Merger Arbitrage: The Risk-Return Profile of Event-Driven Strategies

When RJR Nabisco Broke the Arbs

In October 1988, the management group of RJR Nabisco launched a $17.6 billion leveraged buyout bid at $75 per share. Risk arbitrageurs immediately piled into the stock, buying shares in anticipation of the deal closing. Then Kohlberg Kravis Roberts countered at $90. A bidding war erupted. Over six weeks, the price spiraled through multiple rounds before KKR prevailed at $109 per share. Arbitrageurs who entered early made extraordinary profits. But those who bought near the top of each interim bid, only to see a competing offer restructure the terms, experienced the whiplash that defines merger arbitrage: the strategy looks mechanical until it does not.

Thirteen years later, Mark Mitchell and Todd Pulvino published the most comprehensive empirical study of this phenomenon. Their 2001 paper in the Journal of Finance assembled a dataset of 4,750 mergers and acquisitions between 1963 and 1998, constructing calendar-time portfolios that tracked what a systematic merger arbitrageur would have earned. The findings upended the conventional wisdom about risk arbitrage as a steady, market-neutral income stream.

The Put-Selling Analogy

Mitchell and Pulvino's central discovery was that merger arbitrage returns are not linearly related to market returns. In months when the S&P 500 was flat or positive, risk arbitrage portfolios generated consistent excess returns with near-zero market correlation. But when the market declined sharply, deal spreads widened, deal break rates spiked, and arbitrage returns turned sharply negative — becoming highly correlated with the market at the worst possible moment.

This conditional correlation structure mirrors the payoff of selling uncovered put options on a stock index. The option seller collects steady premium income when markets are calm but faces potentially unlimited losses during a crash. Mitchell and Pulvino demonstrated that a simple portfolio combining short put options on the market with Treasury bills replicated the statistical properties of merger arbitrage returns remarkably well.

The piecewise linear regression told the story cleanly. When market returns exceeded roughly negative 4 percent monthly, the beta of risk arbitrage relative to the market was statistically indistinguishable from zero. Below that threshold, the beta jumped to approximately 0.5. The event-driven strategy that appeared uncorrelated with equities in normal conditions became half as volatile as the market itself during downturns — precisely when diversification benefits matter most.

What Drives the Asymmetry

The non-linear payoff arises from the interaction between completion risk and market conditions. During economic expansions, most announced deals close successfully. Regulatory approval flows smoothly, financing remains available, and acquirer stock prices hold up. The arbitrageur collects the deal spread on the vast majority of positions, generating a stream of small positive returns.

Market downturns disrupt each of these channels simultaneously. Acquirers face tighter credit conditions and falling stock prices that erode the economic logic of their bids. Regulatory agencies become more cautious. Target shareholders grow restive. The probability of deal failure rises across the entire portfolio at once, creating correlated losses that no amount of deal-level diversification can eliminate.

Mitchell and Pulvino documented this mechanism with granular data. In their sample, roughly 23 percent of cash tender offers and 18 percent of stock mergers failed to complete. But failure rates were not uniformly distributed. They clustered during periods of market stress, when multiple deals collapsed simultaneously — the same correlation regime that undermines portfolio-level hedging.

Quantifying the Premium

Across the full 1963-1998 sample, Mitchell and Pulvino estimated that risk arbitrage generated annualized excess returns of approximately 4 percent above Treasury bills after accounting for transaction costs. This premium compensated arbitrageurs for bearing the non-linear downside risk that the strategy entails.

Baker and Savasoglu (2002) extended this analysis by connecting merger arbitrage returns to the limits of arbitrage framework. They found that deal spreads were wider — and expected returns higher — for transactions involving targets with limited institutional ownership and lower liquidity. When fewer well-capitalized arbitrageurs could participate in a deal, the spread available to those who could was larger, consistent with the idea that arbitrage capital is scarce and constrained.

More recently, Jetley and Ji (2010) documented a secular compression in deal spreads from the mid-1990s onward. As dedicated merger arbitrage funds proliferated and capital flowing into the strategy expanded, the average spread narrowed from around 2.3 percent per deal to roughly 1.0 percent. The liquidity premium embedded in deal spreads shrank as the strategy became more crowded, though it did not vanish entirely.

Conditional Risk in Practice

The practical implications extend beyond hedge fund due diligence. Any portfolio that includes merger arbitrage exposure — whether through dedicated funds, multi-strategy allocations, or merger arb ETFs — carries embedded short-volatility risk. The strategy will dampen portfolio volatility during benign markets, potentially inflating risk-adjusted metrics like the Sharpe ratio. But in a severe downturn, merger arbitrage positions can amplify drawdowns rather than cushion them.

Mitchell and Pulvino's framework suggests that the appropriate benchmark for merger arbitrage is not a static risk-free rate or equity index, but a dynamically hedged portfolio that accounts for the conditional market exposure. Evaluating merger arb performance without adjusting for its option-like payoff structure overstates true alpha and understates true risk.

For allocators weighing event-driven exposure, the core question is not whether merger arbitrage generates positive expected returns — decades of evidence confirm that it does. The question is whether the portfolio can absorb the concentrated losses that arrive during the market environments where everything else is also falling.