Skip to content Skip to navigation

Research & Ideas

Search this site

How a plan designed to discourage reckless investing could backfire

A study finds that under certain conditions a new type of financial instrument could create perverse incentives for banks to pre-emptively file for bankruptcy.

CoCos could make our banking system safer — or much, much worse. | iStock/webking; iStock/Maica


During the recent financial crisis, when governments around the world used taxpayer money to bail out giant banks, financial regulators hit on a strategy that was intended to discourage reckless investing and avoid the need for taxpayers to save troubled financial institutions.

This strategy created new financial instruments called “contingent convertible bonds,” CoCos for short. If a bank’s capital reserves decline to a predetermined danger point, it triggers the CoCos to convert into a form that replenishes the balance sheet. One important type of CoCo converts into common stock, which makes former bond owners new shareholders.

The idea is for banks to sell CoCo bonds as a form of financial health insurance while they are still strong. Then, in a crisis, when a bank in need of additional capital would have a hard time finding new investors, many CoCo bonds would automatically convert into shares. Converting CoCo bondholders into new shareholders would wipe out debt, increase the bank’s capital, and improve its cash flow by reducing interest obligations.

The popularity of CoCos differs largely among different countries, mainly due to the different treatments by tax authorities and bank regulators. European and Asian banks represent the largest CoCo issuers. Deutsche Bank of Germany, Societe Generale of France, UBS and Credit Suisse have collectively issued billions of euros worth of CoCos in the past several years. And because European regulators count the convertible bonds as a form of “Tier 1” capital, even before they convert, banks have seen the instruments as a useful way to shore up their balance sheets. For instance, regulators in Switzerland have given banks the option of lower capital requirements if they include CoCos as a substantial part of their debt.

But a new study finds that, under certain conditions, CoCos could create perverse incentives for banks to pre-emptively file for bankruptcy. Markus Pelger, a coauthor of the study and an assistant professor of management science and engineering at Stanford, explains that if a highly leveraged bank issues CoCos on top of the existing debt, this can drastically increase its probability of default.

“CoCos require high coupon rates (yields),” says Pelger. “If a bank is generating low cash flows, the bank’s shareholders might decide that it is optimal to default before the CoCos convert.”

This could lead to a condition that the researchers called “debt-induced collapse,” creating exactly the opposite of what regulators intended when they came up with the CoCo strategy.

Pelger and his coauthors — Nan Chen at the Chinese University of Hong Kong, and Columbia University scholars Paul Glasserman and Behzad Nouri — say the key to avoiding debt-induced collapse involves the trigger mechanism that forces a bank to initiate a CoCo conversion.

The trigger point for most CoCos is now set at 5-7% of capital. That means when a bank’s capital ratio —cash on hand relative to outstanding debt — falls below that percentage, CoCos must convert into shares. This conversion has the effect of reducing the bank’s debt load and improving its capital ratio, but at the same time it adds new shareholders whose influx dilutes the per-share value of existing shares.

The solution, the researchers contend, is to set the trigger point for CoCo conversions sufficiently high. They provide guidance for an appropriate trigger level and show that in particular for “too big to fail” banks, e.g. Deutsche Bank, that level should be well above the current 5-7% trigger points.

The good news is that if CoCos are properly designed to avoid debt-induced collapse, they create incentives for the bank to reduce risk. “If you’re right before the trigger point for an automatic conversion, the bank’s shareholders may well want to do everything possible to avoid a conversion, which is costly for them as it dilutes their shares,’’ Pelger says.

Their study shows that properly designed CoCos could reduce a bank’s appetite for excessive risk-taking and create incentives for shareholders to invest into the firm to prevent conversion.

The researchers developed a general model for CoCos to study their properties and incentive effects. They also applied their models to detailed financial data on 19 of the biggest financial institutions operating in the United States — banks that the Federal Reserve subjected to its first “stress test” in 2009. The researchers found that, on balance, the banks in its study would have been better off if CoCos with a high conversion trigger had accounted for 10% of total debt before the financial crisis erupted in 2008.

Pelger concludes: “CoCos have the potential to make our banking system safer but can end in disaster if used in the wrong way.”

Get Updates from Stanford Engineering