Section 1: Co-ordination Failure as a Prisoners’ Dilemma
Mohamed El-Erian of PIMCO, writing in the FT (‘Why the Greek rescue isn’t going to plan’, April 7), spoke of Greece as a sovereign risk problem that is not going as planned. El-Erian indicated that it was because of what he referred as a co-ordination failure. He wrote
It is a classic co-ordination failure in game theory. Any first mover will become worse off. Indeed, it is in the interest of any single party to wait for others to move first. As a result, no meaningful progress is made, the problems fester, and the risks of a disorderly outcome increase.
This sounded exactly like the Prisoners’ Dilemma (PD) problem in classical game theory that shows why two economic actors may not co-operate despite it being in their best interests to do so. The dilemma is as follows:
Two crime suspects are taken in by the police. The police have insufficient evidence for a conviction. They separate both the prisoners, visit each of them and offer the same deal. If Prisoner A testifies (or defects against the other Prisoner) for prosecution against Prisoner B and the latter remains silent (and hence co-operates with Prisoner A), A goes free and B receives the full sentence of twenty years. If both remain silent, both prisoners are sentenced to a diluted sentence of five year as the more serious charge cannot be proven. If each betrays the other, each receives a ten year sentence. Each prisoner must either betray the other and testify against his partner, or remain silent and thus cooperate with his partner. Each one is assured that the other would not know about the betrayal or cooperation before the end of the investigation.
If such a game is played once and only once, defecting against the other prisoner is the rational outcome over co-operating. Irrespective of the actions of the Prisoner B, Prisoner A will always gain a greater payoff by choosing to defect. Similarly, Prisoner B will always gain a greater payoff by choosing to defect regardless of Prisoner A’s actions.
If the game is played repetitiously with each player (both of whom are aware as to how many times the game will be played and both of whom are in possession of the opportunity to punish the other player for previous non-co-operative play) it turns out that the rational outcome for the two players continues to be to defect again and again, irrespective of how many times the game is repeated.
Only when the players play the game an indefinite or random number of times, is it possible to enforce cooperation as a rational equilibrium outcome. Only in such a case, the incentive to defect can be overcome by the threat of punishment.
Section 2: Greece and Germany as Players in the Game
Greece and EU are playing such a one-shot (unrepeated) PD game.
Greece’s crime is that is did not heed the EU directives and put its fiscal situation in order in line with the Maastricht Criteria. EU’s crime is that it did not enforce the Maastricht Criteria and temporarily suspend Greece as punishment when it failed to follow the guidelines on fiscal debt.
Now that both the EU and Greece have been ‘caught’ by the Sovereign/Euro crisis, each is figuring out its best response. For Greece the optimal outcome, in the context of the prisoners’ dilemma, is to defect, while EU co-operates. Defection would imply that Greece does not rectify its finances and is instead bailed out by the EU. The optimal outcome for the EU in such a game is to force Greece to rectify its finances – Greece cooperates. Thus EU does not bail Greece out –in the context of game, EU defects.
If both the EU and Greece defect as is the optimal rational outcome in the game, both lose out somewhat. Specifically, Greece is likely suspended from the Euro region until it fixes its finances and the EU suffers a bout of credibility as markets begin to question the future of the Euro.
An IMF-led bailout in this context would be akin to both parties pleading guilty to partial charges. This again is not the rational equilibrium outcome and undermines the situation as a non-European solution. It would be somewhat of a co-ordinated outcome.
It is optimal for both Greece and EU to cooperate with each other and put forth a credible plan internally to resolve both the fiscal deficit situation of Greece (and to some extent rest of the imbalances amongst the PIIGS), and – as suggested by Martin Wolf – to stimulate aggregate demand in countries running surpluses (like Germany) for goods from countries with deficits (like Greece). This would prevent Greece from wading deeper into recession, and help preserve the Euro region in its present shape and form. However such an outcome is difficult to achieve as in the context of the game it is not a self-enforcing, rational outcome. Consequently, solutions suggested by Germany hover around a zero-sum world of beggar-my-neighbor policies in which a country attempts to play to its benefit alone by increasing market share from the rest of the region.
Expertise of Trichet notwithstanding, a currency union without a fiscal and political union is difficult to succeed as it distorts incentives for members to part take in the benefits of a monetary union without necessarily incurring all the costs such a union imposes.
Fig 1: Greek over German Spread at historic highs is making Greek Bonds as expensive to issue as Junk Bonds. Source: Bloomberg.
Section 3: A Similar Story
Greece’s actions are no different fundamentally than the actions of banks going into 2007. Banks took on excessively risky exposures, without allocating appropriate levels of capital, to increase potential profits as they implicitly, and correctly, assessed that, on failure, there was a reasonable likelihood of a potential bail out. Thus, they enjoyed the benefits of being the key actors in a subsidized payments system.
Similarly, Greece has enjoyed the benefits of the single currency without having picked up the costs that come with such a union, specifically in terms of strapping its expenses and putting its fiscal house in order.
Worse still, the situation is not getting better. On Saturday, Markit reported that CDS protection on Greece had become more costly than that on Iceland. Earlier, Iceland had to seek a $4.6 billion IMF-led bailout after its three biggest banks collapsed in October 2008.
Ironically despite their demonization, CDS markets have been helping Greek to continue to issue primary debt as CDS trading is providing an avenue to hedge Greek exposure. This has eased hedging concerns in the primary issuance market allowing it to work, albeit at costs of issuance similar to that of junk bonds. Such a prohibitive cost undermines Greek government’s ability to continue to fund itself in the coming months.
Figure 2: Cost of insurance for Greece touched that of Iceland, a country that was bailed out by the IMF. Source: Markit’s CDS Report, Courtesy Gavan Nolan
Figure 3: Cost of insurance for Western European Sovereigns is rising once again indicating possible spreading of concerns. Source: Markit’s CDS Report, Courtesy Gavan Nolan (Click to enlarge)
El-Erian ended his FT article by emphasizing
…Greece is part of a wider, and historically unfamiliar phenomenon – that of a simultaneous and large disruption to the balance sheet of many industrial countries. Tighten your seat belts.
This brings to mind the remarks of CLSA analyst Chris Wood on CNBC (March 01, 2010) that
….Credit Crisis we saw in the West in 2008 and 2009 has simply been deferred, because 95% of the so-called government policy solutions to deal with this Crisis have simply been to extend government guarantees. So the problem has been transferred from the private sector to the public sector. It is just a matter of time before investors’ revolt against these sovereign guarantees. The crisis is going to happen first in Europe. The U.S. will be the endgame.
Two astute market observers may not be completely right but one may ignore them at one’s own peril.
A client comment on Greece – European Sovereign Debt Crisis. The note in pdf is available for Seeking Alpha readers here