Interview: Replicating Private Equity Returns In Public Markets with Verdad’s Brian Chingono

Episode Summary

Executive Summary: Brian Chingono of Verdad Advisors discussed how private equity returns are largely explained by small-cap and value exposure plus leverage, not persistent manager skill alone. He described Verdad’s research-based strategy of buying small, cheap, levered companies and using machine learning to refine selection, plus work on predicting credit rating changes and timing value exposure with high-yield spreads.

Main Topics: Verdad’s origin and strategy (Priority: 5/5): Brian Chingono explained how Verdad grew from a Harvard connection and a private-equity research thesis into a firm built around systematically buying small, cheap, levered public companies that resemble LBOs. What drives private equity returns (Priority: 5/5): The conversation argued that private equity’s historical excess returns are mainly attributable to small-cap and value exposures, with leverage adding risk and return, rather than broad industry skill or a large illiquidity premium. Replicating private equity in public markets (Priority: 5/5): Chingono described the Leveraged Small Value Equities paper, which found that public-market companies with PE-like characteristics can deliver returns similar to or higher than private equity over long horizons. Machine learning in the investment process (Priority: 4/5): Verdad uses ML as a complement to linear factor models, especially to identify nonlinear patterns and interactions in residuals, and to improve deleveraging signals and stock selection. Value investing and market timing (Priority: 4/5): The discussion defended value as a persistent long-term premium and argued that very wide valuation spreads and high-yield spread spikes can signal especially attractive entry points for value exposure. Credit market research and bond rating prediction (Priority: 3/5): Verdad applied random forest and other ML methods to forecast upgrades and downgrades across the credit spectrum, finding that downgrades exhibit nonlinear patterns that are better captured by ML than linear models. Internship program and talent development (Priority: 2/5): The firm’s internship program combines academic-style classes, debates, and company analysis while also supporting interns’ career development and placement.

Key Arguments: Private equity’s net returns in the mid-teens are not as mysterious as they appear; after adjusting for size and value, much of the premium disappears. Historically, PE bought small, cheap companies; in public markets, similar portfolios can replicate those returns using leveraged small value stocks. Any persistent alpha in private equity depends on both skill and discipline around fund size and valuation; scaling too much can erode returns. The main observable effect of PE ownership in the data is higher leverage, while revenue growth, margins, and capex changes look close to random. Machine learning is most useful when layered on top of linear factor models, not used as a replacement for them. Value has not died; its long-term evidence base is far stronger than a single decade of growth outperformance, and today’s valuation spreads suggest attractive expected returns. High-yield spreads above 6% are interpreted as crisis-level fear, creating unusually good forward returns for value even though the window may be brief. In credit, ML can add substantial value because downgrade behavior is nonlinear and credit data is less abundant and less standardized than equity data.

Data Points: Private equity net return: mid-teens (~15% net of fees) - Broad benchmarks such as Cambridge Associates over long periods S&P 500 long-run return: around 10% per year - Used as the public-market comparison for PE Excess return of private equity vs S&P 500: around 5% per year - Initial gap before adjusting for size and value effects Private equity equity check size: about $200 million of equity - Average size of private equity deals mentioned Median S&P 500 market cap: around $40 billion - Illustrating the size mismatch versus PE PE purchase valuation historically: about 7x EBITDA - Average private equity valuation in the 1980s and 1990s Public market valuation historically: about 10x EBITDA - Comparative valuation level for public markets historically Today’s PE purchase valuation: about 12x EBITDA - Indicates PE has become more growth-like as capital flooded in Plausible public-market replication return: about 20% annualized - Claim for public-market leveraged small value over a long period Companies below 7x EBITDA in study: about 25% of companies - Subset of the private equity deal database examined Share of industry profits from those companies: more than half - Those under-7x-EBITDA deals generated most profits LBO dataset size: 390 deals - Public-debt-backed LBOs used to test PE operating claims LBO dataset enterprise value: about $700 billion - Scale of the deal dataset used for analysis Revenue growth effect in PE-owned firms: about 40% increased, 54% slowed - Evidence that revenue growth changes looked roughly like a coin flip EBITDA margin effect: about 55% expanded, 45% contracted - Marginal operating improvement was not strongly systematic Investment/capex effect: about 45% increased, 55% declined - Capital investment changes also looked roughly random Leverage increase after PE ownership: about 70% of deals - Most consistent systematic change observed after buyout Private equity firm flow trend: allocations have ballooned since the 1990s - Capital influx cited as a driver of higher valuations Value history in Ken French data: since 1926 - Long-run evidence cited in defense of value investing High-yield spread crisis threshold: above 6% - Signal used to identify especially attractive times to add value exposure Value recovery window after spread blowout: about 2 years - Typical market-time recovery period cited Credit dataset start year: 1997 - Used in the rating-change prediction work Upgrade model definition: at least one-notch improvement - Examples included single-B- to single-B or BBB+ to BBB Downgrade model definition: move from BB to BB- or below - Or more than one notch in the downgrade framework Intern outcome: 2 interns became Rhodes Scholars - Used to illustrate the internship program’s impact

Pivotal Quotes: "what you find is that the vast majority of the profits in the private equity industry were generated from companies that traded at less than seven times EBITDA" — Brian Chingono: Explaining the original private-equity research that led to Verdad’s strategy "private equity is really delivering small value exposure" — Brian Chingono: Summarizing the core conclusion from benchmarking PE against appropriate public-market factors "the expected return on value is higher today than it was historically" — Brian Chingono: Arguing that wide valuation spreads imply attractive forward returns for value stocks

Implications: The episode suggests investors should benchmark private equity carefully, use factor-aware models, and recognize that patient value exposure may be especially compelling after drawdowns and spread blowouts. It also shows how ML can sharpen, not replace, disciplined long-term factor investing.