Kindred Capital 2019 Investor Letter Excerpt: At the core of our investment process is Equity Data Science (EDS), a software program which affords us enormous leverage in stock selection and in managing the Fund’s investments.
Steve developed an early version of the platform while at Herring Creek, and over the last two years, together with his former colleague and partners (who created a new venture to market the software to other investment managers (for full disclosure, Steve is the largest outside investor in EDS). Over time, we have continued to fine tune and enhance the platform’s capabilities and data mining power.
While initially designed to augment idea generation and stock selection, the latest version of EDS now incorporates tools that help us with portfolio construction and position sizing as well. EDS filters the equity universe in multiple ways: on the basis of valuation relative to peers, relative to the market as a whole, and relative to a company’s own history. It tracks technical trends, leverage, returns, growth, market sentiment, and many other financial measures. It also estimates normalized earnings power based on historical data. Using EDS, we rank securities by quintile from 1 to 5 (best to worst) to define our opportunity set.
To illustrate, with EDS we rank securities using four basic criteria which have generally proven predictive of outperformance: Valuation, Capital Deployment, Earnings Quality and Momentum. Under this construct, potential long investments would typically rank as a 1 in at least the first three categories and potential shorts would rank 5. Of course there are always exceptions to the rule. Value traps whose earnings power may be permanently impaired can screen as a 1 on value (think Kodak). Conversely, a company recovering from a steep decline in earnings can look expensive (a 4 or 5) when its future earnings outlook is bright. Acquisitions score negatively on capital deployment, but can sometimes prove highly accretive or transformative and so on. In this way, EDS serves as a highly effective screening tool, whose scores we then evaluate against a deeper analysis of the company’s history and prospects.