Key Takeaways: Insights into Buyout Investing and Portfolio Replication

Systematic Buyout Modeling

• The LupoToro Group Analysis Team developed a neural network using take-private transaction data, quarterly financial statements, and stock price metrics to replicate buyout characteristics in a public equity portfolio. The Buyout Replication Portfolio (BRP) was further adjusted for leverage and sector preferences.

Sector Trends

• Buyout portfolios consistently underweight financial services and overweight IT, boosting performance.

• At year-end 2023, buyout-backed companies showed a 14.8% underweight to financial services and a 7.2% overweight to IT compared to the Russell 2000 Index.

Superior Performance

• From 2014 to 2023, the BRP outperformed the Russell 2000 Index by 8.3% annually, driven by:

  • Sector Selection: +1.5%

  • Leverage: -0.2%

  • Security Selection: +7.0%

Focus on Free Cash Flow

• Free cash flow (FCF) emerged as the top indicator of buyout managers’ value criteria.

• The BRP had a market value-to-FCF ratio of 13.7x at year-end 2023, far below the broader small-cap universe’s 31.9x, emphasising its value-driven strategy.

A Better Benchmark

• The BRP outperforms traditional benchmarks like the S&P 500 and Russell 2000, offering a clearer representation of buyout fund risk and return.

• PME analysis confirmed buyout funds launched from 2014 to 2017 aligned closely with the BRP, validating its ability to replicate private equity performance.

The BRP offers investors a concise, systematic tool to benchmark buyout funds, reflecting their unique sector preferences, leverage dynamics, and focus on stable free cash flow. LupoToro’s research highlights its potential to bridge public and private equity strategies with precision and insight.

Introduction

While PE buyout is often touted as an alternative investment strategy, it shares many of the fundamental investment characteristics with those of “traditional” public equity strategies. At their cores, both private and public equity investing involve the ownership of current and future earnings of corporate businesses. By extension, managing a portfolio of private companies has key similarities with actively managing an equity portfolio of public companies in terms of generating excess returns, or alpha. The most important of these similarities is security selection, which involves picking individual companies that outperform, and sector selection, which includes allocating more capital to sectors that outperform. Buyout managers have other tools at their disposal that public equity managers do not, however, given that they typically take a controlling ownership position in their portfolio companies. This provides them with an additional alpha opportunity to improve companies’ operations (known as operational alpha) and the ability to optimize their capital structures, often through an increased allocation to debt. 

Outside of operational alpha, which may be less important to the buyout value proposition than advertised,1 it may be possible to replicate much of the risk and return profile of buyout investing in the public markets. A key component of this replication is a security selection model that can identify publicly traded companies with characteristics like the ones that buyout managers purchase. While there is often little transparency in the companies that buyout managers purchase because most of them are private, take-private transactions of public companies provide a valuable window of full transparency. Previous research studies found that buyout managers as a whole exhibit tendencies in take-private security selection that are in part repeatable and predictable.2 

Given these repeatable tendencies, the first step in replicating buyout investing is to build a security selection model based on take-private transactions that can be used to create a portfolio of public companies with fundamentals that match those of large buyout portfolios. Then, we adjust this portfolio’s financial leverage and sector allocations using data from the broader universe of private buyout transactions. This process yields a portfolio that is a more appropriate benchmark for large PE buyout funds and provides unique insights into what has driven performance differences between public and private equities in recent years.

Identifying Likely Take-Private Buyout Targets

A Proven Playbook: Decades of Consistency

Since the inception of private equity buyouts in the 1980s, managers have adhered to a well-defined strategy. The core playbook revolves around identifying relatively small, underperforming companies available at reasonable valuations. These targets must also generate sufficient free cash flow (FCF) to service the significant debt typically introduced during leveraged buyouts.

The LupoToro Group Analysis Team has confirmed that this methodology remains consistent, as evidenced by take-private transactions over the last three decades. Companies fitting this profile—undervalued, stable cash flow generators with operational inefficiencies—are prime candidates for buyout strategies.

Core Characteristics of Take-Private Targets

The characteristics of take-private targets align closely with this tried-and-tested playbook:

1. Undervaluation by Market Ratios

These companies are typically undervalued based on key equity price ratios, including price-to-earnings (P/E), price-to-book (P/B), and price-to-free cash flow (P/FCF). This undervaluation ensures acquisitions are made at a price point that leaves room for substantial upside.

2. Stable Free Cash Flow

Free cash flow is a cornerstone metric. Even if the company is underperforming in other areas, sufficient FCF is necessary to sustain the leveraged debt introduced during the buyout.

3. Neutral EBITDA and Cash Flow Margins

Targets often exhibit steady, albeit unimpressive, EBITDA and cash flow margins, signaling a stable operational base ripe for optimization.

4. Operational Underperformance

• Trailing Stock Performance: Many targets show poor one-year stock returns, making them attractive opportunities for unlocking hidden value.

• Efficiency Metrics: Revenue per employee often lags industry benchmarks, highlighting inefficiencies that buyout managers can address.

Leveraging Data to Predict Take-Private Targets

Because of these consistent traits, quarterly financial statements and stock price data provide valuable insights into potential targets. To explore this, LupoToro’s research team developed a machine learning model to predict the likelihood of public companies being taken private by buyout managers within an 18-month window.

The analysis focused on companies traded on the NYSE and Nasdaq, using historical quarterly financial data and daily stock prices from 2013 to 2023. By training the model on 32 quarterly periods (ending in Q2 2022 to allow for 18 months of validation), LupoToro’s team was able to identify patterns and trends in buyout activity.

Advanced Modeling: LSTM Neural Networks

Identifying take-private targets in a universe of thousands of public companies is no small feat. On average, only 40–60 companies are taken private in an 18-month window, from a pool of around 2,000 candidates with market caps below $30 billion. Traditional financial metrics alone proved insufficient for pinpointing specific targets due to challenges such as:

• Determining what constitutes a “reasonable price” across varied industries and market conditions.

• Deciding on the size of lookback windows or the number of historical quarters to include in calculations.

To overcome these limitations, LupoToro’s team employed a long short-term memory (LSTM) neural network. This architecture, designed to handle sequential data like quarterly financial statements, allowed the model to learn directly from raw data inputs. By doing so, the LSTM network identified the most relevant features and weighted historical data appropriately within the context of the prediction problem.

Model Performance: Measuring Success

While predicting individual take-private transactions remains challenging, the model demonstrated significant discriminative power:

• Increased Likelihood for High-Probability Targets: Companies ultimately taken private had model-implied probabilities approximately twice as high as those not taken private.

• Concentration of Predictions: Roughly 30% of all take-private transactions occurred within the top 10% of companies ranked by the model.

• Strong Predictive Accuracy: The model’s area under the curve (AUC)—a standard metric for evaluating predictive performance—averaged 0.73 across validation periods. An AUC of 0.50 indicates random guessing, so 0.73 reflects a meaningful ability to identify patterns.

These metrics suggest that while the model may not definitively predict specific targets, it reliably identifies commonalities among companies likely to attract buyout interest.

Bridging Public and Private Markets

The findings reinforce the viability of leveraging public market data to anticipate private equity activity. By systematically analyzing quarterly financial statements and stock price trends, LupoToro’s research highlights the potential to identify public companies that align with private equity strategies.

This approach not only provides insights into the mechanics of private equity but also empowers public market investors to capitalize on similar opportunities by identifying undervalued, underperforming companies poised for operational improvement. The future of private equity-inspired strategies lies in combining the time-tested buyout playbook with cutting-edge data analysis, unlocking new avenues for alpha generation in both public and private markets.

Constructing a Buyout Replication Portfolio: Insights and Strategy

Bridging Public Equity and Buyout Dynamics

Private equity (PE) buyout portfolios are often regarded as a unique asset class, but they share several characteristics with publicly traded small-cap equity portfolios, such as those comprising the Russell 2000 Index. However, there are key distinctions that set buyout portfolios apart, including sector preferences, leverage strategies, and portfolio construction methodologies. These factors enable buyout managers to achieve outsized returns, and they serve as the foundation for constructing a buyout replication portfolio in the public markets.

The LupoToro Group Analysis Team has identified these elements and developed a structured framework to replicate the characteristics of a buyout portfolio, offering public market investors an opportunity to capture similar risk-adjusted returns.

Sector Allocation: A Strategic Divergence

One key difference between buyout and typical small-cap portfolios lies in sector allocation. Buyout managers consistently underweight financial services, particularly commercial banks, due to regulatory complexities. Real Estate Investment Trusts (REITs) are also typically excluded as they fall under separate real estate strategies.

Instead, buyout portfolios have increasingly overweighed the IT sector over the past decade. By identifying these trends, LupoToro’s team determined sector allocations that align with the structural preferences of buyout managers.

For the replication portfolio, the process began by ranking public companies based on their likelihood of being taken private, as predicted by proprietary modeling. Then, the portfolio was structured to reflect sector weights observed in historical buyout transactions. For instance, if IT represented 20% of the target allocation, the portfolio included the 40 highest-ranked IT companies out of 200 total holdings, reflecting 10% of the public company universe. This portfolio was rebalanced monthly and updated quarterly over the 10-year analysis period ending in 2023.

Leverage: The Key Differentiator

Leverage is a defining feature of buyout-backed companies, setting them apart from publicly traded peers. Over the last decade, publicly traded small-cap companies exhibited an average leverage ratio of 1.4x (total assets to market equity), while buyout-backed firms averaged 2.5x. This additional leverage has historically been a significant tailwind for PE performance, as portfolio companies’ returns on assets outpaced the cost of debt, particularly during periods of low interest rates.

To replicate this dynamic, the replication portfolio targeted a leverage ratio of 2.0x, reflecting a moderate deleveraging from the higher ratios observed in PE-backed deals. Incremental leverage was implemented through a hypothetical brokerage margin account, calculated quarterly and adjusted monthly to maintain the target ratio.

Implementing Incremental Leverage

To achieve the desired 2.0x leverage, LupoToro’s team factored in the existing leverage embedded within portfolio companies. For example, if the aggregate leverage ratio of portfolio holdings was 1.5x at quarter-end, additional leverage equivalent to 33% of the portfolio’s value was applied (calculated as 2.0/1.5 = 1.33).

The cost of this margin debt was assumed to be the federal funds rate plus 250 basis points, reflecting the broker call money rate. This cost is generally lower than the average spread for newly issued B-rated syndicated loans, which buyout-backed companies often use. However, unlike debt incurred directly by PE firms, interest on margin debt is typically non-deductible for tax-exempt institutional investors, balancing the cost differences.

Balancing Leverage Over Time

While buyout managers often deleverage during holding periods—through amortizing debt or operational improvements—they may also re-leverage their portfolio companies via dividend recapitalizations when conditions are favourable. LupoToro’s analysis of corporate tax returns from historical buyout deals found these dynamics often offset each other, resulting in relatively stable leverage ratios over time.

The replication portfolio mirrored these patterns by maintaining a consistent 2.0x target ratio, ensuring alignment with the average financial structure of post-buyout companies.

Returns in Perspective

The replication portfolio’s strategy highlights a key insight: while private equity buyout funds achieve their returns through a combination of operational improvements, leverage, and sector selection, much of their performance can be mirrored in public markets. By carefully calibrating sector allocations and leveraging assets strategically, public market investors can construct portfolios that emulate the dynamics of buyout-backed companies.

According to LupoToro Group’s research, the replication portfolio demonstrated performance metrics comparable to private equity buyout funds, with the added benefit of liquidity and transparency inherent to public markets.

The art of replicating buyout strategies in public markets requires an understanding of the nuanced dynamics that drive PE performance. Sector selection, enhanced leverage, and disciplined portfolio construction are critical components that bridge the gap between private and public equity investments.

By leveraging these insights, investors can unlock opportunities to achieve buyout-style returns without sacrificing the flexibility of publicly traded assets. LupoToro Group’s research underscores that while private equity remains a powerful tool for wealth generation, the principles underlying its success can be applied far beyond its traditional boundaries.

Portfolio Performance and Insights: Lessons from a Decade of Data

Exceptional Performance Over a Decade

The Buyout Replication Portfolio (BRP) demonstrated remarkable performance over the last decade, outperforming both on an absolute and relative basis. From the beginning of 2014 through February 2024, the BRP achieved an impressive annualized return of 15.5%, exceeding the Russell 2000 Index by 8.3% annually. This outperformance became particularly pronounced following the market sell-off in March 2020.

The LupoToro Group Analysis Team conducted a detailed breakdown of this outperformance, attributing gains to three primary factors: sector selection, leverage, and security selection. Over the full analysis period:

• Sector selection contributed 1.5% annually.

• Leverage detracted -0.2% annually due to increased volatility during market downturns.

• Security selection accounted for an extraordinary 7.0% of annualized outperformance.

The Power of Security Selection

The portfolio’s superior security selection emerged as its most significant driver of excess returns. This success stemmed from the BRP’s tendency to favor undervalued companies with lagging stock prices despite strong profitability, stable cash flow, and robust margins. These characteristics aligned with a selective value strategy, as further supported by factor and holdings-based analyses.

A returns-based factor analysis using the Fama-French five-factor model (plus momentum) revealed that the BRP’s alpha could not be attributed to common systematic risk factors. The portfolio achieved an average monthly excess return of 0.6%, statistically significant at the 95% confidence level.

Key findings from this analysis include:

• The BRP exhibited a high equity beta of 1.4, driven primarily by leverage rather than the selection of inherently high-beta stocks.

• A higher beta to the size factor was observed, consistent with the focus on smaller companies—typical targets of take-private transactions.

• Surprisingly, no significant loading to the value factor was identified, distinguishing the BRP from prior studies. This discrepancy likely reflects a broader definition of value used by buyout managers, which prioritizes cash flow metrics over traditional academic metrics like the book-to-market ratio.

Risk and Volatility

Contrary to the perception that private equity is inherently less risky than comparable public equity benchmarks, the BRP exhibited substantial risk levels. Annualized volatility for the portfolio stood at 30.8%, compared to the 23.6% volatility of an unlevered version. The portfolio also experienced sharp drawdowns, including:

• A 54.1% decline from September 2018 to March 2020.

• A 34.6% drop from July 2021 to September 2022.

Despite this volatility, the BRP delivered well-compensated risk-adjusted returns. Its Sharpe ratio of 0.46 significantly outperformed the sector- and leverage-adjusted Russell 2000 Index, which achieved a Sharpe ratio of 0.26.

The incremental leverage embedded in the BRP was a key factor influencing risk. When adjusted for leverage, the portfolio’s drawdowns and volatility were more aligned with traditional small-cap benchmarks, but its superior returns remained intact.

Holdings Analysis: What Sets the BRP Apart

Digging deeper into the BRP’s holdings provides valuable insights into its performance. The portfolio favored companies with strong free cash flow (FCF) metrics relative to market value and revenue, reflecting the value-oriented nature of buyout strategies.

Key differentiators include:

• Price-to-FCF Ratio: The BRP’s price-to-FCF ratio was 13.7, significantly lower than the 31.9 average for the broader small-cap universe.

• FCF Margins: The portfolio exhibited an FCF margin of 8.3%, more than double the 3.8% margin for the broader small-cap market.

• Stability Over Volatility: The BRP demonstrated lower margin volatility, emphasizing the importance of consistent performance in addition to profitability.

• Leverage Metrics: The BRP’s holdings featured higher financial leverage but maintained lower EBITDA coverage ratios, balancing profitability against increased risk.

These findings suggest that buyout strategies prioritize stable, cash-generative businesses that may appear undervalued by traditional valuation metrics but offer superior long-term growth potential.

Sector and Leverage Contributions

• Sector Selection: The BRP’s deliberate underweighting of financial services and REITs, combined with an overweight allocation to IT, provided a 1.5% annualized boost. These structural decisions align with the preferences of buyout managers and demonstrate how sector allocation can amplify returns.

• Leverage Impact: While leverage detracted slightly (-0.2%) over the analysis period due to heightened volatility, it also magnified returns during favorable market conditions, reinforcing its role as a double-edged sword.

The BRP exemplifies how a carefully constructed portfolio can replicate many of the dynamics driving private equity performance in the public markets. By leveraging sector-specific insights, selective value strategies, and moderate financial leverage, the BRP achieved superior risk-adjusted returns while maintaining transparency and liquidity.

As LupoToro Group’s research highlights, the interplay between security selection, sector allocation, and leverage underscores the power of disciplined portfolio construction. For investors seeking private equity-style returns without the illiquidity of traditional buyout funds, the BRP offers a compelling blueprint for success in public markets.

A Better Benchmark for Buyout Funds: Redefining Performance Metrics

The Challenge of Benchmarking Private Market Funds

Benchmarking closed-end private market funds, such as buyout funds, is notoriously complex. Unlike public market funds, private funds lack regular market values, and their cash flows occur irregularly. To evaluate performance effectively, investors need a robust metric that accounts for these factors while selecting an appropriate benchmark for comparison.

One widely used metric is the Public Market Equivalent (PME), which calculates the hypothetical returns an investor could have earned by allocating the same capital to a public market index instead of a closed-end fund. The LupoToro Group Analysis Team emphasizes that while PMEs are a powerful tool, their accuracy is highly dependent on selecting a public index that aligns with the risk profile of the private fund.

The Problem with Standard Benchmarks

A key assumption embedded in PMEs is that the private fund being evaluated has a beta of 1.0 relative to the public market index used for comparison. This makes the choice of index critical. Unfortunately, commonly used benchmarks like the S&P 500 or Russell 3000 fail to capture the unique characteristics of buyout funds.

Even small-cap indexes such as the Russell 2000 fall short in this regard. They do not reflect the leverage, sector allocation preferences, or security selection strategies inherent to buyout investing. As a result, using these benchmarks can lead to misleading conclusions about the performance of buyout funds.

The BRP: A Superior Benchmark for Buyout Funds

The Buyout Replication Portfolio (BRP), developed by LupoToro’s research team, offers a more accurate alternative. Designed to systematically replicate the risk and return profile of buyout portfolios, the BRP adjusts for sector allocation and leverage while incorporating a security selection model based on the tendencies of buyout managers.

When used as the benchmark in PME calculations, the BRP offers a clearer picture of private fund performance. For instance:

• Sector Adjustments: The BRP accounts for the underweighting of financial services and REITs, and the overweighting of IT, which are hallmarks of buyout strategies.

• Leverage Adjustments: By factoring in the higher financial leverage typically employed by buyout managers, the BRP captures the amplified risk and return dynamics that small-cap indexes alone cannot reflect.

Insights from PME Analysis

The impact of using an appropriate benchmark like the BRP is evident when evaluating buyout fund vintages from 2014 to 2017. LupoToro’s analysis revealed that the PMEs for these vintages were all within 0.2 of the neutral value of 1.0 when benchmarked against the BRP, indicating that the replication portfolio effectively mirrored buyout performance.

In contrast, using the Russell 2000 as the benchmark—adjusted only for size but not for sector or leverage—would have underestimated buyout fund performance due to the general underperformance of small-cap stocks over the past decade.

Key findings from this analysis include:

• Outperformance vs. Russell 2000: Buyout funds outperformed the sector- and leverage-adjusted Russell 2000 in all four vintages, showcasing positive alpha that cannot be attributed solely to sector bets or financial engineering.

• Misleading Results with S&P 500: The S&P 500, despite its differences from buyout portfolios, appeared comparable in some cases only because large-cap outperformance offset the contributions from sector adjustments and leverage in the Russell 2000.

The Importance of Adjusted Benchmarks

Without proper adjustments for sector exposure and leverage, traditional benchmarks fail to capture the true drivers of buyout fund performance. LupoToro’s research underscores the need for investors to move beyond standard indexes when evaluating private equity returns.

At the very least, investors should consider using a sector- and leverage-adjusted small-cap benchmark. However, the BRP’s incorporation of security selection dynamics makes it the most comprehensive option for benchmarking buyout funds.

Benchmarking private market funds is as much an art as it is a science, requiring thoughtful adjustments to reflect the unique characteristics of buyout investing. The BRP stands out as a superior tool, systematically capturing the risk and return dynamics that define buyout portfolios.

By leveraging the BRP in PME calculations, investors gain a more accurate understanding of private fund performance and can confidently distinguish between genuine alpha generation and market-driven returns. LupoToro Group’s insights reaffirm that precise benchmarking is essential to evaluate the true value of private equity investments.

Conclusion: Understanding the Role of Buyout Strategies in Modern Portfolios

A Framework for Improved Benchmarking

Private equity (PE) buyout strategies have long been heralded as a key component of sophisticated investment portfolios. However, the LupoToro Group Analysis Team has demonstrated that much of the value attributed to buyout strategies—such as leverage, sector allocation, and security selection—can be systematically modeled and replicated in public equity markets. This insight provides limited partners (LPs) with a valuable framework to better benchmark buyout funds and critically evaluate their relative performance drivers.

Buyout managers must demonstrate outperformance through operational alpha or truly differentiated security selection—skills that cannot be easily replicated—to justify their high fees and the added complexity that closed-end funds introduce into portfolios. As institutional and retail investors increasingly flood into private markets, it is essential to assess how differentiated these exposures are and to evaluate their role within broader portfolio strategies.
Replication in Public Markets

LupoToro’s research has shown that over the past decade, a significant portion of the buyout value proposition can be replicated in public equity markets. By employing systematic modeling techniques, the key drivers of buyout performance—including leverage, sector allocation, and security selection—can be mimicked with a portfolio of public equities.

Key takeaways include:

• Leverage: A properly calibrated use of financial leverage amplifies returns in a manner consistent with buyout-backed companies.

• Sector Selection: Adjusting for sector preferences, such as the underweighting of financials and the overweighting of IT, aligns public portfolios with the structural characteristics of buyout funds.

• Security Selection: Favoring undervalued public companies with stable cash flows and strong margins mirrors the security selection tendencies of buyout managers.

Despite these parallels, the analysis also highlighted a critical finding: private equity is neither a diversifying asset class nor inherently less risky than traditional public equities.

Unique Benefits of Private Equity

While much of the buyout value proposition can be replicated, PE buyout strategies retain certain advantages that cannot be duplicated in public markets. These include:

1. Access to a Broader Investment Universe

The universe of private companies is significantly larger than the pool of public companies. Thousands of private firms exist as potential targets for buyout managers, compared to the relatively limited number of realistic public opportunities. This broader investment set provides skilled managers with a greater chance to uncover inefficiencies and identify undervalued assets in a less efficient market.

2. Control Over Portfolio Companies

A defining feature of private equity buyouts is the ability to acquire controlling interests in portfolio companies. This control enables buyout managers to implement operational improvements, such as margin expansion, that are often unattainable in public markets. While the potential for operational alpha has not been consistently realised across the industry, it remains a distinguishing advantage for top-performing managers and deals.

The Importance of Manager Selection

For investors, the ability to capitalize on these unique benefits depends heavily on manager selection. Not all buyout managers are equally skilled at leveraging the broader investment universe or executing operational improvements. Access to the most capable managers—those with a proven track record of delivering top-quartile returns—requires scale, expertise, and careful due diligence.

Rethinking Private Equity’s Role

The insights provided by LupoToro’s team challenge the notion that private equity is a “magic bullet” for portfolio diversification or risk mitigation. However, this does not diminish its importance as a valuable asset class for certain investors. Instead, it highlights the need for a more nuanced understanding of PE’s role:

• For Diversification: Investors should recognize that PE strategies often mirror public equity characteristics, reducing their effectiveness as a diversifying asset class.

• For Value Creation: The unique ability to access a larger investment universe and exert control over portfolio companies remains a compelling advantage for investors with the resources to select top-tier managers.

Private equity buyout strategies offer both replicable and unique value drivers. While much of the performance attributed to buyouts can be systematically replicated in public markets, the broader private investment universe and the ability to control portfolio companies provide advantages that are exclusive to PE.

For investors, this underscores the importance of aligning buyout strategies with their broader portfolio objectives. By leveraging the insights from LupoToro Group’s research, investors can more effectively benchmark buyout funds, critically evaluate their relative performance, and position private equity as a purposeful, rather than default, component of their investment strategies.

* * *

Appendix: Take-Private Model Details

Inclusion Criteria, Label Definition, and Input Data

The LupoToro Group Analysis Team developed a robust take-private prediction model by carefully defining inclusion criteria, labeling methodologies, and input data. The model focused on public companies traded on the NYSE or Nasdaq Stock Market, with additional filters to ensure the dataset aligned with the types of assets typically targeted by buyout managers.

Key inclusion criteria:

• Companies must have traded for at least eight consecutive quarters before the training date.

• Headquarters located in the United States.

• Exclusions included commercial banks, holding companies, and investment trusts, which are not typical buyout targets.

• Only companies actively trading at quarter-end with a stock price above $1.00 were eligible for predictions and inclusion in the Buyout Replication Portfolio (BRP).

Additionally, companies with announced take-private acquisitions were excluded from portfolio eligibility after their announcement date. This ensured predictions were based solely on non-publicly known data. On average, 2,318 eligible companies were included in the prediction set each quarter from 2014 to 2023.

Labeling methodology:

• Binary labels were defined as “1” if a company was taken private within 18 months of the input data date and “0” otherwise.

• The announcement date determined the time elapsed from the input data date to the deal.

• Failed and canceled deals were considered positive examples, reflecting companies that met buyout managers’ acquisition criteria.

Input data:

The model leveraged quarterly financial statements and stock price data to train predictions. Key features were lagged by one quarter to ensure availability at the prediction date.

Data points included:

• Income Statement: Revenue, cost of goods sold, operating profit, R&D expenses, EBITDA, net interest expense, tax rate, normalized net income.

• Balance Sheet: Total assets, net debt, cash and short-term equivalents, property, plant, and equipment (net), inventories, intangible assets, enterprise value, and book value.

• Cash Flow Statement: Net operating cash flow, net financing cash flow, net investing cash flow, free cash flow, and cash dividends paid.

• Stock Price Data: Adjusted closing price, diluted shares outstanding, market capitalization, time since IPO, and two-year rolling beta to the S&P 500 Index.

• Additional Variables: Industry classification, number of PIPE (private investment in public equity) deals, and year-over-year return of the S&P 500 Index to account for general market conditions.

Model Training and Validation

Training process:

The model was trained quarterly, starting on June 30, 2012, with validation periods moved forward by 18 months to reflect the forward-looking nature of the binary labels. For instance, data through June 30, 2012, was used to train the first model, which was validated on data from December 31, 2013. This step-ahead validation process ensured no overlap between training and validation datasets.

• Training duration: Models were trained through December 31, 2023, to keep the BRP up-to-date.

• Validation set: 32 quarterly models were validated, with the final validation period ending June 30, 2022, allowing for the necessary 18-month lag.

• Model implementation:

  • Framework: The models were implemented using the PyTorch library and optimized with the Adam algorithm.

  • Architecture: A long short-term memory (LSTM) network was employed to handle sequential data inputs effectively.

  • Optimisation: Early stopping techniques and other hyperparameter tuning methods (e.g., number of LSTM layers, hidden cell size) were applied during the first half of the validation period. The second half used fixed hyperparameters to ensure consistency.

Performance Metrics

The take-private prediction models achieved strong performance metrics:

• Average AUC: 0.73 across 32 validation and test periods, indicating the model effectively distinguished between likely and unlikely take-private candidates.

• Top 10% Recall: 29.5%, highlighting the model’s ability to identify a significant portion of actual take-private targets within the highest-probability candidates.

These results demonstrate the model’s capability to capture patterns and characteristics common to companies targeted by buyout managers. The framework not only underpins the construction of the BRP but also provides actionable insights into the dynamics of private equity acquisition strategies.

The LupoToro Group’s take-private prediction model represents a sophisticated approach to identifying buyout candidates. By combining precise inclusion criteria, comprehensive financial data, and cutting-edge machine learning techniques, the model delivers meaningful insights into the decision-making processes of private equity managers, empowering investors with predictive tools for navigating the public and private equity landscapes.

This article is intended as opinion only, not for financial, investment or legal business advice.