The 'company age' factor is not predictive on its own. MDT's decision tree model uses it to create context, asking different questions about young companies versus mature ones. For example, valuation proves to be a much more important factor for older, established businesses.
Related Insights
Contrary to the 'get in early' mantra, the certainty of a 3-5x return on a category-defining company like Databricks can be a more attractive investment than a high-risk seed deal. The time and risk-adjusted returns for late-stage winners are often superior.
Traditional valuation models assume growth decays over time. However, when a company at scale, like Databricks, begins to reaccelerate, it defies these models. This rare phenomenon signals an expanding market or competitive advantage, justifying massive valuation premiums that seem disconnected from public comps.
For consumption-based models, simple size-based segmentation (SMB, Enterprise) is insufficient. Stripe and Vercel use a two-axis model: company size (x-axis) and growth potential (y-axis). A small company growing at 200% YoY is more valuable and warrants more sales investment than a large, stagnant one.
In AI M&A, recency is key. Companies pre-ChatGPT often had to rewrite their entire stack and relearn skills, making their experience less relevant. Acquiring a company with post-ChatGPT experience ensures their tech and knowledge are current, not already obsolete.
Rather than building one deep, complex decision tree that would rely on increasingly smaller data subsets, MDT's model uses an ensemble method. It combines a 'forest' of many shallow trees, each with only two to five questions, to maintain statistical robustness while capturing complexity.
Instead of opaque 'black box' algorithms, MDT uses decision trees that allow their team to see and understand the logic behind every trade. This transparency is crucial for validating the model's decisions and identifying when a factor's effectiveness is decaying over time.
The firm doesn't just decide a factor is obsolete. Their process begins by observing within their transparent 'glass box' model that a factor (like book-to-price) is driving fewer and fewer trades. This observation prompts a formal backtest to confirm its removal won't harm performance.
This provides a simple but powerful framework for venture investing. For companies in markets with demonstrably huge TAMs (e.g., AI coding), valuation is secondary to backing the winner. For markets with a more uncertain or constrained TAM (e.g., vertical SaaS), traditional valuation discipline and entry price matter significantly.
While many investors screen for companies with high Return on Invested Capital (ROIC), a more powerful indicator is the trajectory of ROIC. A company improving from a 4% to 8% ROIC is often a better investment than one stagnant at 12%, as there is a direct correlation between rising ROIC and stock performance.
MDT deliberately avoids competing on acquiring novel, expensive datasets (informational edge). Instead, they focus on their analytical edge: applying sophisticated machine learning tools to long-history, high-quality standard datasets like financials and prices to find differentiated insights.