Xavier Gine and Inessa Love, May 2009
The current financial crisis has pushed many firms to the brink of bankruptcy. A key policy question is thus whether bankruptcy laws are efficient, in the sense of allowing better firms to reorganize while liquidating unviable firms. The sixth in our impact series presents lessons from a reform in Colombia that achieved this objective.
Nearly ninety countries around the world have reformed their bankruptcy codes since World War II and over half of them have done so during the last decade. The extent to which these reforms will improve efficiency depends on their design and the context in which the new codes are binding. A key aspect of an efficient bankruptcy system is its ability to encourage the reorganization of viable firms and the liquidation of unviable ones. This requires a delicate balance. On the one hand, if the law is lenient towards failing firms, it will inevitably allow inefficient firms to continue operations. On the other, if the law favors liquidation, it will also liquidate viable firms.
While there is a growing literature estimating the costs of bankruptcy, there is little empirical evidence assessing how these costs affect the ability of the bankruptcy system to separate viable businesses from unviable ones, a key to ensuring efficiency. A recent evaluation of a bankruptcy reform that took place in Colombia provides some first such evidence.
Bankruptcy Reform in Colombia
In the midst of (a previous) financial crisis, facing a backlog of failing businesses entering a very inefficient bankruptcy process, Colombia adopted a new reorganization code in late 1999. This law, known as Law 550, streamlined the reorganization process by establishing shorter statutory deadlines for reorganization plans, reducing opportunities for excessive appeals by debtors (who typically delayed the process in order to suspend the debt service) and requiring mandatory liquidation in cases of failed negotiations.
The Colombian reform can be seen as a natural experiment with two regimes: a pre-reform regime with high reorganization costs and the post-reform regime with lower costs. In a regime with high costs, some viable businesses may be liquidated if the costs they face are too high. In this case, the bankruptcy system fails to separate viable from unviable firms, resulting in inefficient outcomes. In contrast, when reorganization costs are low, better quality firms are more likely to choose reorganization, resulting in a clear separation between firms that reorganize and those that liquidate. In this regime, as a result of both lower costs and better selection, the recovery after reorganization is faster.
This study uses a natural experiment (i.e. introduction of the new law) to evaluate its impact on different classes of firms. We exploit the fact that the reform only changed the reorganization proceeding, leaving the liquidation process unchanged. This allows us to compare the pool of firms that file for reorganization to the pool of firms filing for liquidation before and after the law (i.e. a difference in difference methodology). In addition, we control for macroeconomic conditions by comparing firms filing for bankruptcy with a sample of matching active firms (matched by industry, size and year of filing).
The new law increased the efficiency of the bankruptcy system in Colombia.
We find that:
1. The duration of reorganization proceedings significantly decreased in practice after the introduction of the law in accordance with its text.
2. After the reform viable firms were more likely to reorganize and unviable firms to be liquidated. Under the old law, firms filing for reorganization were indistinguishable from those filing for liquidation in terms of several measures of financial health. In contrast, relatively weaker (and hence less viable) firms were more likely to file for liquidation after the new law, resulting in a clear separation in the distribution of the two types of firms.
3. A firm’s recovery after reorganization was significantly improved under the new law. While under the old law firms hovered at about the same low level of financial health for years after entering into reorganization, we observed a clear recovery under the new law.
1. Effective Bankruptcy procedure should seek to minimize reorganization costs. High reorganization costs in the form of lengthy procedures or biases in the law undermine credit recovery and prevent viable firms to be reorganized thereby leading to inefficient liquidations.
2. A financial crisis provides a good opportunity for serious bankruptcy reform. A high number of bankruptcies can put the issue high on the policy agenda, and generate the political will to pass the required legislation.
3. An effective bankruptcy process has to ensure that viable firms are reorganized and unviable ones are liquidated. Inefficient firms need to be liquidated and the resources reallocated to more efficient ones. However, liquidations and fire-sale of assets are costly due to market illiquidity and transaction costs. Therefore, firms that are potentially viable are better off being reorganized.
4. An effective bankruptcy process should support speedy recovery of reorganized firms. Once a firm emerges from reorganization, it should recover quickly and become profitable, which will support most effective utilization of resources. Prolonged distress and insufficient debt relief is likely to lead to repeated bankruptcy episodes, and ultimately to failure of the firm.
5. An effective bankruptcy process is important for financial sector development. Financial transactions are impossible without assurance that creditors will get their money back (with a return). An effective bankruptcy system is important in returning funds to creditors when the enterprise fails, or allowing for rescheduling or reducing debt levels to allow a firm to recover and ultimately maximize the return of capital to the creditors.
For further reading see:
Gine, X and I. Love “Do Reorganization Costs Matter for Efficiency? Evidence from a Bankruptcy Reform in Colombia”, World Bank Policy Research Working Paper 3970, revised May 2009.