Failing the test of truth-telling
Latin American social philosopher George Santayana said that those who forget history are doomed to repeat it.
Why didn't the International Monetary Fund see this global crisis coming? This was the question posed by the Independent Evaluation Office of the IMF in its report published last month.
Some readers will remember that the office was established after the Asian financial crisis to evaluate independently where the IMF had gone wrong during that crisis. To the IMF's credit, it understood that it had to be transparent to be credible and the evaluators' specific mission is to provide independent feedback to the IMF board on the fund's effectiveness.
This time, the evaluation office focused not on the fund's handling of the crisis, but on its surveillance performance from 2004 to 2007.
The office should be commended on its methodical analysis. Since the IMF did not warn its members of the unfolding crisis, the study concentrated instead on whether the fund evaluated the proper risks and vulnerabilities before the crisis, what message the fund gave to its board and members, and what constraints the IMF had in conveying the difficult messages.
The facts are that, even up to April 2007, the IMF's main message was continued optimism. Even though it warned of the global imbalance, there was insufficient focus on the balance sheet fragilities, so it missed the key elements that underlay the crisis. For example, in the annual review of the United States, it did not discuss the mortgage problems until the crisis had already surfaced. In a 2006 study of Iceland, IMF staff concluded that 'hedging behaviour and generally sound balance sheets and asset-liability management made the financial system relatively robust to the recent shocks'. In the next year, the largest Irish banks failed.
This was like a doctor giving a good health report to a patient before he suffered a heart attack.
One oft-repeated criticism of the fund by its smaller members is the uneven treatment of members, since it could be quite tough on them, but less tough on members with more shareholdings. This was particularly true in the mandatory annual review of the US, its largest shareholder. The fund did not conduct a 'financial sector assessment programme' for the US before 2009 because its authorities did not agree to one.
The report itself doubted whether, even if such a review had been done, the current methodology would have identified the scale of the risk. For example, the 2006 assessment update for Ireland had concluded that the 'outlook for the financial system is positive'.
The annual country reports - known as 'Article IV surveillance reports' - on the US and Britain, the world's main financial centres, were sanguine on financial innovation and behind the curve on risks. Overall, IMF staff had felt that these authorities were on top of their risks. In the 2006 report, the credit rating agencies were said to be 'uniquely positioned to assess a wide range of structured transactions'. True, they were uniquely positioned, but they told the world that these toxic products were triple-A-rated.
In most countries, the person who certifies that an unsafe drug is safe goes to jail. Yet, no rating agency has been punished.
The next question is: why did the fund fail to give clear warnings? The funny thing is that Raghuram Rajan, the IMF's then economic counsellor, had warned in 2005 that leveraged financial innovation would leave countries more exposed than in the past. His warnings to leading central bankers fell on deaf ears and his work was not followed up even within the IMF's own programme.
It goes to show that if the crowd's mindset goes in one direction, a lone voice will not necessarily stop the tide. Sometimes, it takes a crisis to force a change of mindset.
The fund's internal failings were identified as analytical weaknesses, particularly cognitive biases such as group thinking, intellectual capture and confirmation bias. If the group thinks that financial innovation is good, it becomes very difficult for lone dissenters to say otherwise. Intellectual capture refers to a situation in which students bow to the intellectual power of a teacher. This was the position of IMF staff who were overly influenced by, and perhaps in awe of, the largest members' reputation and expertise.
Confirmation bias refers to selective attention - people focusing on what they expect and ignoring information that is inconsistent with their theories. If conventional wisdom says global imbalance is the problem, then other areas, such as balance sheet vulnerabilities, may be ignored.
The IMF economists relied heavily on their macroeconomic models, but these did not incorporate balance sheet data and therefore underestimated macro-financial linkages. The report also pointed out that stress tests do not reveal second-round effects, such as liquidity shocks.
The report told of organisational impediment in terms of departmental silos, where staff did not share information and did not 'connect the dots'. And, the incentives failed to 'foster the candid exchange of ideas that is needed for good surveillance'.
Finally, to what extent was the fund subject to political constraints? The report quoted staff as saying that there were limits to how critical they could be regarding the policies of the largest shareholders - that 'you cannot speak the truth to the authorities' since 'you are owned by these governments'.
Every wise government must have an independent voice to give 'ruthless truth-telling', one of the key roles for the fund according to one of its intellectual founders, John Maynard Keynes.
The Independent Evaluation Office has set an example of what should be done for the fund. Perhaps its member countries can learn from this.
Andrew Sheng is author of the book From Asian to Global Financial Crisis