Merger Arbitrage (Announced Deal Spreads)

Thesis (edge)

When one company agrees to buy another, it typically offers a premium. If the target was trading at 40 and the acquirer offers 50, the target's shares will jump on the announcement, but usually not all the way to 50. They might settle at 48.

That remaining gap of 2 exists for a reason. The deal is not done. It needs shareholder approval, regulatory clearance, financing, and time. Some deals fall apart. The market discounts the offer price to compensate whoever holds the shares for bearing the risk that this one is among them.

Merger arbitrage is the business of collecting that gap. Buy the target at 48, wait for the deal to close, receive 50. If it takes four months, that 4 percent gain annualizes to something like 12 percent, which is attractive. Do it across many deals, diversify the risk, and you have a strategy whose returns depend on deal completion rather than on the direction of the stock market.

The payoff structure is the thing to understand before anything else. It is a long string of small wins interrupted by occasional large losses. It is, in economic substance, an insurance business. You are selling protection against deal failure and collecting a premium for it.

Where it works (regimes)

The strategy earns steadily in normal markets when deal flow is healthy, financing is available and regulators are permissive. In those environments most deals close, spreads are collected, and the return stream looks smooth and impressive.

It fails when several things go wrong at once. In a market crash, financing dries up, acquirers get cold feet, and deals that looked safe start to wobble. Target prices fall, spreads widen sharply, and the strategy takes losses across many positions simultaneously. This clustering is the crucial risk and the reason the strategy is not the market-neutral haven it is often sold as. In 2008 merger arbitrage funds lost money at the same time as everything else, precisely when their supposed independence from the market would have been most valuable.

It also suffers during periods of aggressive antitrust enforcement. When regulators shift from permissive to hostile, deals that priced at tight spreads because they seemed certain suddenly do not close, and the losses land on positions that had been sized as if they were safe.

Signals

• Gross spread: the offer price minus the current price. This is the raw opportunity.
• Annualized spread: the gross spread scaled by the expected time to close. A 2 percent spread on a deal closing in one month is a much better trade than a 4 percent spread on a deal closing in a year, and only the annualized number reveals that.
• Implied break probability: this is the most useful lens. If the target trades at 48, the offer is 50, and the shares would fall to 40 on a break, then the market is implying a break probability of roughly 20 percent. Your job is to decide whether the true probability is higher or lower than that. The strategy is not about collecting spreads. It is about identifying deals where the market's implied break probability is too high.
• Risk factors: antitrust overlap between the two businesses, national security review if the acquirer is foreign, political attention, financing that is not yet committed, and a hostile rather than agreed structure. Each of these raises the true break probability, and the spread will already reflect some of them.

Portfolio construction

For an all-cash deal, the position is simply long the target. For a stock-for-stock deal, where the acquirer is paying in its own shares, you must also short the acquirer in the exchange ratio, otherwise you are exposed to the acquirer's price and not to the deal itself. This is a common and expensive mistake for newcomers.

Diversification is not optional. A concentrated merger arb book is a coin flip with extra steps. Run twenty to thirty live deals, and make sure they are not all facing the same regulator, the same sector or the same financing market, because that would leave you with one risk rather than thirty.

Size positions by the loss on a break, not by the size of the spread. The instinct is to put more money into the widest spread, but the widest spread is usually the deal most likely to fail. Correct sizing means calculating what you lose if the deal collapses and making sure no single break can take out a year of gains.

Some managers buy put options on the target to cap the downside on the riskiest deals. This turns a large occasional loss into a smaller, constant cost, and whether it improves the strategy depends entirely on how the options are priced.

Risk model

The deal break is the entire risk model. When a deal fails, the target does not drift gently lower. It gaps down, often to below its pre-announcement price, because the market now knows something is wrong with the company that made a buyer walk away. A 4 percent spread turned into a 25 percent loss in a single session is a completely ordinary outcome.

The three main causes of a break, in rough order of importance:

• Regulatory. An antitrust authority sues to block the deal, or a national security review objects. These decisions are political as much as economic, and they are extremely hard to model. Regulatory attitudes shift with administrations, and a merger that would have sailed through five years ago can be blocked today.
• Financing. The acquirer cannot raise the money, usually because credit markets have turned. This risk rises sharply in exactly the market conditions where everything else in your portfolio is also falling.
• Buyer's remorse. The acquirer finds a reason to walk, invokes a material adverse change clause, or simply renegotiates the price downward. Even a renegotiation, without a full break, produces a loss.

Because breaks cluster in bad markets, the strategy has a hidden correlation to equities that only appears in a crisis. Model it with expected shortfall or scenario analysis, not with volatility, because the volatility of a merger arb book in calm times is deceptively low.

Costs & implementation

Trading costs are modest, but the strategy demands genuine research. Reading merger agreements, assessing antitrust exposure and judging regulatory temperament are not things a price-based model can do. This is the most fundamentally driven strategy on this list, and a purely mechanical version that just buys every announced deal will collect the average spread and take the average losses, which historically has been a mediocre outcome after costs.

For stock deals, borrow cost on the acquirer matters and can be the difference between a profitable and an unprofitable trade. Hard-to-borrow situations should simply be avoided.

Capacity is limited by the size of the deals themselves. You cannot deploy unlimited capital into a finite number of announced mergers, and crowding into the same deals compresses spreads for everyone.

Failure modes

• Sizing by spread rather than by downside. The widest spreads are the deals the market is most worried about, and usually the market is right to worry.
• Ignoring the stock leg in stock-for-stock deals, which leaves you long the acquirer by accident.
• Concentrating in deals facing the same regulator, so that one policy decision breaks several positions at once.
• Backtesting only on completed deals, which is a spectacular form of survivorship bias and will produce a strategy that appears to print money.
• Believing the strategy is market neutral. It is neutral until it is not, and the moment it stops being neutral is the moment you needed it to be.
• Underestimating how much regulatory regimes can change. The historical base rate of deal completion is not a constant of nature.

Our Notes & Suggestions

Think of this as underwriting rather than trading. You are writing insurance policies on deal completion, and like any insurer your success depends on pricing the risk correctly, diversifying properly, and holding enough capital to survive the bad year.

The systematic edge, to the extent one exists, comes from being more accurate than the market about break probabilities in the middle of the distribution, and from avoiding the deals where the spread is wide because the deal is genuinely doomed. The easy deals with tight spreads pay very little. The wide spreads are wide for reasons. The money is in judgment about the ones in between, which is why this strategy has stubbornly resisted full automation.

If you build a purely quantitative version, be honest that you are collecting an insurance premium with a fat left tail, size it accordingly, and never let a single deal be capable of ruining the year. The managers who blew up in this strategy were almost always the ones who found a deal they were certain about.