Rethinking Risk w/ Charlesbank’s Michael Choe

Michael Che, CEO of Charles Bank, critiques traditional five-year LBO models, highlighting their inherent flaws and human biases. He advocates for a probabilistic approach to risk management, focusing on two-year performance metrics to better predict outcomes. This method, which considers a range of scenarios and emphasizes asymmetric upside and downside risks, encourages investment teams to think more critically about potential outcomes. By fostering a culture of probabilistic thinking, Charles Bank aims to enhance decision-making processes, improve risk assessment, and accelerate learning within the organization, ultimately leading to more informed investment strategies.

• Traditional five-year leveraged buyout (LBO) models are often inaccurate, as they focus on precise growth rates and internal rate of returns (IRRs) that do not account for real-world variability.
• These models tend to fixate on specific entry multiples, which are merely snapshots in time and do not reflect the full range of potential outcomes.
• There is a tendency for models to cluster around a base case, neglecting significant divergences that can occur both to the downside and upside.

"Whenever we looked at a traditional five-year LBO model, we would say something like, well, the one thing we know for sure about this model is that it's not going to be right."

• This quote highlights the inherent uncertainty and inaccuracy of traditional LBO models, emphasizing the need for a more flexible approach.

"You're fixated on artificially precise five-year growth rates. You're fixated on artificially precise five-year IRRs. You get trapped in looking at endless variations of what your entry multiple is, which is just a snapshot in time."

• This comment underscores the limitations of focusing on precise metrics in a model that cannot accurately predict long-term outcomes.

• Human biases, such as anchoring bias, often influence investment models, leading to overly optimistic projections.
• There is a natural tendency to underestimate downside risks, with models rarely predicting capital impairment despite historical occurrences.

"What you're observing there we think is really kind of a natural human tendency of having anchoring bias."

• This quote points out the influence of human bias on investment modeling, suggesting that it skews projections toward more optimistic outcomes.

"I have never seen a downside case that had capital impairment as a result. That means, you know, every single downside case was a positive result."

• This observation reveals a common bias in investment models where downside risks are underestimated, potentially leading to unexpected capital losses.

• The new approach involves analyzing two-year performance as a predictor of five-year outcomes, allowing for more precise and actionable insights.
• By focusing on a shorter time horizon, teams are encouraged to think about specific upside drivers and downside risks with greater precision.
• This method helps in understanding the probability distribution of outcomes, such as the impact of macroeconomic events like recessions.

"Instead of trying to predict five years...why don't we actually discuss two-year probabilities?"

• This quote suggests a shift in focus from long-term predictions to more immediate, two-year forecasts, which can be more precise and reliable.

"Once we moved the conversation into a two-year modeling time horizon, well then you start talking about real specifics."

• This statement highlights the benefits of a shorter forecasting period, which allows for more detailed and actionable analysis.

"A recession in that time period will actually show you how macro exposure could really negatively impact the company in any time period."

• This quote illustrates the importance of considering macroeconomic risks within a shorter time frame, providing a more realistic assessment of potential impacts.

• Investment models are shifting focus from traditional base cases to considering asymmetric upside and downside outcomes.
• Teams are encouraged to think probabilistically about potential outcomes, evaluating the probability of negative EBITDA scenarios.
• The emphasis is on understanding the probability of impairment cases, which traditional models often overlook.

"There's a couple of probabilities that we are forcing teams to express. One probability is to ask them if you ran this company or invested in this company in 100 parallel universes, in what percent of those cases are we experiencing impairment as a result of EBITDA decline?"

• This quote highlights the new approach of assessing risk by considering the probability of negative outcomes across multiple hypothetical scenarios.

• Teams are required to calculate the likelihood of both downside and upside scenarios using probabilistic models.
• The focus is on understanding the percentage of instances where a model predicts impairment or a significant positive return, such as a 30% gross IRR over two years.
• This approach provides a more nuanced understanding of risk and potential returns compared to traditional LBO models.

"We are forcing teams now to actually explicitly show the percentage of instances in which the model is producing an impairment or a decline in EBITDA."

• This quote emphasizes the importance of explicitly calculating and presenting the probability of negative financial outcomes.

• Introducing probabilistic models poses challenges due to their complexity and the additional workload they create.
• The resistance to adopting these models is mainly due to the physical difficulty of integrating them into existing processes within tight timeframes.
• Despite these challenges, the new approach provides significant insights, especially in identifying high volatility risk events.

"The resistance hasn't so much been that teams might be forced to admit that the distribution is wider than what they might have modeled in the past. But it's really more the physical difficulty of adding a brand new analytical tool that's much more complicated."

• This quote explains that the main challenge in implementing these models is their complexity and the additional workload they entail, rather than reluctance to acknowledge wider risk distributions.

• Investment committees often analyze the forward price curve of commodities like oil and assess the standard deviation of prices to anticipate market trends.
• Human biases can lead to overconfidence in certain outcomes and underestimation of unpredictable variables.
• Example: A chemical carve-out with significant pre-tax income may seem promising, but macroeconomic factors like oil price fluctuations introduce unforecastable risks.
• Investment strategies now focus on avoiding companies with imponderable risks that cannot be precisely forecasted.

"The human mind gets seduced by the certain and tends to discount some of the imponderable variables."

• This quote highlights the cognitive bias towards certainty, often leading investors to overlook unpredictable factors.

"The conversation has changed a lot for us on companies like that where there's just imponderable risk relating to macro factors that we just don't have the ability to forecast with any precision."

• The quote explains the shift in investment strategy to avoid companies with unpredictable macroeconomic risks.

• The Phantom Outcomes Tool aids in making faster and higher fidelity investment decisions by assessing a broader range of potential outcomes.
• Example: Analyzing a software company for a take-private deal revealed risks of negative growth, despite potential cost efficiencies and market position.
• Acknowledging the proximity of zero growth to negative growth helps in honest risk assessment and probability distribution.

"0% growth is really close to negative growth. Right? 0% is adjacent to a company that is shrinking."

• This quote underscores the importance of recognizing the risk of stagnation leading to decline, affecting investment decisions.

"We had to be honest about that in the probability distributions."

• The quote emphasizes the necessity of incorporating realistic growth probabilities in investment analyses.

• Investments are evaluated based on their probability of achieving a 30% rate of return over two years, providing a benchmark for success.
• The goal is to minimize impairment risks while maximizing potential returns, starting with quantitative benchmarks.
• A rich dialogue about management partnerships, capability matches, and other intangibles follows the initial quantitative assessment.

"We're rallying around is trying to filter and really home in on investment return distributions that give us a greater than average probability of beating a 30% rate of return over a two year period of time."

• This quote outlines the primary financial goal and benchmark for evaluating potential investments.

"That subsequent conversation is much richer when we are actually starting with the basis of what really drives upside and downside in a business."

• The quote highlights how quantitative benchmarks lead to more meaningful discussions on investment potential and risks.

• The process involves back testing the current tool to evaluate past predictions and decisions.
• This retrospective analysis helps improve input distributions and decision-making accuracy.
• The use of "fan of outcomes" language has permeated decision-making, fostering a more precise and focused analysis of potential decisions.

"The early results of that exercise suggest that we were more right than wrong. And for sure it helped us make better decisions three, four years ago than we would have using simple LBO models."

• This quote highlights the effectiveness of the current tool in improving decision-making accuracy compared to traditional models.

"Every decision we make, what we are noticing is that our team members are starting to use fan of outcomes language to describe the decision."

• The adoption of "fan of outcomes" language illustrates a cultural shift towards more nuanced and detailed decision analysis.

• The decision-making process has evolved organically, with mid-level professionals increasingly using fan of outcomes thinking.
• This shift has led to a bottom-up flow of ideas, enhancing the diversity and quality of investment ideas.
• The approach accelerates pattern recognition among less experienced team members, improving their decision-making capabilities.

"Our VPs, senior vice presidents, principals, our mid-level professionals, have started to use fan of outcomes thinking to triage ideas and filter them up to the top."

• This quote emphasizes the organic change in idea flow, with mid-level professionals playing a more active role in decision-making.

"That direction of idea flow going from mid-level upwards...is super exciting for us."

• The excitement about this change indicates its positive impact on the organization's decision-making processes.

• The ultimate goal is to instill probabilistic thinking across the organization to enhance collective decision-making.
• This approach has led to vigorous debates about the probabilities of various decisions, enriching the decision-making process.
• The cultural transformation is not just about using a tool but changing the organization's entire approach to decision-making.

"The fundamental objective for us is to think probabilistically and by thinking probabilistically, think better together."

• This quote underscores the importance of probabilistic thinking in improving collaborative decision-making.

"It's really cultural transformation approach that you're talking about. It's a way of thinking that's evolving the entire organization's approach to decision making."

• The quote highlights the broader cultural shift towards a more analytical and thoughtful decision-making process.