SINGAPORE For those who see North Korean leader Kim Jong-il as a dangerous lunatic prepared to risk the annihilation of his regime by launching a devastating attack on his neighbors, there is no shortage of supporting evidence.
He uses platform shoes and bouffant hair to appear taller. His official media churns out hagiographic tales of his almost superhuman talents -- including the world's greatest round of golf, 38 under par with 5 holes in one. And his stance in nuclear negotiations frequently appears contradictory and confounding.
But for analysts and policymakers trying to gauge the chance of a catastrophic war, game theory offers a crucial insight. In Kim's position, it is perfectly sane to seem mad. And it would be a disaster for him if the world believed he was rational.
This has fundamental implications for risk analysis. Above all, it means investors should avoid overreacting to signs of rising tension. The doomsday scenario of full-scale war on the Korean peninsula is an extremely remote possibility.
"He is frequently described as irrational, eccentric and dangerous. Most assuredly he is the last of those, but irrational and lightweight -- I think not," said Bruce Bueno de Mesquita, who specializes in political forecasting based on game theory, and who advised the U.S. Defense Department on Korea in 2004.
"No fool stays in power for years ... when there are so many generals, sons and wives waiting in the wings to launch a coup."
For years, Kim has veered unpredictably between belligerence and conciliation. In May he stunned the world with a nuclear test after earlier walking out of disarmament talks, vowing never to return. This month he made a conditional offer to restart talks, before wrong footing observers again with short-range missile tests.
Viewed in terms of game theory, this behavior is entirely rational. In fact, it is Kim's only viable strategy.
THE MADMAN THEORY
For game theorists like Bueno de Mesquita, the overriding objective of political leaders is the same -- survival.
Viewed in this context, Kim's actions should be interpreted as designed to maintain his internal grip on power and protect his regime against any foreign efforts to overthrow it. This is why he has put military strength above even feeding his people.
A second insight from game theory is that the decisions of players are shaped by their beliefs about how other players will respond. Each side tries to anticipate the actions of the other.
Despite the world's highest military spending as a percentage of GDP, North Korea is at a fundamental disadvantage -- South Korea is backed by the might of the United States. Kim could inflict devastating damage on South Korea in any war, but the result would be massive retaliation that finished his regime.
So it would be irrational for Kim ever to launch a major attack -- the inevitable result would be the end of his regime. The United States and its allies know this. The only way Kim can leverage his weaker military power is by making his opponents believe he may be crazy enough to launch a war anyway.
This strategy not only allows Kim to deter foreign attacks, but also gets him the maximum concessions in talks to give up the nuclear bomb. Because the other negotiating nations believe the threat Kim would use nuclear weapons cannot be ruled out, they are prepared to offer much more to remove this risk.
By signaling that he is irrational and unpredictable, Kim can turn a weak set of cards into a winning hand.
Washington and its allies face the opposite problem -- they are unable to make Kim believe their threats are credible.
Despite the frequent U.S. insistence that a nuclear North Korea is unacceptable, Pyongyang has never faced significant consequences for breaking promises and making threats, beyond economic sanctions which have only limited effectiveness.
Kim would not be the first world leader to adopt a strategy of calculated irrationality. It has been used repeatedly in history, not least by former U.S. President Richard Nixon.
Nixon tried to convince the Soviet Union, North Vietnam and other enemy states that he was unbalanced enough to launch a nuclear attack at the slightest provocation. "I call it the Madman Theory, Bob," he told his Chief of Staff H.R. Haldeman.
Nobel economics laureate Thomas Schelling explained "the rationality of irrationality" in his seminal 1960 work on game theory, The Strategy of Conflict. And as far back as the 16th Century, Italian political theorist Niccolo Machiavelli wrote that it can be "a very wise thing to simulate madness."
It is possible Kim would risk a suicidal war he cannot win. But a far more likely scenario is that he is bluffing -- or, in game theory terms, strategically leveraging a weak bargaining position by faking irrationality to make his threats credible.
Such a strategy would entail sudden policy changes, frequent brinkmanship and grandiose threats, and a willingness to initiate isolated military skirmishes, while preventing conflict escalating. This is exactly how Pyongyang has behaved for years.
In this interpretation of North Korean behavior, investors and markets should not take North Korean threats and provocations seriously. Periodic escalations of tension should be regarded as distractions which don't change the underlying risk calculus. Any panicky market sell-off should be seen as buying opportunities.
A second implication is that while Kim remains leader of North Korea, the country will remain a constant source of regional scares. Kim's strategy has repeatedly won him concessions, and also helps him maintain his internal grip on power. There is no reason for him to throw away his best cards.
(Editing by Dean Yates)