Predatory pricing

27 October 2008

1 Predation and entry deterrence

• Broadly defined, entry deterrence and predation are business practices adopted
by a firm, usually the industry’s incumbent, to reduce the level of expected
profits that the firm’s rivals, actual and potential, could hope to earn. If suc-
cessfully pursued, these policies would force the exit of the firm’s rivals, or at
the very least, cause the rivals to cut back output. When this happens, the firm
may earn higher profits in return for pursuing a predatory pricing policy.

• With the primary goal of lessening the competition the firm may face in the
long run, such business practices are often called anti-competitive conducts.

• Examples of entry deterrence and predation strategies include low prices, output
expansions, the introduction of new products, the redesigns of existing products,
promotions, capacity expansions, foreclosing input sources and exclusionary
contracts.

• Such practices, though can serve many other purposes, may be adopted to
convince the rivals that staying in the industry will not be profitable, and that
they are better off exiting the industry.

• In practice and in theory as well, it is however often difficult to distinguish these

strategies from normal market rivalry that economists salute for promoting the
common good. Take for example the introduction of new products aimed at
driving out rivals, actual and potential. There will be less competition as a
result. Consumers may nevertheless benefit now that they are able to choose
from a greater variety of higher quality products. On the other hand, this
will be at the expense of having had to pay higher prices. The firm, if only
confronted with lesser competition, will raise prices above the marginal cost of
production to a larger extent. Whether consumers in fact are better off can
only be determined via a careful cost-benefit analysis on a case by case basis.

1
• From an efficiency viewpoint, whether entry deterrence and predatory strategies
are beneficial or otherwise would depend on the very specifics of the market
environment. No simple, blanket conclusion is available that would apply as a
general rule.

• Still there are situations where the conclusion is less ambiguous. The canonical
example is that of predatory pricing defined as the incumbent charging a low
price for some time in order to bankrupt the entrants so that the incumbent
may be able to charge the monopoly price afterward. There may be a price
war for a short period of time. In the long run, consumers are surely worse
off, absent the price competition the entrants could have helped instill in the
market.

• The U.S., the E.U., and Japan have enacted legislation against such anti-
competitive behaviors. Firms that are found to have engaged in such anti-
competitive practices are liable to prosecution by the relevant government agency.

• As in almost any debates in economics, the need for active public policy against
anti-competitive behaviors is not entirely without controversy. The ultracon-
servatives have argued passionately that low prices and many other apparently
predatory business practices are part and parcel of normal and benevolent mar-
ket competition. Punishing and prohibiting businesses for competing in prices
and perhaps through other channels serve no useful purposes while inhibit the
competitive process. According to the more severe criticisms against anti-
competitive legislation, public policies and regulations against anti-competitive
behaviors have the only effect of helping the inefficient firms to survive the
competitive pressures coming from the truly innovative entrepreneurs, slowing
down economic progress and making humanity poorer as a result.

• Needless to say, any such sweeping conclusions can only be irresponsible. In-
deed, many such claims are mere rhetoric that are positively harmful to the
cause of rational discourse. Surely less competition hurts consumers and re-
duces the efficiency of resource allocation. If a certain industry incumbent suc-
ceeds in driving out the entrants and discouraging future entries in the process,
prices will only be higher in the long run. Worse, the incumbent’s urgency to
innovate is no more.

• It is however much harder to dismiss the more sensible concern that successful
predation can be very rare in practice and indeed may not be possible at all.
Relatedly, it is far from trivial to distinguish true predation from normal compe-
tition in prices and product qualities. The possibility to confuse vigorous price
competition from predatory pricing aimed at driving out rivals is very real.
One particular damning criticism is that predatory pricing in particular and

2
 predation and entry deterrence in general are seldom feasible if profitable busi-
ness strategies at all. If this is true, punishing low prices is punishing vigorous
competition.

• Our major focus in the following is to examine the environment in which pre-
dation is possible and profitable for the industry’s incumbent to carry out.

2 Predatory pricing
• More formally, we may define predatory pricing as the industry’s incumbent
waging a price war, cutting the price below the marginal cost of production, in
order to bankrupt the rivals and to force them out of the industry. If successfully
pursued, the incumbent’s monopoly reign is protected, and it will be able to
continue charging the monopoly price, earning a higher return on investment
as a result.

• There may certainly be instances of predatory pricing. The more relevant ques-
tion is really: How widespread is it? If predatory pricing is both feasible and
profitable in a variety of different environments, we would expect it to be quite
prevalent. And the need for vigilant policies against anti-competitive behaviors
in general and predatory pricing in particular cannot be downplayed.

• The same low prices that hurt the rivals would also hurt the incumbent too and
probably to a much larger extent. For the rival can always choose to scale back
production during the price war to minimize loss, whereas the incumbent would
have to produce at a high level of output to sustain the low price. In fact, in
the absent of any sunk cost, it is not clear predatory pricing can ever work at
all. During the war price, the entrant can simply shut down production and
redeploy the fixed assets elsewhere. In doing so, the entrant will never incur
any losses. The incumbent, in waging a price war, hurts no one but itself. If
the incumbent ever tries to raise the price above the average and marginal cost
of production, the entrant can move the assets back into the industry to take
advantage of the high price.

• Industries that have no sunk costs are termed perfectly contestable by William
Baumol of New York University. Predatory pricing is always a losing strategy
in perfectly contestable market. For it to work, the entrant must not be able
to avoid any losses altogether in shutting down production.

• There may be certain industries that come quite close to the ideal of perfect
contestability. But most industries in the real world are probably not perfectly
contestable. There are surely assets that may be redeployed in other industries
relatively easily. Examples include PC, office furniture, and transportation

3
 equipment. But then there are many other assets that have limited uses outside
the industry to various extents. A useful rule of thumb is that predatory pricing
is more likely to take place in industries whose production has a larger sunk
cost.

3 Deep pocket theory of predatory pricing

• Suppose production in the given industry is in fact associated with a large
sunk cost. Will the industry’s incumbent necessarily find it profitable to adopt
predatory pricing, and hence a clear necessity to restrain the firm’s behaviors?

• Since the incumbent itself would also suffer losses during the period of low
prices, the firm must be able to sustain the losses longer than the entrant can
do just that for predatory pricing to work. If the incumbent and the entrant are
identical in every respect, predatory pricing can not be a profitable strategy.
Perhaps in this case, it is equally plausible for the entrant to price predate the
incumbent.1

• Only when the incumbent is better financially endowed that predatory pricing
is likely to be a profitable strategy. The better financing could be due to the
firm having greater retained earnings and better access to bank credit and the
stock market. Sometime this is called the long purse or deep pocket model of
predatory pricing.

• Specifically, suppose the entrant will be losing some F dollars each period in the
price war started by the predatory industry incumbent, and its total financial
resource available to sustain the loss is equal to some B dollars. Then it may
last no longer than
B
n≤
F
periods in the price war.

• If the incumbent is operated by a much wealthier person, or maybe someone

who has much better access to bank credit and the stock market, then even
though it may lose at least as much as F dollars each period and most likely
more, it may well outlast the entrant in the price war. This makes predatory
pricing feasible.
1
It may be profitable for the incumbent to price predate the entrant if it has a lower average
and marginal cost of production. In particular, suppose the incumbent has constant marginal cost
equal to cI < cE the constant marginal cost of the entrant. The incumbent may thus set the price a
shade below cE and still be able to earn a positive profit. This, however, should not be considered
as predatory pricing as such but is more like normal price competition.

4
• Predatory pricing will be profitable as well if the difference in the monopoly
profit and the duopoly profit, the returns on adopting the predatory pricing,
exceeds the loss sustained during the price war. Consumers may gain in the
short run but will have to suffer the higher monopoly price in the long run.

• What kind of entrants would most likely be predated on? The sooner the
entrant can be bankrupted, the more likely predatory pricing will be deemed
profitable. Hence, predatory pricing should be used against financially weaker
entrants more often than entrants that are better financially endowed.

• According to J.P. Benoit of New York University, we will never actually observe
predation to take place in a world of no uncertainty. If the entrant will surely
lose in the price war and eventually forced to exit, it may as well save the losses
to be incurred in the process and exit at the first sign the incumbent is to wage
a predatory price war.

• To begin, assume that both firms have limited financial backings, but the in-
cumbent can survive in the price war longer before being forced into bankruptcy.
Predatory pricing is deemed feasible as a result. Whether it is profitable would
depend on the cost and the benefit for the incumbent in waging a price war.

• Let π m and πd be the monopoly profit and the duopoly profit the incumbent
may earn respectively, and C the cost to be incurred by the incumbent per
period in the price war. Assume that the following holds
³ ´ ³ ´ ³ ´
π m − π d δ + π m − π d δ 2 + πm − π d δ 3 ...
³ ´
πm − πd δ
= > C. (1)
1−δ

This has the interpretation that if the price war lasts merely one period, it
would be profitable for the incumbent to price predate.

• For the sake of making the story more interesting, let’s also assume that
³ ´
πm − πd δ2
< (1 + δ) C. (2)
1−δ

The left side is the PV of the benefits for the incumbent by a successful price
war that lasts for two periods, whilst the right side is the cost of doing so.
The inequality says that it should not be profitable for the incumbent to price
predate if the entrant chooses to fight for at least two periods.

• If the entrant can be made bankrupted in just one period, by (1) it would be
profitable for the incumbent to price predate. The more interesting case is when

5
 the entrant can fight for two periods. The inequality in (2) seems to suggest
that predatory pricing cannot be profitable for the incumbent and therefore
should not take place. However, according to a backward induction argument,
predation can be profitable even in such circumstances.

• To see why, suppose the price war has in fact gone on for one period. And
we are at the beginning of the second period. Should the incumbent fight the
entry? By then the entrant can only last one period longer. By the inequality
in (1) it would be profitable for the incumbent to price predate. The entrant,
knowing that it will be fighting a losing battle, should choose to retreat; i.e.,
exit the industry altogether to minimize loss.

• Next we back up to the previous period. The incumbent knows that if it fights
the entry by waging a price war in this period, the entrant will at the end of the
period have sufficient financing to fight for one period longer. In this scenario,
at the beginning of the next period, the incumbent will choose to threaten a
price war. The entrant’s best response is to exit right away. The entrant then
knows that fighting the price war for survival in this period unfortunately can
only be a losing battle. Hence, if the incumbent chooses to fight in the second
to the last period, it will similarly be best response for the entrant to exit the
industry immediately to avoid incurring any expense in fighting a pointless price
war in the given period. It is thus optimal for the forward—looking incumbent
to threaten to wage a price war in this period to force the exit of the entrant.

• This argument suggests that it will be optimal for the incumbent to threaten
to fight every period and best response for the entrant to exit at the first sign
that the incumbent will initiate predatory pricing.

• This is a remarkable proposition. Predatory pricing can be profitable for the

incumbent even in apparently very marginal circumstances. The incumbent will
find it profitable to price predate so long as the gain from predation exceeds
the cost of waging a one period price war, whereas the entrant may in fact be
able to survive a price war of many periods.

• The practical implication is that predatory pricing is easily a profitable practice

and therefore could be quite common. The best use of the model is perhaps
as a counterargument that predation should not happen often in reality since
many better financially endowed entrants may be able to survive a price war
for quite some time, whereas predation is typically only profitable if the price
war is over in a short period of time.