Definitions of the term “Material Adverse Effect” (“MAE”) in business combination agreements typically contain exceptions from the definition for events arising from certain systematic risks.[1] Common exceptions pertain to general changes in business or economic conditions, changes in financial or securities markets, and changes in the industries in which the target operates. Other common exceptions pertain to force majeure events or changes in law or generally accepted accounting principles (“GAAP”). Such exceptions have figured prominently in the MAE cases arising from the COVID-19 pandemic, such as AB Stable[2] and KCake[3] in Delaware and Fairstone[4] in Canada. Indeed, these cases have revealed a latent ambiguity in the typical MAE definition: sometimes the causal background of a material adverse effect on the target is complex, with an earlier event E1 (such as a pandemic) causing a later event E2 (such as governmental lockdown orders) and the later event E2 (the lockdown orders) causing the material adverse effect on the target. If the MAE definition allocates the risk of both events to the same party, then clearly that party bears the risk of the material adverse effect, but what happens if the definition allocates the risk of one event to one party and the risk of the other event to the other party? If the target bore the risk of a pandemic but the acquirer bore the risk of lockdown orders (under an exception related to changes in law), has there been a Material Adverse Effect or not?
Albeit only in dicta, the courts in AB Stable[5] and KCake[6] both resolved this particular problem by saying that there would be no Material Adverse Effect. The court in Fairstone reached a similar conclusion.[7] Further complicating the problem, the court in KCake also considered the effect of language in the MAE definition introducing the exceptions and expanding their scope to include not only events falling into the exceptions but also events “arising from” or “related to” such events. The problems that these cases raise thus involve the meaning of the language in the exceptions, the relation of the exceptions to the main part of the MAE definition and to each other, and the effect of the language introducing the exceptions.
Starting with First Principles
In a new article on Pandemic Risk and the Interpretation of Exceptions in MAE Clauses[8] forthcoming in the Journal of Corporation Law, I argue that, to make sense of these issues, we should begin by taking a step back and reflect on the general structure of a typical MAE definition. In my view, there are two aspects of that definition that are critically important in this context. First, in MAE definitions, the (capitalized) term “Material Adverse Effect” is defined to mean any event that has or would reasonably be expected to have (the exact language varies) an (uncapitalized) material adverse effect on the target. The definition thus involves two separate things, an event and a material adverse effect that the event causes, which are related as cause to effect. It is counterintuitive but nevertheless apparent from the face of the definition that a (capitalized) Material Adverse Effect is not an (uncapitalized) material adverse effect. Rather, a Material Adverse Effect is an event that causes (i.e., “has or would reasonably be expected to have”) a material adverse effect.
Of course, good transactional lawyers know all this, but they nevertheless often speak imprecisely, running together Material Adverse Effects and material adverse effects, an infelicitous habit that is greatly facilitated by the fact that the three-letter abbreviation “MAE” is used indiscriminately for both concepts. Confusions of Material Adverse Effects with material adverse effects crop up in ordinary speech when someone says that a company “has suffered a Material Adverse Effect”; if anything, the company suffered a material adverse effect because some event occurred that was a Material Adverse Effect. Some confusions even appear in the text of merger agreements negotiated by expert counsel, as when a representation is required to be true except for “inaccuracies that would not have a Material Adverse Effect”; if anything, the inaccuracies just are a Material Adverse Effect because they are facts or events that would reasonably be expected to have a material adverse effect on the company. Because, as we shall see, such confusions are not always harmless, I shall distinguish in this article clearly and consistently between Material Adverse Effects and the material adverse effects that they cause.
The second feature of MAE definitions I want to emphasize concerns how they allocate risk between the parties. In particular, such definitions allocate risk on the basis of events causing material adverse effects, not on the basis of material adverse effects arising from such events. That is, the definition takes the universe of all possible events and divides these events into two classes—events the risk of which is allocated to the target, and events the risk of which is allocated to the acquirer. In particular, MAE definitions effect this division by saying all events causing a material adverse effect are Material Adverse Effects, except for events falling into one of the exceptions enumerated in the definition. Hence, the risk of an event causing a material adverse effect is allocated to the target, unless the event falls into an exception, in which case the risk of that event is allocated to the acquirer. Exceptions in the MAE definition apply to events, and whether the risk of an event is allocated to the acquirer or the target depends on whether the event falls into an exception or not.
What Is Included with an Allocated Risk
The fact that Material Adverse Effects are not material adverse effects but events causing material adverse effects, and the fact that MAE definitions allocate risks on the basis of events not effects are both evident from the plain language of the typical MAE definition. For this reason, I think both are well-nigh indisputable. There is another key point about the typical MAE definition, however, that is implicit in the definition and so not immediately evident from the text. This point concerns exactly which risks are allocated when an MAE definition allocates to one party or another the risk that a certain event will occur. To grasp this point, notice that MAE definitions allocate the risks that certain events may occur not because the parties care about those events in and of themselves; MAE definitions allocate the risks that certain events may occur because those events may have a material adverse effect on the target. That is why MAE definitions define Material Adverse Effects in terms of their effects—that is, they define an event to be a Material Adverse Effect in terms of the event’s having, or being reasonably expected to have, a material adverse effect on the target. Therefore, when an MAE definition allocates to one party or the other the risk that a certain event may occur, what is being allocated is the risk that the event occurs along with all of the event’s reasonably-expected consequences, up to and including any reasonably-expected material adverse effect on the target.
To see just what that means, reflect that events rarely have material adverse effects on a company immediately and directly. On the contrary, events typically result in material adverse effects, if at all, only through generally predictable causal pathways. Indeed, parties single out certain kinds of events and specifically allocate the risks of those events precisely because the parties know that such events tend to set in motion a sequence of events that often lead to a material adverse effect. For example, suppose the MAE definition allocates to one party or the other the risk of “a change in interest rates,” and after the agreement is signed the Federal Reserve’s Federal Open Market Committee (the “FOMC”) announces that it has decided to increase its target Federal Funds Rate. This decision is a “change in interest rates,” and the risk of any material adverse effect reasonably expected to follow from that change is allocated to the party that bears the risk of such change. But any adverse effect on the target resulting from this decision by the FOMC would come about only through a long and complicated, but nevertheless reasonably-expected, sequence of events. In particular, the FOMC’s decision to increase the target Federal Funds Rate will result in the Trading Desk at the New York Federal Reserve Bank making certain purchases of securities in the open market, and this, via the decisions of bond traders and commercial bankers, will likely result in a change in the effective Federal Funds Rate, which is an average of the rates depository institutions actually charge each other for overnight loans of banking reserves. That change, via the trading decisions of innumerable market participants, will filter through the credit markets, eventually increasing the yield on long-term Treasury bonds, which will (under accepted principles of corporate finance) increase the company’s cost of equity capital, thus reducing the present value of its future cashflows. When the parties allocate the risk of a change in interest rates, they do so precisely because they understand that changes in interest rates tend to have such consequences, up to and including a reduction in the value of the target when the present value of its future cashflows is reduced in the manner stated. It would subvert the intention of the parties were someone to later argue that, because the causal sequence from the decision by the FOMC to the reduction in the present value of the company’s future cashflows runs through a great many other events, the material adverse effect on the target should be attributed not the decision by the FOMC but to some intermediate event in the causal sequence, such as the decisions of bonders traders reacting to the FOMC’s decision. On the contrary, in allocating the risk of a change in interest rates, the parties are allocating the risks of all events that would reasonably be expected to follow from a change in interest rates, up to and including any reasonably-expected material adverse effect on the company. More generally, when an MAE definition allocates the risk of a certain event, it allocates the risk of all other events reasonably expected to follow from that event. Indeed, the whole point of allocating the risk of the event is to allocate the risk of the reasonably-expected consequences of the event, again, up to and including any material adverse effect on the company.[9]
How to Apply MAE Definitions in Causally Complex Cases
Such considerations show us how we should apply MAE definitions when the causal background of a material adverse effect runs through multiple events and the definition allocates the risks of different events to different parties. Thus, suppose that a first event E1 (say a pandemic) causes a second event E2 (such as governmental lockdown orders) and the second event E2 (the lockdown orders) causes a material adverse effect on the company (because the orders curtail its operations), and suppose further that the MAE definition allocates the risk of E1 (the pandemic) to the target and the risk of E2 (the lockdown orders, under an exception for changes in law) to the acquirer. In such cases, when event E1 (the pandemic) occurs, the relevant question under the MAE definition is whether that event would reasonably be expected to have a material adverse effect on the target. If so, since the target bore the risk that event E1 (the pandemic) would occur, E1 is a Material Adverse Effect. This is true even though the causal pathway from E1 (the pandemic) to the material adverse effect runs through another event E2 (the lockdown orders) that is of a kind that, generally speaking, falls into an exception, which would shift the risk of such an event to the acquirer. The reason is that, in allocating to the target the risk of a pandemic, the MAE definition allocated to the target everything that is reasonably expected to follow from a pandemic, and if a pandemic actually occurs and is such that it would reasonably be expected to result in certain lockdown orders, then the risk of such orders was included with the risk of the pandemic. As a matter of contract interpretation, the specific governs over the general, and even if the risk of changes in law was generally allocated to the acquirer, nevertheless in allocating the risk of a pandemic to the target, the MAE definition also allocated to the target the specific risk of lockdown orders reasonably-expected to follow from a pandemic. This is the reasonable way of giving effect to both allocations of risk made by the MAE definition.
Now, although in allocating the risk of a pandemic to the target the MAE definition also allocated the risk of all events reasonably expected to follow from a pandemic, including any reasonably-expected lockdown orders, nevertheless the acquirer could concede arguendo that the risk of lockdown orders resulting from a pandemic was allocated to it under the exception for changes in law and yet still succeed in showing that there had been a Material Adverse Effect. The reason is that, in situations like this, where the risk of the more remote event E1 (here the pandemic) is allocated to the target and the risk of the more proximate event E2 (here the lockdown orders) is allocated to the acquirer, there is an even stronger argument available to the acquirer, an argument that does not require the acquirer to rely on the fact that, in allocating a risk, an MAE definition allocates the risk of everything reasonably expected to follow from that risk. That is, the acquirer can argue that, if the more remote event E1 (the pandemic) occurs and would reasonably be expected to have a material adverse effect on the target, then that event is a Material Adverse Effect under the express terms of the MAE definition. It may well be perfectly true, the acquirer may say, that some other event E2 (the lockdown orders) is not a Material Adverse Effect because the risk of such an event was allocated to the acquirer. But the fact that some other event is not a Material Adverse Effect in no way changes the fact that E1 (the pandemic), the risk of which was allocated to the target, is a Material Adverse Effect. The acquirer will say that, in arguing that the risk of a change in law was allocated to the acquirer and so cannot be a Material Adverse Effect, the target is like the defendant who, charged with the murder of Jones, proves he did not kill Smith.
Now, what is good for the goose is good for the gander, or, more accurately, MAE definitions treat all events in the same way, whether the risk of the event is allocated to the target or to the acquirer. Thus, suppose again that a first event E1 (a pandemic) causes a second event E2 (lockdown orders), with event E2 (the lockdown orders) causing a material adverse effect on the company, but this time suppose that that the MAE definition allocates the risk of event E1 (the pandemic) to the acquirer and the risk of event E2 (the lockdown orders) to the target (because there is no exception for changes in law). If a pandemic occurs and is such that it would reasonably be expected to have a material adverse effect on the target, then that risk has been allocated to the acquirer, and so the pandemic is not Material Adverse Effect. If the acquirer says that it bore the risk of a pandemic but not the risk of changes in law, and the material adverse effect on the target resulted from governmental lockdown orders (albeit ones that themselves resulted from the pandemic), the answer is that in allocating the risk of a pandemic to the acquirer, the MAE definition also allocated to the acquirer the risk of all events reasonably expected to follow from the pandemic up to and including any material adverse effect on the target. Hence, if the pandemic results in lockdown orders of a kind reasonably expected to follow from the pandemic, and those orders have a material adverse effect on the target, the risk of such orders was allocated to the acquirer along with the risk of the pandemic itself. It does not matter that, in general, the MAE definition allocated the risks of changes in law to the target. In this specific case (and the specific governs over the general), the risk of these changes in law (the lockdown orders reasonably to be expected to follow from the pandemic) were specifically allocated to the acquirer when the acquirer agreed to bear the risk of a pandemic and, by implication, everything reasonably to be expected to follow from any pandemic that actually occurs.
It is important to understand, however, that the target cannot concede arguendo that the risk of the change in law was allocated to it and still claim it is entitled to prevail in the way the acquirer could when the roles were reversed and the target bore the risk of the more remote event E1 (the pandemic) and the acquirer bore the risk of the more proximate event E2 (the lockdown orders). The acquirer could make this concession and still prevail because it could argue that the more remote event, E1 (the pandemic), the risk of which was allocated to the target, was still a Material Adverse Effect even if some other event, such as E2 (the lockdown orders), was not a Material Adverse Effect because that other event fell into an exception. The target cannot make an analogous argument because such an argument would amount to saying that, since the more remote event E1 (the pandemic) is not a Material Adverse Effect (because the risk of such an event was allocated to the acquirer), it does not matter that the more proximate event E2 (the lockdown orders), the risk of which was allocated to the target, is a Material Adverse Effect. Clearly, such an argument does nothing to help the target, for it most certainly would matter that there was another event that was a Material Adverse Effect. Indeed, such a point would be decisive. This form of the argument (as long as one event is a Material Adverse Effect, it does not matter that some other event is not) is available to the acquirer while the analogous form of argument (as long as one event is not a Material Adverse Effect, it does not matter than some other event is) is not available to the target because of the fundamental asymmetry between acquirers and targets in relation to MAE clauses. That is, for the acquirer to prevail, there need be only one event that is a Material Adverse Effect, and so the acquirer can shrug off the existence of events that are not Material Adverse Effects; but for the target to prevail, there need be no events that are Material Adverse Effects, and so the target cannot shrug off events that are Material Adverse Effects.
This asymmetry puts the target at a disadvantage in cases where the risk of the more remote event E1 is allocated to the acquirer and the risk of the more proximate event E2 at least appears to be allocated to the target in the sense that it is not expressly covered by an exception from the MAE definition. As explained above, under the correct interpretation of the MAE definition, the allocation of the risk of event E1 includes the allocation of the risk of all events reasonably expected to follow from event E1, including event E2 (assuming E2 really would reasonably be expected to follow from E1). The target is at a disadvantage, however, because, unlike the acquirer, it cannot make the kind of argument explained above but would have to argue that, in allocating the risk of the more remote event E1 (the pandemic) to the acquirer, the MAE definition also allocated the risk of all events reasonably expected to follow from E1, which would include event E2 (the lockdown orders). This argument is perfectly sound, but the target would still have to make it, which puts the target at a disadvantage relative to the acquirer in the analogous situation. There is, however, a simple drafting solution to this problem: the MAE definition can make explicit what is already implied by expressly providing that, when the risk of a certain event is allocated to the acquirer under an MAE exception, allocated along with that risk are the risks of all events reasonably to be expected to follow from that event. And, in fact, the language introducing the exceptions in the MAE definition typically does exactly this by expressly stating that excepted from the definition are not only events falling into the exceptions but also all events “arising from” events falling into the exceptions. This language thus restores a certain parity between the target and the acquirer by ensuring that the risks of events allocated to them are treated in the same way.
Why the Courts Have Misread the Exceptions in MAE Definitions
If all this correct, then the courts that have thus far confronted these issues have been misreading the exceptions in the MAE definitions before them in rather serious ways. Take AB Stable, by far the best-reasoned of these cases. Albeit only in dicta,[10] the court in that case suggested that, even if the target had borne the risk of a pandemic, exceptions in the MAE definition for changes in business conditions, changes in industry conditions, and changes in law would each have applied, with the result that there was no Material Adverse Effect.[11] Now, the court’s argument for that conclusion has three main steps. In the first, the court stated that the various exceptions to the MAE definition are to be read independently of each other in the sense that whether an event falls into an exception depends on the language in that exception and is independent of whether the event would fall also into some other exception.[12] That proposition, I think, is entirely correct.
In the second step, however, the court said that whether an event falls into an exception is also independent of whether any event causing that event falls into an exception, and thus, for example, governmental lockdown orders would fall into an exception related to changes in law even if the orders resulted from a pandemic and the pandemic was not excepted.[13] Although I myself once argued for exactly that proposition[14] (and the court cited one of my working papers in reaching that conclusion),[15] for the reasons given above I now think that this is mistaken. The better view is that, when the MAE definition allocates the risk of an event, it allocates along with that risk the risk of all the reasonably-expected consequences of the event. Hence, if a pandemic is such that it would reasonably be expected to result in lockdown orders, then the risk of those orders is allocated along with the risk of the pandemic. It was a mistake for the court to treat events reasonably expected to follow from the pandemic as being separate events under the MAE definition.
But even making this mistake, the AB Stable court could still have reached the right result, for it was still open to the acquirer to argue that, although the lockdown orders were not a Material Adverse Effect (because they fall into the exception for changes in law), nevertheless the pandemic, a quite different event, was a Material Adverse Effect, for it did not fall into any exception and would reasonably be expected to have a material adverse effect on the target. As explained above, this reasoning is perfectly sound, and although it is not clear whether the acquirer made this argument, it is clear that the court would have rejected it because of an additional and even more serious mistake in the third and final step of the court’s argument. That step is left mostly implicit, but from what the court says explicitly elsewhere in the opinion, we know that the court thinks that if “the cause of the [material adverse] effect fell within an exception to the MAE Definition,” then “the effect could not constitute a Material Adverse Effect.”[16]
This innocuous-sounding sentence conflates material adverse effects and Material Adverse Effects and leads to a near reversal of the internal logic of the MAE definition. That is, in the situation that the court is considering, an event causing a material adverse effect falls within an exception. What follows from this is that this event is not a Material Adverse Effect. What the court thinks follows from this is that the material adverse effect caused by the event is not a Material Adverse Effect. But this latter conclusion is nonsense, for it is events causing material adverse effects that are Material Adverse Effects (if the event is not excepted) or not Material Adverse Effects (if the event is excepted); material adverse effects themselves cannot be Material Adverse Effects. Thinking they could be is a category mistake. Moreover, confusing material adverse effects with Material Adverse Effects in this context leads the court to the erroneous view that exceptions in the MAE definition apply to material adverse effects in the sense that, if an excepted event causes a material adverse effect, then not only is that event itself excepted but the material adverse effect it causes is excepted too, and so any other event causing that material adverse effect is treated as if it were excepted as well, even if it is plainly not excepted. That is why, in the court’s view, if the lockdown orders (an excepted event) cause a material adverse effect, the pandemic that caused the same material adverse effect (via the lockdown orders) is not a Material Adverse Effect: on the court’s view, the pandemic is not a Material Adverse Effect because the material adverse effect it causes has been “excepted” since it was also caused by an excepted event (the lockdown orders). This is all quite wrong, of course, because the exceptions in an MAE definition plainly apply to events causing material adverse effects and not to material adverse effects themselves, and the fact that one event is not a Material Adverse Effect does not generally imply that some other event is not a Material Adverse Effect.
Now, as I argued above, on the correct reading of the MAE definition, acquirers enjoy a certain natural advantage in MAE disputes. That is, the acquirer prevails if there is even one Material Adverse Effect, that is, even one unexcepted event that would reasonably be expected to have a material adverse effect. The existence of other events that are not Material Adverse Effects is irrelevant. By contrast, the target prevails only if every event is not a Material Adverse Effect, that is, every event is either excepted or would not reasonably be expected to have a material adverse effect. The court’s reading of the MAE definition negates this natural advantage enjoyed by the acquirer and confers an analogous advantage on the target. Under the court’s reading, if there is even one excepted event that would reasonably be expected to have a material adverse effect, then that material adverse effect is excepted, and no event, excepted or unexcepted, causing that effect is a Material Adverse Effect. Hence, for the acquirer to prevail, every event causing a material adverse effect must be unexcepted, and if there is even one excepted event causing that material adverse effect, the target wins. Put yet another way, if there are many events causing a material adverse effect, under the proper reading of the MAE definition, the acquirer wins if even one of them is unexcepted; under the court’s reading, the acquirer wins only if every one of them is unexcepted. By making the MAE exceptions apply to material adverse effects rather than the events causing them, the court’s reading has negated the natural advantage of the acquirer and conferred an analogous advantage on the target.
Furthermore, recall that, when the risk of a remote event E1 (such as a pandemic) was allocated to the acquirer and the risk of a proximate event E2 (such as lockdown orders) was allocated to the target, the acquirer’s natural advantage under the proper reading of the MAE definition (it takes only one unexcepted event having a material adverse effect for there to be a Material Adverse Effect) creates the possibility that the acquirer may ignore the principle that for purposes of the MAE definition an event includes all its reasonably-expected consequences and argue speciously that, even if the remote event E1 (the pandemic) was not a Material Adverse Effect, the proximate event E2 (the lockdown orders) resulting from E1 was a Material Adverse Effect. Recall, too, that such specious arguments can be blocked by language in the MAE definition introducing the exceptions and providing that events arising from excepted events are excepted. The court in AB Stable, by treating exceptions in the MAE definition as if they applied to material adverse effects rather than events causing material adverse effects, has not only negated the acquirer’s natural advantage under the MAE definition and conferred an analogous advantage on the target, but it has also created the possibility of the target making analogously specious arguments against the acquirer. That is, when the risk of a pandemic is allocated to the target, but the risk of changes in law is allocated to the acquirer, the target ought not be permitted to distinguish between the pandemic and any lockdown orders reasonably expected to arise from it. To block such a move, there is a drafting solution analogous to the language introducing the exceptions and providing that events arising from excepted events are excepted: such a solution would involve having the MAE definition provide that events arising from non-excepted events are not excepted.
Unsurprisingly, therefore, the court in AB Stable pointed out that its interpretation of the MAE exceptions could be blocked if the language in the exceptions included a proviso to the effect that otherwise excepted events would not be excepted after all if they arose from unexcepted causes.[17] Of course, such a proviso should be doubly unnecessary. It is unnecessary in the first instance because, in allocating the risk of an event to the target, the MAE definition allocates as well the risk of any event reasonably expected to follow from the event. It is unnecessary a second time because, if exceptions are understood to apply to events causing material adverse effects and not to material adverse effects themselves, even if an event arising from an unexcepted event is excepted and so not a Material Adverse Effect, nevertheless the unexcepted event from which the excepted event arises may be a Material Adverse Effect in its own right, which is all the acquirer needs to prevail. Hence, the fact that the court noted that parties could add a proviso to the MAE definition to make events arising from unexcepted events unexcepted shows again that the court had reversed the basic logical structure of that definition.
In sum, there are two problems with the reasoning in AB Stable. The first and less serious is a failure to appreciate that, in allocating the risk that a certain event will occur, the MAE definition also allocates the risk of all events reasonably expected to follow from the event, up to and including any reasonably-expected material adverse effect on the target. The second and more serious is that, by conflating Material Adverse Effects with material adverse effects at a critical point in the argument, the court has made exceptions in MAE definitions apply to material adverse effects rather than Material Adverse Effects,[18] which largely reverses the internal logic of the MAE definition.
The Origin of the Confusion in AB Stable
Now, as I mentioned above, even the most expert and experienced transactional lawyers sometimes conflate Material Adverse Effects and material adverse effects, which suggests that the AB Stable court ought not be judged harshly for falling into this mistake at a critical point in the argument. In fact, however, the particular form of the confusion in AB Stable, making exceptions in MAE definitions apply to material adverse effects instead of events causing them, is not the court’s fault at all. If anyone is responsible for this mistake, I am. The mistake existed in both scholarly and practitioner literature for years before AB Stable, but the earliest example of this mistake that I have found is in one of my own law review articles from 2009.[19] There, I said, “When … exceptions are present [in an MAE definition], adverse changes to the company resulting from such causes are not MACs within the meaning of the definition.”[20] Replace the word “changes” with “effects” and the abbreviation “MAC” with “MAE,” and what I said in 2009 is exactly what the court in AB Stable said in 2020: if an event causing a material adverse effect falls into an exception, the material adverse effect that the event causes is not a Material Adverse Effect, which makes the exceptions apply to material adverse effects rather than the events causing them, and so entails all the erroneous consequences exposed and deprecated above. The confusion I condemn here is thus one that I myself invented. Whether everyone who has fallen into this mistake has done so under the influence of my writings is doubtful; because of the pervasive confusion of Material Adverse Effects with material adverse effects, this was a mistake waiting to happen. But given the unpleasant choice, I would rather accept the whole blame and be accused of grandiosity in appropriating the responsibility to myself than blame others for making the mistake independently and be accused of shirking the responsibility for the mistake.
The Effect of the Language Introducing the Exceptions
There remains one final point about the language introducing the exceptions in the MAE definition. A new article by Professor Guhan Subramanian (Harvard University) and Caley Petrucci (Wachtell, Lipton, Rosen & Katz)[21] forthcoming in the Columbia Law Review includes an impressive empirical study of MAE definitions in public-company merger agreements and reports that the language introducing the exceptions in an MAE definition tends to come in one of two forms. Sometimes, as discussed above, the language introducing the exceptions provides that, besides events falling into the exceptions, events “arising from” excepted events are excepted; in other cases, the language provides that, besides events falling into the exceptions, events “related to” excepted events are excepted.[22] The authors suggest that the latter language is broader than the former, that is, would result in more events being excepted.[23]
To make sense of this, I think it helps to take a step back and recall that, in the simplest case, the language introducing the exceptions would say only that no event falling into an exception “is” or “shall constitute” a Material Adverse Effect. That is, if we understand the exceptions as listing certain kinds of events the risk of which are being allocated to the acquirer, the introductory language need only say that no event falling into an exception will count as a Material Adverse Effect even if the event, should it occur, would have or would reasonably be expected to have a material adverse effect on the target. Moving to more complex cases, we see that the introductory language can expand the set of excepted events to include as well events “arising from” events falling into the exceptions. Given the discussion above, the purpose of such language is clear: when an MAE definition allocates the risk of an event, by implication it allocates as well the risk of all events reasonably expected to follow from the event up to and including any material adverse effect on the target. But this is by implication, and adding the “arising from” language to the introduction of the exceptions makes this explicit. That is particularly important for the target, because, as explained above, given the logical structure of the MAE definition, the acquirer enjoys a natural advantage in that, under a proper reading of the definition, all events are non-excepted events (unless they happen to fall into an exception), and so in particular events arising from non-excepted events are non-excepted events (unless they happen to fall into an exception). To restore some form of parity between events the risk of which are allocated to the target and events the risk of which are allocated to the acquirer, parties can agree to language introducing the exceptions that provides that events arising from events falling into exceptions are also excepted.[24]
But what if the language introducing the exceptions excepts not just events arising from excepted events but also events merely “related” to excepted events? Wouldn’t this, as Subramanian and Petrucci say, greatly expand the scope of the exceptions? I don’t think so. In fact, reading this language expansively—i.e., reading “related” to mean “related in any way whatsoever”—would make nonsense of the definition and play havoc with the allocation of risks that the parties intend. Indeed, reading the language in this way would mean that an event that (a) causes a material adverse effect on the target, and (b) does not itself fall into an exception, would nevertheless not count as a Material Adverse Effect, if the event (c) was “related” in any way whatsoever to any other event that did fall into an exception. That would make the set of excepted events extremely large. For example, in Akorn,[25] the emergence of new competitors had a material adverse effect on the target, and this event—the emergence of the new competitors—did not fall into an exception. Suppose that, while the new competitors were emerging, the economy slipped into recession. The recession would be a general change in economic conditions and so would fall within an exception covering such changes. Had the Fresenius-Akorn merger agreement used the “related” language rather than the “arising from” language, would the emergence of the new competitors, which caused a material adverse effect on Akorn, not have counted as a Material Adverse Effect because the emergence of the competitors occurred simultaneously with (and thus in a sense was “related to”) the recession, an excepted event, even though the recession had no causal connection either with the emergence of the new competitors or with the material adverse effect on the company? Reading the “related” language in this unrestricted manner implies that the emergence of the new competitors would no longer count as a Material Adverse Effect, but I cannot see how such an interpretation of the contract would serve any rational purpose. Sophisticated commercial parties would attribute no importance to the occurrence of events “related” to the event causing the material adverse effect if such events were not themselves causally related to either the material adverse effect or the event causing it.
Furthermore, if the phrase “related to” in the language introducing the exceptions meant “related to in any way whatsoever,” then there would almost certainly never be a Material Adverse Effect. For, whenever a non-excepted event caused a material adverse effect, we could always find some excepted event—some change in business, economic, market or industry conditions, some change in law or GAAP, some force majeure event—that was “related,” in some way or other, to the event causing the material adverse effect, which would mean that this event would not count as a Material Adverse Effect. Thus, when the language introducing the exceptions includes events “related” to excepted events, the meaning cannot be that the related events are related to the excepted events in just any way whatsoever, such as occurring simultaneously or being discussed in the same edition of The Wall Street Journal. The meaning has to be more restricted. The obvious way of restricting the meaning is to say that “related to” means “related to as effect to cause.” This interpretation has the virtue of making the resulting allocations of risk rational and consistent with the rest of the MAE definition, but it also makes “related to” equivalent to “arising from” and thus deprives the phrase “related to” of any independent meaning.
In KCake, the only Delaware case ever to consider the import of the language introducing the MAE exceptions, the court stated in dicta that the phrase “arising from or related to” “is broad in scope under Delaware law,” and the court clearly implied that “arising from” and “related to” have different meanings.[26] Beyond that, however, the court did further construe either phrase, and on the facts it was clear that the events “arising from or related to” the excepted events had arisen from the excepted events as effects from causes. The problem remains, then, whether the “related to” language can have some meaning beyond that of “arising from” but still not be so expansive as to amount to “related to in any way whatsoever.” One possibility would be to construe “related to” as meaning “correlated with.” In that case, an event would be “related to” an excepted event if the related event either arose from the excepted event or else both the related event and the excepted event arose from the same cause. This sounds promising, but the first possibility reduces to “arising from,” and the second possibility, which is doing all the work of giving “related to” some independent meaning, seems to produce anomalous results. That is, interpreted in this manner, the “related to” language would imply that there could be an event that (a) would reasonably be expected to have a material adverse effect on the target, and (b) does not itself fall into an exception, and yet (c) would not count as a Material Adverse Effect merely because it (d) arose from some event that also caused an event that did fall into an exception.
It is difficult even to construct a plausible example, but imagine that the President of the United States takes some highly controversial action, such as moving the American embassy in Israel from Tel Aviv to Jerusalem or declaring Turkish atrocities against the Armenians during World War I an act of genocide. As a result, the target company’s chief executive officer, who is deeply offended by the President’s decision, demands that the company cease doing business with the government of the United States, and when the board of directors refuses, he abruptly resigns. The resignation is not excepted under the MAE definition, and it results in a material adverse effect on the company. Meanwhile, the President’s decision also spurs various terrorist attacks against Americans around the world, but these events have no effect at all on the company. Terrorist attacks, however, are excepted events under the MAE definition. Now, the departure of the chief executive officer is both not excepted and reasonably expected to have a material adverse effect on the company. It would thus seem to be a Material Adverse Effect. Would anyone say that since the departure and the terrorist attacks, which had no effect on the company, arose from the same cause, and since the terrorist attacks were excepted, the departure should also be excepted because it was “related” to the terrorist attacks in the sense that both events arose from the same cause? It is hard to see why sophisticated commercial parties would attribute any importance to this fact. There would seem to be no economic rationale for treating the risk of the chief executive abruptly resigning merely because the cause prompting him to resign also resulted in terrorist attacks that had no effect on the company.
For such reasons, it is difficult to see what independent meaning, distinct from the meaning of “arising from,” can plausibly be ascribed to the phrase “related to” in the language introducing the exceptions in the MAE definition. Perhaps the best view is to construe “arising from or related to” as a legal doublet like “null and void” or “cease and desist” and limit the meaning to the causal interpretation.[27] If that is right, then it makes no difference whether the language introducing the exceptions excepts events “arising from” or “related to” events falling into such exceptions. Although transactional lawyers may sometimes argue about this issue as if it mattered,[28] the case may be analogous to distinctions between the various kinds of efforts clauses, about which transactional lawyers also argue about but among which the Delaware courts have been unable to meaningfully distinguish.[29]
[1] Professor of Law at the University of Iowa College of Law and Fellow and Program Affiliated Scholar at the Classical Liberal Institute at New York University Law School. This article is an expanded version of How Much Do Exceptions in MAE Definitions Except? 26.2 Deal Points 10 (Spring 2021), and it draws on material from Robert T. Miller, Pandemic Risk and the Interpretation of Exceptions in MAE Clauses, 46 J. Corp. L. __ (forthcoming, 2021).
[2] AB Stable VIII LLC v. Maps Hotels & Resorts One LLC, No. 2020-0310-JTL, 2020 WL 7024929 (Del. Ch. Nov. 30, 2020).
[3] Snow Phipps Group, LLC v. KCake Acquisition, Inc., No. 2020-0282-KSJM, 2021 WL 1714202 (Del. Ch. April 30, 2021).
[4] Fairstone Fin. Holdings Inc. v. Duo Bank of Can., 2020 CanLII 7397 (Can. Ont. Super. Ct. J.).
[5] AB Stable VIII LLC, No. 2020-0310, at *55-56.
[6] KCake Acquisition, Inc., No. 2020-0282, at *35.
[7] Fairstone Fin. Holdings Inc., 2020 CanLII 7397, at paras. 105-108.
[8] Robert T. Miller, Pandemic Risk and the Interpretation of Exceptions in MAE Clauses, 46 J. Corp. L. __ (forthcoming, 2021).
[9] Notice that, when we ask what is reasonably expected to follow from an event, we ask what is reasonably expected to follow from the actual event that occurs, not what is reasonably expected to follow from an “average” or “typical” event of the relevant kind. In the example in the text, what matters is what is reasonably expected to follow from the actual decision by the FOMC in the actual circumstances in which the decision was made. Thus, what is reasonably expected to follow from an increase in the target Federal Funds Rate of 25 basis points is not what is reasonably expected to follow from an increase in that rate of 500 basis points. Similarly, if one party bears the risk of a pandemic and a pandemic occurs, we ask what is reasonably expected to follow from the actual pandemic that has occurred, which will depend on the nature of the actual pandemic occurring. Some pandemics are much worse than others (compare the H1N1 pandemic of 2009 with the COVID-19 pandemic of 2020), and so what would reasonably be expected to follow from one particular pandemic might not reasonably be expected to follow from another particular pandemic. In asking what would reasonably expected to follow from an event, we are asking what would reasonably be expected to follow from the actual event that has occurred, in all its existential particularity. The reason for this is that the MAE definition expressly speaks of events and what is reasonably expected to follow from them, not of kinds or types of events or of “average” or “typical” events of certain kinds.
[10] The court ultimately held that the acquirer bore the risk of a pandemic because an exception in the MAE definition included the terms “natural disaster” and “calamity,” and the COVID-19 pandemic was both of these. AB Stable VIII LLC, No. 2020-0310, at *57-59. Hence, the court’s discussion of whether other exceptions in the MAE definition would have applied if the target had borne the risk of a pandemic is dicta.
[11] Id. at *55-57.
[12] Id. at *56. Court also says about root cause, but we can let that pass. “each exception applies on its face, not based on its relationship to any other exception or some other root cause
[13] Id. at *56-57.
[14] Robert T. Miller, Material Adverse Effect Clauses and the COVID-19 Pandemic (Univ. Iowa Legal Stud., Rsch. Paper No. 2020-21), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3603055 [https://perma.cc/NHA6-EXL7] [hereinafter Miller, Covid-19 Pandemic], at 25-26.
[15] AB Stable VIII LLC, No. 2020-0310, at *55 n. 204.
[16] AB Stable VIII LLC, No. 2020-0310, at *48 (emphasis added).
[17] AB Stable VIII LLC, No. 2020-0310, at *56.
[18] The court in KCake makes exactly the same mistake. In that case, the acquirer had pointedly refused to agree that “pandemics” would be excepted events, thus allocating the risk of a pandemic to the target, KCake Acquisition, Inc., No. 2020-0282, at *6-7, but the parties had agreed that general changes in the economy, id., and changes in law would be excepted, id. at *35, and thus that the risk of such changes would be allocated to the acquirer. The court held that the target had not suffered a material adverse effect, which suffices to dispose of the case, but it then went to say in dicta that that “revenue declines arising from or related to changes in law fall outside of the definition of an MAE, regardless of whether COVID-19 prompted those changes in the law.” Id. As in AB Stable, this confuses events causing material adverse effects with the material adverse effects they cause. Perhaps “changes in law” and “revenue declines” are excepted events and so not Material Adverse Effects, but that does not prevent COVID-19, which caused the revenue declines, from being a Material Adverse Effect, unless the declines are treated as a material adverse effect, and this effect is deemed excepted because the court was applying the exceptions to effects rather than events. As in AB Stable, if a material adverse effect arises from an excepted event, then no event causing that material adverse effect can be a Material Adverse Effect.
[19] Robert T. Miller, The Economics of Deal Risk: Allocating Risk Through MAC Clauses in Business Combination Agreements, 50 Wm. & Mary L. Rev. 2007 (2009).
[20] Id., at 2047.
[21] Guhan Subramanian & Caley Petrucci, Deals in the Time of a Pandemic, __ Columb. L. Rev. __ (2021), https://papers.ssrn.com/sol3/abstract_id=3799191.
[22] Id. at 50. The authors state that only 47% of the agreements in their sample use the causal “arising from” (or similar) language, while 53% use the non-causal “related to” (or similar) language. Id. It turns out, however, that the authors counted such expressions as “impact of,” “resulting directly or indirectly,” and “arising in connection with” as non-causal. Id. at 37 n. 211. These expressions seem plainly causal to me. Accordingly, I doubt whether Subramanian and Petrucci’s breakdown of 47% (causal language) and 53% (relational language) is correct.
[23] Id. at 50.
[24] Note that, to accomplish this goal, the phrase “arising from” in the language introducing the exceptions must be construed in the same way as the phrase “reasonably expected” in the base part of the definition. The “reasonably expected” language suggests the concept of proximate causation commonly used throughout the law. The “arising from” language certain can suggest the same concept, but it can also suggest the notion of but-for causation, which is also commonly used throughout the law. I think the two expressions should be read as having the same meaning, and that the meaning should be the proximate-causation meaning of the base part of the definition. That is, if the language in the base part of the definition speaks in terms of events “reasonably expected” to have a material adverse effect on the target, then any “arising from” language introducing the exceptions should be read as meaning “reasonably expected to arise from.” If the “arising from” language were read in terms of but-for causation, the results would be unpredictable and often irrational because the but-for effects of an event are multifarious, extremely far-reaching, and highly unpredictable. It is difficult to imagine that sophisticated commercial parties would allocate risks in such a haphazard manner.
[25] Akorn, Inc. v. Fresenius Kabi, AG, No. 2018-0300, 2018 WL 4719347 (Del. Ch. Oct. 1, 2018)
[26] KCake Acquisition, Inc., No. 2020-0282, at *35.
[27] This insightful point was suggested to me by Glenn West.
[28] Subramanian & Petrucci, supra note 24, at 53.
[29] Akorn, Inc., 2018 WL 4719347, at *8687; Williams Cos. v. Energy Transfer Equity, L.P., 159 A.3d 264, 272 (Del. 2017).
