is unlikely to compensate for the risks even over the long-term, if the quality thesis deteriorates or is proven wrong, or if the portfolio is fully invested and a competing idea takes precedence.