Merger arbitrage: how to invest when one company buys another

A company announces it's buying another one for $50 a share, in cash. Right before the announcement, the target was trading at $36. A few weeks later, it trades at $48.50.

If someone has already agreed to pay $50, why can you buy that share today for $48.50?

That $1.50 gap is the whole idea behind a strategy called merger arbitrage: you buy shares of a company that already has a buyer, and you pocket the difference between what you pay today and what you get paid when the deal closes.

If everything goes to plan, in a few months someone hands you $50 for something that cost you $48.50. You make $1.50 a share, and that depends mostly on the deal going through, not on the market going up (although, as we'll see, a crisis can put the deal at risk and make you lose more if it breaks).

Put like that, it sounds like free money. But that $1.50 isn't there by accident: the deal can still get delayed, renegotiated, or fall apart entirely. So let's look at where that $1.50 comes from and when it's worth going after. I've left the more technical stuff for the end, for anyone who wants extra credit.

What merger arbitrage is (the simple version)

Imagine your neighbor is selling their apartment. They already have a buyer and a signed contract for $300,000, but the buyer won't pay for six months, until the mortgage comes through and the paperwork is done.

Your neighbor is in a hurry and offers you a deal: "Give me $291,000 today and the contract is yours. In six months, you collect the $300,000."

Good deal? Well, it depends on two things:

• How sure is it that the buyer pays? If the mortgage gets turned down, you're left holding an apartment that might be worth $250,000.
• How long will it take? If it's two years instead of six months, that $9,000 doesn't look like much anymore.

That, in a nutshell, is merger arbitrage. Swap the apartment for shares of a listed company, the buyer for another company, and the paperwork for regulators, votes, and banks. The logic is exactly the same.

You'll also see it called risk arbitrage. And no, it has nothing to do with legal arbitration (when two companies settle a dispute in front of an arbitrator). This is about investing.

The gap between what the share costs today and what the buyer will pay is called the spread. In the example, $1.50. If the deal closes, you keep it. If it breaks, the stock drops and you lose a lot more than you stood to make.

So why is anyone handing me that money?

Nobody's handing it to you. The market pays you that $1.50 for doing three things a lot of people would rather not do:

• Waiting. Plenty of shareholders would rather sell now, take 97% of the price, and forget about it, than spend months waiting for the last 3%.
• Living with uncertainty. Many funds can't (or won't) hold a stock that can drop 20% in a day because a regulator doesn't like the deal.
• Reading the fine print. Analyzing a deal properly means reading contracts, understanding regulators, and following the news for months. Not everyone has the time or the patience, and honestly, I get it.

The spread is what you get paid for all of that: for the wait, and for the chance that something goes wrong before the deal closes.

Pennies in front of a steamroller

The problem is that what you can make and what you can lose look nothing alike. If the deal closes, you make $1.50 a share. If it breaks, the stock goes back to being worth what it's worth on its own, with nobody trying to buy it. Say about $38: you lose $10.50.

In other words, you're risking about $7 for every $1 you can make.

That's why people say merger arbitrage is like picking up pennies in front of a steamroller. Most of the time you pick up your pennies and nothing happens. But the day the steamroller gets you, a single deal break wipes out the gains from several deals that went fine.

Merger arbitrage: +$1.50 if it closes, −$10.50 if it breaksT-bills: +$0.97 per deal

That's 20 deals with the numbers from the example, one after another, six months each: ten years in total. Each time you buy one share, so you never have more than $48.50 in. If 18 go well and 2 break (a 90% hit rate, which sounds great), after ten years you've made $6. If you'd left that same $48.50 in T-bills for those ten years, you'd have $19.40. In other words, you'd have done a lot better doing nothing.

And reinvesting what you make doesn't help: every time a deal breaks, you lose more than 20% of everything you have at that point. You'd end up about $3 ahead, versus about $23.60 in T-bills.

That said, it's not all bad news. The strategy has some really good things going for it, and we'll go through them properly further down, in What's good about merger arbitrage:

And to be fair, the steamroller example is the worst case: all your money in a single deal each time, bought at a price that was already tight. Nobody does it that way. The normal thing is to hold lots of deals at once, with a little in each, and only buy when the price makes it worth it.

Here's the same math done that way: $10,000 split across 20 deals at once, $500 in each, all bought at $47. When one closes, you open another. Over three years that's 120 deals, and 12 break, the same 90% hit rate as before:

Portfolio: +$32 if it closes, −$96 if it breaksT-bills: the same $10,000

Now when a deal breaks, you lose about $96, 1% of the portfolio, and the deals that keep closing make up for it quickly. In the end you make about $2,300, almost double the $1,200 from T-bills.

One thing, though: what flips the result is mostly the price, not having lots of deals. Spreading out turns each break into a scratch instead of a punch, but it doesn't change what you make on average. That same portfolio, bought at $48.50, would make about $370 in three years, well below T-bills.

Napkin math

Before we get into the details, here's a summary of the math we'll do further down:

	In the example
You make if it works out	$1.50 a share (3.1%)
You lose if it doesn't	About $10.50 a share (21.6%)
Time to close	About 6 months
What you'd make risk-free in that time	2% in T-bills
Minimum closing probability to beat T-bills (we work it out below)	95.6%

Put another way: of that 3.1%, about 2 points you'd also get from T-bills without risking anything. The real reward for taking the risk is a lot smaller than it looks at first glance.

Where the $1.50 spread comes from3.1% in 6 months

What T-bills would pay you

$0.97 · 2.0%

Reward for the risk

$0.53 · 1.1%

Since the example is in dollars, I compare it with US Treasury bills of a similar maturity, held to maturity. I use 4% a year as an example rate; for a real deal, check what they pay that day. All the math is in dollars, before costs and taxes. If you invest from another currency, the exchange rate matters too (more on that at the end).

The only thing you need to figure out

If you take away one thing from this post, make it this. For every deal, all the work boils down to answering three questions:

1. Will it close, and on what terms?
2. When?
3. What happens to me if it doesn't?

Everything else - contracts, regulators, votes, formulas - is just tools for answering them.

Merger arbitrage already came up briefly in the post on return engines, as one of those ways of making money that don't depend on the market going up. If you want to see where it fits next to other strategies:

From announcement to payday

From the day a deal is announced to the day it closes, it usually takes several months. Along the way there are several stages (some run in parallel), and at each one something can go wrong. And when do you get paid? Since several run at the same time, the slowest one sets the pace: figure out which one will take longest and add the paperwork that comes after it.

1. The announcement

   The two companies sign the agreement and make it public. The stock jumps toward the offer price but stays a bit below it: that's where the spread is born.

2. The fine print

   The merger agreement gets filed. The press release tells the nice story; the contract says what has to happen for the deal to close, by when, and what happens if someone walks away.

3. The approvals

   Regulators check that the deal doesn't hurt competition and sometimes also look at foreign investment or the industry itself. This can take weeks or years.

4. The vote

   The target's shareholders vote on whether to accept. In a tender offer, instead of voting, each shareholder decides whether to sell.

5. The money

   The buyer has to have the cash on closing day. If it depends on banks lending it, there's risk here.

6. The deadline

   If the deal still hasn't closed by an agreed date, it can be called off. That date is usually called the outside date.

7. Closing: you get the agreed price

Most announced deals end up closing. But that average mixes near-certain deals with very shaky ones: on its own, it tells you nothing about the odds of the one you're looking at.

What about before anything is signed?

Before the definitive agreement there are usually rumors, strategic reviews (the classic "we're exploring alternatives"), and non-binding preliminary offers. There can be more upside there, sure, but it's a different bet: neither the price nor the deal itself is certain. If you're just starting out, I'd wait for a signed agreement.

Are you paid in cash or in stock?

In cash. The buyer pays a fixed amount per share, like in the example. It's the simplest kind: you buy, you wait, you get paid. If you're starting out, start here.

In stock. The buyer pays with its own shares ("I'll give you half a share of mine for each share of yours"), so what you'll receive goes up and down with the market. To get rid of that risk, professionals sell borrowed shares of the buyer at the same time. That's called short selling. It works, but it adds costs and complications. I explain it at the end.

A mix. For example, $20 in cash plus 0.3 shares of the buyer for each of your shares. The cash part is fixed; the rest moves with the stock price. There are even stranger variants, but my advice is to leave them until you have a few simple deals under your belt.

The math (it's just subtraction, promise)

You don't need to be a mathematician: it's all subtraction and division. Let's start with the easy part, what you'd make if the deal closes.

The $50 in the headline isn't always your $50

The headline says $50, but what you end up collecting isn't always the offer price, and what you pay isn't just the share price either. It's worth tightening up both numbers in the subtraction:

• Dividends. Check what the agreement says. Some deals let you collect them on top of the offer price; others deduct them or ban them until closing. If you collect $0.50 and it's later taken off the $50, you haven't made anything extra, so don't count it twice.
• What's already been paid. If the $50 announced included a $3 advance paid before you bought, you only have $47 left to collect. Do the math with what you still have to collect.
• What you pay. The purchase price plus commissions and, if you invest from another currency, the cost of converting.
• When you get paid. Even once the deal has legally closed, it can take a few more days for the cash to show up.

The catch in that 6.3% a year

Plenty of websites will tell you this deal "yields 6.3% a year". Sounds great, but there's a catch:

• It assumes it closes on time. If it takes a year instead of six months, you still make the same 3.1%, but now over a year: 3.1% a year. That's less than T-bills pay in the example, 4% a year (the 2% in six months from before). And delays are very common.
• It flatters deals that are about to close. Making 1% in two weeks "works out" to more than 25% a year. Lovely, until those two weeks turn into two months and the 25% shrinks to 6%.
• It's not what you'll make in a year. Annualizing is only useful for comparing deals with different timelines. To actually make that 6.3% in a year, you'd have to chain similar deals back to back, with no idle cash and not a single loss.

Look at what happens to the same 3.1% depending on how long the deal takes to close:

Annualized return on a 3.1% gain, by how long the deal takes to close

That's why it pays to take the timeline seriously. If a deal that should have closed in four months takes fourteen, you make the same, but over more than three times as long.

If it breaks, where does the stock land?

For me, this is the most important question in the whole analysis. The easy answer is that the stock goes back to where it was before the announcement, $36. But it's almost never exactly that:

• The market has moved. If similar companies are up 10% since the announcement, TargetCo without a buyer would probably be worth more than $36.
• The company has been frozen. While it waits to be bought, it can't make big moves, and sometimes that takes a toll.
• If it breaks on bad news, it can fall further. For example, if the buyer found something ugly.
• Or it can fall less. Someone being willing to pay $50 says something about what the company is worth, and sometimes another buyer shows up.

So you have to do the classic thing: value the company as if nobody wanted to buy it. In the example, let's say that comes out at about $38.

But careful: valuing the stock at $38 doesn't mean you'll be able to sell it at that price. For this math, what you need is an estimate of what you'd sell for if the deal breaks, and when. If you think the company is worth $38 but can see yourself selling at $32 in the stampede after the break, use $32. If your plan is to wait for it to get back to $38, that's another story, with its own timeline. That estimated sale price after a break is usually called the break price. In the example we'll use $38.

How sure do you need to be?

You now have the three numbers: today's price, the offer price, and the price if it goes wrong. With those you can work out how often the deal needs to go through for what you make to make up for what you lose when it doesn't.

Think of the answer as a bar to clear. If you think the real probability is above it, the numbers work. If it's below, they don't.

To keep things simple, for now we'll assume the closing and the break would happen on the same date. Further down we'll see what happens when they don't. Here's the formula:

Prob=

Price − Break

Offer − Break

Where

Prob = closing probability

The closing probability above which, on average, you don't lose money buying at today's price.

Price = today's price

What the target's stock costs right now.

Offer = offer price

What you get paid per share if the deal closes.

Break = break price

Your estimate of what you'd sell the stock for if the deal falls apart.

With the numbers from the example:

Prob=

$48.50 − $38

$50 − $38

=$10.50$12=87.5%

What does that mean? That at an 87.5% chance of closing, what you expect to collect matches what you pay. Put another way: if you could repeat this exact deal many times, what you make on the ones that close would exactly cover what you lose on the ones that break. On average, you neither make nor lose money (before costs). That "on average" is the key idea for everything that follows.

But something's still missing: time. While you wait, that money could be earning interest somewhere else. If you leave that $48.50 in T-bills for six months at 4% a year, you make 2%. So part of the spread isn't a reward for risk at all: it just pays you for waiting, like it would pay anyone.

To account for that, instead of comparing what you expect to collect with what you pay today, you compare it with what you'd have if you'd left that money in T-bills:

Prob=

Price × (1 + r) − Break

Offer − Break

r = what you'd make risk-free while you wait

For example, what T-bills pay over the time left until closing (2% over six months, not the 4% annual rate).

With the numbers from the example, and T-bills at 2% over those six months:

Prob=

$48.50 × (1 + 2%) − $38

$50 − $38

=$49.47 − $38$12≈95.6%

That's a huge difference. Of the $1.50 spread, almost $1 is what the Treasury would pay you to wait without risking anything. If everything goes well, you only make about 53 cents more than with T-bills.

The bottom line: with these assumptions, you only do better than T-bills, on average, if the closing probability is above 95.6%. All before costs. And even if you clear it by a hair, I'd think long and hard about risking $10.50 to make $1.50.

Move the price you'd sell at if the deal breaks and watch the bar move:

If it breaks, you sell at$38

Before you get this number tattooed, a few warnings:

• It's not what the market "thinks". It comes from your assumptions. Part of the spread is what other people charge for carrying the risk, even if they believe exactly what you believe. So a spread existing doesn't mean the market is wrong.
• It depends hugely on your break price. With $32 instead of $38, the bar goes up from 95.6% to 97.1%.
• It only allows for two endings. It's useful for getting your bearings, but don't take it too literally.

So where do I get my probability from?

Now comes the hard part, and also the part where it's easiest to fool yourself. No formula gives you your probability: in the end it's always an estimate. But you can make it with a bit of sense:

And if the deal is only worth it at 98% but stops being interesting at 95%, you have another problem: can you really tell those two numbers apart? I certainly can't. When a small change in such an uncertain estimate flips the whole decision, the numbers are held together with tape.

So, is it worth it?

You have the bar and you have your estimate. All that's left is to put them side by side.

How much would you pay?

So far we've asked what probability you need at today's price. But when it's time to decide, I find it more useful to flip it around: with my own assumptions, what's the most I'd be willing to pay?

It takes two steps. First you work out what you expect to collect on average. Then you divide it by 1 plus the return you want from the deal until it closes: if you want 4%, divide by 1.04. That's your maximum price. It's a before-costs calculation; if you want to be precise, subtract the commissions for buying and selling.

Max=

Prob × Offer + (1 − Prob) × Break

1 + Required

Where

Max = maximum price

The most it would make sense to pay for the stock today, given your assumptions.

Required = required return

What you want from the deal over the time until it closes (not annualized). At the very least, what T-bills pay.

Say you give the deal a 97% chance of closing. On average you expect to collect 0.97 × $50 + 0.03 × $38 = $49.64. Then:

• If you only want the same as T-bills (2% over six months), the maximum price is $48.67.
• If you want 4% over six months for the risk and the work (an example figure, not a recommendation), the maximum drops to $47.73.

There are two separate filters here. The first: does the deal beat the 2% T-bills would give you? The second: does it reach the 4% total you're asking for the risk and the work? You can pass the first without passing the second.

And if you don't have one number but a range (say, 94% to 97%), do the math with the low end. At 94%, the maximum when you want 4% drops to $47.38. At $48.50, the deal isn't for you: even if it were certain to close, you wouldn't get to that 4%. At $47, on the other hand, it would pass the valuation filter (which doesn't spare you from analyzing the risks).

Thinking this way also protects you from a classic trap: when the price doesn't work, nudging your probability up bit by bit until it does. If you catch yourself raising your estimate so the numbers add up, ask yourself what new information justifies it. If there isn't any, be suspicious: the only thing that's probably gone up is how badly you want to buy.

To do all this math with your own numbers, I built a small free calculator. It gives you the break-even probability, the maximum price, and what happens if you get the probability or the timeline wrong.

Real life has more than two endings

So far we've played with two endings: it closes at $50, or it breaks and drops to $38. In real life there are more. It can close late, the offer can get cut, another buyer can show up, or it can break on bad news. And each ending arrives at a different time:

Put it on a chart and you can see where that −$0.34 comes from: the most likely ending adds little, and the two breaks, unlikely as they are, take away a lot more.

What each ending adds to or takes from the average, against T-bills

Closes on time

83% · +$0.53 if it happens

+$0.44

Closes, but late

7% · −$0.44 if it happens

−$0.03

Price gets cut

2% · −$2.95 if it happens

−$0.06

Higher offer

2% · +$5.21 if it happens

+$0.10

Breaks

4% · −$11.31 if it happens

−$0.45

Breaks on bad news

2% · −$17.31 if it happens

−$0.35

Average

−$0.34

A deal can close and still lose you money (if the price gets cut) or make you less than T-bills (if it takes too long). And you don't need to get the percentages exactly right: the useful part is the exercise, which forces you to think through everything that can happen, not just the happy ending.

That's why I wouldn't go looking for the biggest spread, but for deals where what you can make is worth the risk you're taking.

What's good about merger arbitrage

If you've made it this far, you might be thinking this is a terrible idea. Fair enough: I've spent quite a while talking about steamrollers and losing money. But I picked the deal in the example on purpose, with a very thin spread, so the risks would be easy to see. When the price is right, things look quite different.

And the strategy has a lot going for it. These are the things I like most:

The historical evidence is on its side. In the stock market, the usual rule is that if you want to make more, you have to accept more risk: stocks pay more than T-bills because sometimes they crash. Two classic studies checked whether merger arbitrage followed that rule. And it turns out it made more than its risk called for. In other words, on top of paying you for the risk, there was something extra.

It makes sense if you go back to why the spread exists: lots of people pay to avoid waiting and reading contracts, and somebody pockets that money. Here are the numbers, if you're curious:

Buffett did it. In his 1988 letter to shareholders, Warren Buffett explains that Berkshire did arbitrage as an alternative to holding cash, because it often had "more money than ideas". And he gives some striking numbers: at Graham-Newman, the firm run by his mentor Benjamin Graham, arbitrage made an average of 20% a year between 1926 and 1956, without leverage. And Buffett reckoned his own had beaten that 20% by a wide margin between 1956 and 1988, though he made clear he hadn't done the exact math. Over those 63 years, the stock market returned a bit under 10% a year. It's a striking historical reference, not a return you should expect. And note that Buffett uses "arbitrage" in a broader sense: it also includes liquidations, reorganizations, and other corporate events, so those numbers aren't a pure merger arbitrage track record.

And it's done its own thing. In Mitchell and Pulvino's sample, when the market was flat or rising, merger arbitrage returns had little to do with the market's. If your portfolio is already full of stocks, a return that most of the time doesn't depend on what the market does is worth a lot.

It's not completely independent, though. Using monthly data from the last twenty years, The Merger Fund's correlation with the stock market is around 0.5 (my own calculation using monthly returns for MERFX and the S&P 500 with dividends, from July 2006 to June 2026, with data from Yahoo Finance): a long way from the near-1 of a stock portfolio, but not zero. And in big crashes it falls with the market, though much less (more on that in When a crisis hits).

So how much has it actually made? The right comparison here is T-bills, which is the real alternative. Over the last twenty years, from July 2006 to June 2026, The Merger Fund beat T-bills by a bit under two points a year, with far less volatility than the stock market and nowhere near its return:

July 2006 to June 2026	The Merger Fund	T-bills	S&P 500
Annual return	3.3%	1.6%	11.4%
Annual volatility	3.1%	—	15.3%
Worst drawdown	−12.2%	—	−55.3%

These are returns with dividends reinvested and, for the fund, after its fees, over years when rates spent a long time near zero. My own calculation with data from Yahoo Finance (MERFX and the S&P 500 with dividends) and from the Federal Reserve for three-month T-bills (FRED, TB3MS).

Almost all of them work out. As a historical reference, a guide from Accelerate, a Canadian fund manager that specializes in this, puts the completion rate for announced US deals at 94%. Within those 94 out of 100, in about 6 the buyer ended up paying more than announced, and in 1 less:

How 100 announced US deals end

Closes as agreed87Closes at a higher price6Closes at a cut price1Breaks6

It's not a probability you can apply directly to any deal or to your portfolio, and the guide doesn't say which years the data covers. But it gives you a sense of scale: out of every 50 deals, about 3 break and about as many get a better price. Westchester, the firm behind The Merger Fund, says more than 98% of the 5,000 deals it has invested in since 1989 have closed, although those are the ones they picked after analyzing them.

You know what you make if it works out, and roughly when. With a normal stock, your profit depends on the market someday seeing what you see, and nobody tells you when that'll happen. Here there's a price agreed in writing, a set of conditions, and some dates. Your job comes down to checking whether someone will pay what they've already promised, which is a lot more concrete than guessing what a company is worth. And almost everything you need for that is in public documents.

At a good price, the numbers work out really well. Go back to the example, but imagine you can buy at $47 instead of $48.50:

• The bar for beating T-bills drops from 95.6% to 82.8%.
• If your estimate is 95%, you expect to collect $49.40 on average: 5.1% in six months, more than 10% annualized with these timelines and assumptions. That's about $1.46 a share more than T-bills would give you.
• And you have room to be wrong. Even if the real probability were 90%, you'd still expect $0.86 more than with T-bills.

So the same deal can be bad at $48.50 and pretty good at $47. All of this assumes the probability, the timeline, and what you'd lose if it breaks haven't changed. If the stock has dropped because the deal has gotten more complicated, it's not enough to redo the math with the new price: you have to revisit the other numbers too (we'll see this in the month-by-month example). And those prices do come up: when the market is scared of a regulator, when a crisis hits and lots of people are forced to sell, or in small companies the big funds don't even look at, the spread sometimes widens well beyond what the real risk justifies.

Your money comes back quickly. A deal usually lasts a few months. When it closes, you get paid and move on to the next one, so you can go around several times in a year, each deal analyzed from scratch and at that moment's price.

When rates are high, spreads tend to be wider. Since part of the spread is what T-bills pay, when interest rates go up, spreads tend to widen with them. An analysis by New York Life Investments using merger arbitrage indexes from 1994 to 2020 found that relationship, although it was clearer before 2008 than after. Bonds work the other way: when rates go up, their price falls. That said, what goes up is the total return. What you make above T-bills, the reward for the risk, doesn't have to change. And what's good for someone getting in now isn't necessarily good for someone already in: if spreads widen, their position loses value, even though what they'll collect at closing stays the same. There's a flip side, too: with high rates, buyers that depend on debt, like private equity funds, find it more expensive to borrow, and that can put their deals at risk.

You have the edge in a few things. Many funds have rules that force them to sell or keep them out of small companies; you don't (more in Can you beat the pros?).

Sometimes there's a bonus. Every now and then another buyer shows up or the offer gets raised. In the six-endings example, that case added $5.21 a share versus T-bills.

You learn a ton. Following a handful of deals from start to finish teaches you how companies get bought and sold, how regulators think, and how to read a contract. That's useful for investing in general, not just for this.

All of this comes down to the same thing: getting in at a price that pays for the risk. What comes next is for exactly that.

How deals break

Every deal has its own mix of risks, and the spread is the price of all of them together. These are the most common ones, and the clues for seeing them coming.

The regulator says no

This is the headline risk in big deals. Competition authorities ask a pretty simple question: if we put these two companies together, will customers end up with fewer options or paying more? In the US that's the FTC or the Department of Justice; in the European Union, the European Commission or each country's national authority, depending on the size of the deal; in the UK, the CMA. Some clues:

• If the two companies compete head to head and they're big, be careful. That's the classic case for a block.
• Not competing doesn't mean no risk. Regulators also worry about the buyer being able to cut rivals off from something they need. That was the big argument against Microsoft buying Activision.
• An in-depth review doesn't mean the deal will fail, but it can add several months of waiting. In the US, when the regulator asks for a lot more information to do that review, it's called a second request.
• The political climate matters. Some eras and administrations are a lot tougher than others.

Not all approvals are about competition. A deal may also need sign-off on foreign investment or national security (in the US, from CFIUS), or from an industry regulator, as in banking or telecoms. In some EU deals, foreign subsidies control also comes in, which is a separate review.

A buyer out of nowhere

When the buyer is Microsoft, you know pretty well who's on the other side. But in small companies, buyers nobody has heard of sometimes show up, offer a very generous price and, when the time comes, can't or won't pay. That's when you have to play detective.

As a first filter, spend ten minutes checking who they are. Start with their website and look up their address on Google Maps: is it an office or a mailbox? Look up their executives on LinkedIn and Google, and ask yourself whether it makes sense for that company to do a deal like this. And check whether they've backed out of a deal before: if they've already left one hanging, you have every reason to pass.

That said, a nice-looking website doesn't mean they have the money. For that you have to go to the documents: who's behind the entity that signs? (It's often a subsidiary set up just for the deal.) Who's putting up the money? What commitments are there in writing, and who's on the hook if they're not met? Until I'm clear on where the money comes from, I wouldn't call the analysis done.

Where's the money coming from?

It's not enough to know someone wants to buy the company. You also need to know where the millions to pay for it are coming from. Does the buyer have the cash, or does it depend on borrowing it? A big company with lots of cash is very low risk. A private equity fund with its loans already signed, low but not zero. Someone who still has to line up the financing, high.

Then there are the termination fees, and there are two of them:

• The one the target pays if, for example, it later accepts a better offer. That's the termination fee (or break fee). If it's very high, it's not always a good sign: sometimes it's there to scare off other buyers.
• The one the buyer pays if the deal doesn't close for certain reasons, such as a missing approval or financing. That's the reverse termination fee.

The amount matters less than it seems here. What counts is in which cases the agreement can be terminated and when each fee kicks in, because a $1 billion fee doesn't help much if the contract lets the buyer walk away without paying it in exactly the scenario you're most worried about. And remember these are paid between the companies: you, as a shareholder, don't collect them directly.

Don't mix up two things either: the banks having committed to lend, and the buyer being able to cancel if the financing falls through, are separate questions. Check separately when the banks can refuse to lend, when the buyer can refuse to close, and what the target can demand if either of them doesn't deliver.

In the UK and Spain this risk is much smaller. In a UK cash offer, a third party has to confirm the buyer has the resources. In a Spanish cash tender offer, payment has to be guaranteed with a bank guarantee or a deposit before the offer is authorized.

The buyer gets cold feet

Sometimes the buyer changes its mind: the market has tanked, or it has realized it overpaid. But a signed agreement can't be broken just because. It can only be broken in the cases the contract allows.

Even so, it's worth keeping an eye on whatever might make the buyer regret it. If one gold miner buys another and the price of gold collapses while they wait for approvals, the appetite to pay the agreed price drops quite a bit, and that's when it starts looking for clauses to get out.

One of those cases is usually the target suffering a serious deterioration (a material adverse effect, or MAE). In the US, judges almost never accept it: in Delaware, where most large US companies are incorporated, the first ruling that accepted such a deterioration as grounds not to close was Akorn v. Fresenius, in 2018. On top of that, many contracts let the target ask a judge to force the buyer to close, instead of settling for damages. That's called specific performance.

Where you really do need to look closely is at the unusual clauses. An unusual condition can be exactly the way out. And don't just look at the deterioration clause. The contract also usually requires the target to keep operating normally until closing. During covid, in the AB Stable case, the Delaware courts sided with a buyer that refused to close because the seller had changed the way it operated significantly, even though there was no material adverse effect under the contract. It depends on how each agreement is written.

The shareholders say no

Another classic, especially if there are big shareholders or activist funds (the ones that buy shares to pressure the company) who think the price is too low. Ask yourself:

• Is the price low compared with where the stock traded recently? If the buyer takes advantage of a recent drop to offer little, a lot of shareholders will feel ripped off.
• Who's in charge? A couple of big shareholders are not the same as thousands of scattered small investors.
• What are they saying? The best sign is that the main shareholders have already committed in writing to vote in favor.
• Do the buyer's shareholders have to vote too? In the US this happens, for example, when the buyer pays with lots of new shares: Nasdaq rules require it if the buyer issues 20% or more of the shares or voting power it already had. That vote opens a door: a third party can bid for the buyer itself, the buyer's shareholders vote down the original deal to take that offer, and the deal falls apart. If you were also short the buyer, you lose on both sides. Julian Klymochko, who manages a merger arbitrage fund, says this does happen, and that it has caught him too.

And check what majority is needed, because it changes by country and by the company's bylaws.

What if it takes twice as long?

Even if the deal ends up going through, it can take twice as long as expected. You don't necessarily lose money, but the same spread stretched over a lot more months can turn a good investment into a mediocre one (we saw this with the annualized return). Look at the contract's deadline, the outside date, and whether it can be extended.

Sometimes someone shows up and pays more

Sometimes the surprise is a good one: a second buyer shows up willing to pay more, or shareholders push back and the buyer raises the offer. There are clues that it might happen: the agreement including a go-shop period (a few weeks when the company can go looking for better offers), and the price looking low compared with what's been paid in similar deals in the industry.

When the market smells something like that, the stock can trade above the offer. If you pay more than you'll collect in total, it's no longer enough for the deal to close. You need the terms to get better.

How to analyze your first deal

OK, all very nice, but how do you actually do this? This is roughly how I'd go about it, and it's a process you can repeat with every deal:

1. Find deals

   There are websites and newsletters that list pending deals with their spreads, some free and some paid. Start with cash deals that are already signed, in companies whose shares are easy to buy and sell. Avoid rumors and preliminary offers.

2. Read the essentials

   You don't need to read every page, but you do need to find the closing conditions, the financing, the cases in which the agreement can be terminated, and the key dates. If you couldn't explain it to a friend, it's too early to invest.

3. Go over the risks

   Who the buyer is and where the money comes from, which approvals are needed, and whether any vote is in doubt.

4. Estimate the timeline and the break price

   How long it will take (and what's left for you if it takes twice as long), and what you'd sell for if it breaks, with a central scenario and a worse one.

5. Run the numbers and set your maximum price

   Compare the break-even probability with your probability range; the calculator does it for you. If the price is above your maximum, or the edge disappears as soon as you nudge your assumptions, pass. If the numbers only work if you're right to the decimal point, that's a bad sign.

6. Ask yourself what risk you're being paid for

   Does the spread pay you for the time, the possible losses, and the work? If it's very wide for a deal that looks safe, find out why; there's usually a reason. You don't need to uncover a secret, but you do need to understand what you're taking on.

7. Decide how much to put in

   Based on what you'd lose if it goes wrong, using your bad scenario, not on what you'd make if it goes right.

8. Write down what would change your mind

   Do it before you buy, while you're still neutral. That way you'll know which news matters.

9. Follow the news until it closes

For the step where you run the numbers, the calculator is right here:

And where do you find all this? Here's the basic map:

• In the US, it's all on the SEC's website (EDGAR). The merger agreement is attached to the announcement, in a filing called an 8-K. If it's a merger with a vote, check the proxy statement, the document for the vote: its Background of the Merger section tells you who negotiated, how much each party offered, and who dropped out (it's usually the most entertaining part). If it's a tender offer, the offer is in the Schedule TO, and the target's position is in the Schedule 14D-9.
• Outside the US, look for the offer document approved by the local market regulator (in the UK, the scheme document; in Spain, the prospectus the CNMV publishes when it authorizes the offer).

How much should I put in?

The basic rule: decide how much to put in by looking first at how much you'd lose if it goes wrong, not at how much you'd make if it goes right.

For example, if a break could knock the stock down 25% and you don't want a single deal to cost you more than 1% of your portfolio, don't put in more than 4% (a 25% drop on 4% of the portfolio is exactly that 1%).

But that 25% is an estimate too, so the 1% isn't a guaranteed maximum loss. If the drop ends up being 50%, the same position costs you 2% of the portfolio. Do the math with your bad scenario, not just the central one. And set yourself a size cap per deal too: estimating a small drop doesn't make the position safe.

Eight deals aren't eight separate bets

You can hold eight deals and be a lot less diversified than you think. Several positions can soften the blow of a break, but they don't guarantee the gains will make up for the losses, especially if they share risks. Write down what each one depends on (which regulator, whether there's a fund with loans involved, whether it depends on a vote, when it's expected to close) and look at the whole list.

If five of your eight deals depend on the same competition authority and that authority decides to get tough, you could eat all five at once. So rather than counting positions, I'd look at what could make several of them go wrong at the same time.

When a crisis hits

In normal times, merger arbitrage pretty much does its own thing relative to the stock market. Until a crisis hits. In the big crashes, like 2008 or March 2020, spreads blew out across every deal at once. Not because they were all going to break, but because lots of funds using borrowed money had to sell at the same time, and they sold whatever they could.

The Mitchell and Pulvino study we saw earlier sums it up well: merger arbitrage is a lot like selling insurance against market crashes. You collect a little almost all the time, and you pay out when the market really tanks. Right when you least feel like it.

Still, it helps to put those drops in context. Here's how The Merger Fund, one of the oldest merger arbitrage funds, did in five of the biggest market drops since 2000:

How much each fell, crash by crashS&P 500The Merger Fund

2008–09

-52.5%

-1.4%

2000–02

-46.2%

-5.9%

2020

-33.8%

-6.9%

2022

-24.3%

-0.2%

2018

-19.3%

+1.1%

This fund did much, much better than the market. But it's one specific fund, with a professional team behind it, and not everyone did that well: in the February–March 2020 crash (from February 20 to March 23), a merger arbitrage ETF (MNA) lost 12.8%, and a leveraged fund would have suffered quite a bit more. A portfolio of your own with only a few deals can behave very differently. And a note on the chart: it uses the dates of the market's drops. The fund's own worst drawdown was a different one: 12.2% between October 2007 and October 2008.

It does its own thing most of the time, sure. But when the market really crashes, it falls too.

That's why you shouldn't use borrowed money (don't use leverage): it's what turns a temporary rough patch into forced selling with real losses. And keep some cash on hand: the best times to get in are usually when other people are forced to sell. Although careful, in a real crisis some buyers do try to wriggle out, so not every spread that blows out is a bargain.

Following a deal without losing your mind

These deals last months and the information keeps changing along the way. What follows works whether you've bought or you're just following a deal without investing.

Write everything down

The facts of a deal are scattered across press releases, contracts, and filings that come out over months. If you don't write them down, in three weeks you won't remember why on earth you decided to buy or pass. A simple worksheet solves that.

The trick is the last column, which separates three things. For facts, also note where they come from (in the example, the section of the agreement), so you can find them again three months later:

• Facts: what you know and can back up with a source. "Shareholders vote on October 15."
• Estimates: what you think will happen. "I think they'll approve the deal."
• Questions: what you still need to check. "Not sure whether approval in China is needed."

Don't mix the three. And note when you last checked each one.

See the rows about the agreement

What to write down	In our example	Type
Who's putting up the money	The parent company, with its own cash and a loan already signed	Fact · agreement § 5.8
Is financing a condition to close?	No	Fact · agreement § 7.2
What can the target demand if the buyer doesn't close?	Ask a judge to force the buyer to close (specific performance), if the agreement's requirements are met, or the fee in the cases provided for; check whether they're alternatives and what limits apply	Fact · agreement § 9.5
Fees and when they apply	3% from the target if it accepts another offer; 5% from the buyer if an approval is missing	Fact · agreement § 8.3
Outside date	12 months out; extended by 3 months if only an antitrust approval is missing	Fact · agreement § 8.1
Required majority and what it's based on	Majority of all outstanding voting shares (not just those present at the meeting)	Fact · agreement § 7.1
If it were a tender offer: minimum acceptance	What percentage is needed, what shares it's based on, whether the buyer can waive that condition, and what happens if it isn't reached	Question (doesn't apply here)
Buyer's commitments to get the approvals	Sell businesses adding up to 5% of its revenue; not required to go to court	Fact · agreement § 6.2

Don't mix up three dates that show up on the worksheet: when you think you'll get paid, the agreement's outside date, and the last day your broker gives you to act. They don't mean the same thing and they're not for the same thing.

When something changes, it's not enough to add it to a list. Write it under "last change" in this format: before → now → what I do. For example: closing expected in month 6 → month 12 at the earliest, because the regulator wants more information → I update the timeline, revisit the break price, and recalculate the maximum price.

Also look at the "what's left before closing" row (✓ done, ○ pending). Thinking about each hurdle separately (the vote, each regulator, the money) shows you very quickly where it can really break. The risk is almost always concentrated in one or two of them, and that's where you should spend your time. But don't treat them as independent: what happens to one usually affects the others, so don't just multiply their probabilities.

And a big open question doesn't get fixed by using a more cautious probability. If you still don't know which approvals are needed or which conditions allow the deal to be called off, the worksheet isn't ready for a decision. Keep reading.

There's news. Now what?

While you follow a deal, things keep coming out: filings, earnings, regulator news, lawsuits, extensions... Almost all of it is noise, but a few things change everything. To tell them apart, write down before buying what would make you change your mind. For example:

I think shareholders will vote in favor, because a third have already committed and the offer pays 40% more than the stock traded at before the announcement. I'd change my mind if a major shareholder comes out against it publicly or if the proxy advisors recommend voting against.

Writing it down before the news arrives helps you notice if, later, you start bending the story to justify what you already wanted to do (spoiler: we all do it). And when something comes out, you'll know where to look first. That said, the list is a guide, not a closed list: if something important comes up that you didn't see coming, redo the math anyway.

So passing on a deal can also be a good outcome of the analysis, even if the deal ends up closing.

Following one deal like this is fine. Following ten at once is another story: ten worksheets, ten deadlines, and a pile of filings that keep coming out without warning, so it's very easy to miss something. That's why I'm building Tracking Alpha: a free list of every live US cash merger and, for the ones you hold, a worksheet like the one above that gets checked every day. I'm aiming to open it in December, and you can already join the waitlist:

When to get out early

You don't always have to wait until the last day. It makes sense to sell early if:

• There's almost nothing left to make. If the spread has narrowed so much that it pays less than T-bills, you're carrying the risk for free.
• The story changes. Ask yourself a very simple question: if you didn't own the position today, would you buy it at this price with what you know now? If the answer is no, think about getting out. The price you paid doesn't change what the deal offers today.
• You find something better. Your money is limited. Put it in the deal that pays you best for the risk.

Don't sell automatically just because the stock has dropped: that's what people with borrowed money are forced to do. But don't assume nothing has changed either. Check the sources and your assumptions. If you can't find an explanation for the drop, own that uncertainty before holding or adding to the position.

And if the deal breaks? Your merger arbitrage thesis is over. Keeping the shares can make sense, but that's a different investment: owning a company with no buyer. Ask yourself whether you'd buy it today, on its own, at the current price, and look at the business and the position size through that lens. And "waiting to get back to what I paid", however tempting it is for all of us, is not a reason to stay.

Payday

It depends on the kind of deal:

• In a merger with a vote, you don't have to do anything. When the deal closes, your shares are swapped for cash, which shows up in your account a few days later depending on your broker.
• In a tender offer, check whether you need to give your broker instructions to tender and what their deadline is (usually a few days before the official expiration). In the US you can generally withdraw your shares while the offer is still open; once it expires and the buyer accepts them, your shares are committed until you're paid.

If you don't tender, you can end up as a minority shareholder. In many deals (in the US, quite a lot) there's a merger afterwards that also converts the shares of those who didn't tender, but check the documents.

On your own or through a fund?

On your own

You need a broker with access to the markets where the deals are (mostly the US). With cash deals you don't need to short anything: you buy the stock and wait (if it's a tender offer, keeping an eye on your broker's instructions and deadlines). With stock deals, if you want to hedge against what the buyer's stock does, you'll have to short it, and that's a whole other level.

It's the option that gives you the most control and also the one that asks for the most time.

Through a fund

There are funds that specialize in merger arbitrage, including mutual funds and ETFs. Two things to keep in mind:

• Fees weigh a lot. If the strategy makes a few points more than T-bills, a 1.5% annual fee eats a good chunk of that extra. Compare the fund with T-bills, not with the stock market.
• If you invest from Europe, many US ETFs are off-limits. EU rules (PRIIPs) require a key information document (the KID) that most of them don't have, so you'd usually look at European (UCITS) funds instead. Individual stocks don't have this problem: you can buy those directly.

Can you beat the pros?

In some ways, yes (without getting carried away):

• You're small. You can look at deals that aren't worth a big fund's while (with the money they manage, they barely move the needle), and there may be less competition there. That said, those deals aren't safer: a small company can dominate a very specific market, and that's exactly where shady buyers show up most.
• You're patient and you're not leveraged. Nobody forces you to sell during a long review, and you can buy when others have to sell.
• Maybe you know something. If you know an industry or its regulators well, you can have an edge in certain deals.

In other ways, no: you pay more in commissions, you have worse information, and reading contracts takes time.

Taxes and currency

Taxes depend on where you live, so treat this as a starting point and check your own situation.

• In the US, a cash deal is a sale. When the deal closes, you're taxed on the gain as if you'd sold. Most deals close in less than a year, so the gain is usually short-term and taxed at your ordinary income rate, not the lower long-term rate. That takes a real bite out of a spread that was thin to begin with, so do the math after tax.
• If it breaks and you sell at a loss, watch out for the wash sale rule: if you buy the same stock (or something substantially identical) within 30 days before or after the sale, you can't deduct that loss yet.
• Stock deals and foreign dividends. Some stock-for-stock deals are tax-free and others aren't, and foreign dividends can come with withholding. The proxy usually has a section on the tax consequences; read it before you buy.
• If your money isn't in dollars, the exchange rate matters. A 3% gain can turn into a loss if the dollar drops 4% while you wait. You can hedge it (at a cost) or live with it, but factor it in.

Extra credit

How to hedge a stock deal

If the deal is paid in stock and you only buy the target, you're really making two bets: that the deal goes through and that the buyer's stock doesn't fall. To keep only the first one, you short the buyer's stock in the same ratio as the exchange.

In practice, there's fine print:

• Borrowing shares costs money, sometimes quite a lot in popular deals, and they can be recalled at a bad time.
• You pay the buyer's dividends while you're short.
• You need margin. The $50,000 from the short stays as collateral, the broker asks for more if things go wrong, and it can close your position before the deal closes.
• The buyer can end up being bought. If someone bids for BuyerCo, its shares go up and your short loses money. And the TargetCo deal can fall apart along the way, so you lose on that side too.
• If the deal breaks, the hedge can make it worse. Usually the buyer goes up and the target goes down, so you lose on both sides.

Collars and CVRs

Some stock deals come with a collar: a formula that changes depending on where the buyer's stock trades relative to a price band. Depending on the agreement, inside that band either the value you receive or the number of shares stays fixed, and outside it a different rule applies. To hedge, you need to understand the specific formula for your deal.

Others include a CVR: an extra payment you only collect if something specific happens, like a drug getting approved. The market usually values them at quite a discount, and sometimes for good reason: the buyer controls many of the decisions that determine whether you ever get paid.

How long each kind of deal takes

• In the US, a tender offer with no regulatory problems can close in just over a month. Two clocks are running. One is the offer's minimum period, normally 20 business days (since April 2026, some negotiated, all-cash tender offers for 100% of the shares can go down to 10). The other is the antitrust waiting period, usually 30 calendar days (15 for cash tender offers), as the FTC explains. A merger with a vote usually takes several months.
• The US antitrust review (known as HSR). After the deal is filed, either the waiting period simply expires (a good sign, although it's not the same as an explicit approval), or the FTC or the Department of Justice issues a second request, the request for a lot more information we saw earlier, and that opens an investigation that lasts months.
• In the UK, a scheme of arrangement is very common. Shareholders vote (it takes a majority of those voting and 75% of the value voted, under the Companies Act) and then a court approves it. If it goes through, it applies to everyone.
• Competition regulators have their own timelines. The European Commission has 25 business days for a first review and 90 for an in-depth one. The UK CMA, 40 business days and 24 weeks. These are per-phase timelines, not for the whole process. They start with a valid filing or the formal opening of that phase, not with the deal announcement, and they can be extended.

Tender offers in Europe

European tender offers follow their own rules, set country by country. A few things to look for in the offer document, using Spain as an example:

• A minimum acceptance. Many voluntary offers only go ahead if a minimum percentage of shares is tendered. If it isn't reached, the offer lapses, unless the buyer waives that condition.
• Squeeze-out and sell-out. In Spain, if the buyer reaches 90% of the voting capital and at least 90% of the voting rights the offer was aimed at have accepted, it can force the rest to sell at the same price, and you can force it to buy your shares. Both rights only last three months after the acceptance period ends (Royal Decree 1066/2007).
• If those thresholds aren't reached, you can get stuck as a minority shareholder in a company someone else controls, with very few people buying and selling. Not a comfortable place to be.

If you want to keep reading

• You Can Be a Stock Market Genius, by Joel Greenblatt. A classic on special situations, with a chapter on mergers and arbitrage. Very easy to read.
• How to Profit from Special Situations in the Stock Market, by Maurece Schiller. Old, but with the timeless ideas.

My advice: start by watching

From the outside, merger arbitrage looks simple: you buy a stock for $48.50 and wait for someone to pay you $50. The hard part is knowing how likely you are to collect that $50, how long it'll take, and what happens to the stock if it never comes. To decide whether a deal is worth it, compare what you can make with how long it'll take and with what you'd lose if it breaks. And don't put too much into any single one.

If you're curious, start by watching. Pick three or four cash deals that are in progress, run the numbers (the calculator saves you the division) and follow them to the end, without putting money in. After you've watched one close, another get delayed, and maybe another break, you'll understand the strategy much better than from reading twenty definitions. And you'll know pretty quickly whether this is for you. And if you get a few going and keeping up with them gets to be a pain, that's what I'm building Tracking Alpha for.