At the time of the authors’ study, insurance companies held more than $2.3 trillion of corporate and foreign debt—more than mutual funds and pension plans combined. Because the insurance industry is thus the largest investor in the US corporate bond market, this bias has implications for the broader economy. The authors provide evidence that insurance companies with the highest propensity to reach for yield experience bigger losses of equity in economic downturns and that periods of reaching for yield by insurance companies lead to a higher issuance of risky assets by corporations.

Insurance companies’ allocation to riskier debt increases the risk profile of the underlying insurer, although these allocations do not represent a breach of regulatory constraints. Specifically, the lagged nature of credit ratings creates incentives for companies to circumvent the very controls that are intended to mitigate credit risk. During the global financial crisis, the insurance companies with the highest propensity to reach for yield saw their equity values fall the most. The authors demonstrate a strong negative correlation between stock performance and the amount of reaching behavior.

Because of insurance companies’ significant role in the corporate bond market, the authors explore whether reaching for yield generates a shift in the supply of credit to individual firms. They find that high-yield issuers within the BBB rating category issue more than low-yield issuers in the same rating category when insurance companies reach for yield.

The findings are robust to controlling for various bond and issuer characteristics.

To test the reaching-for-yield hypothesis, the authors examine the promised yield spreads on bonds and CDS conditional on regulatory risk constraints. That is, the authors measure how often insurance companies overweight the highest-yielding securities allowable.

Regulations restrict insurance companies’ ability to invest in risky securities by imposing punitive charges that escalate based on the perceived riskiness of the security (delineated by assigned credit ratings). These capital requirements often lead insurers to have a target distribution of their corporate bond portfolio across the various risk categories with a primary allocation to the highest-rated cohort (AAA to A rated securities). It is within this highest-rated cohort that the authors test the reaching-for-yield hypothesis. They find that although insurance companies hold 72% of all bonds in the top-rated cohort, 88% of these bonds are concentrated in the highest-yielding securities. Interestingly, the authors find no evidence of portfolio management skill with bond performance, largely owing to substantial systematic risk; bonds with higher insurance company holdings are more likely to be downgraded.

The time period reviewed is from the beginning of 2004 through 30 September 2010, and the sample is composed of more than four million individual corporate bond holdings. The dollar value of sample holdings is $336 billion in Q1 2004, rising to $472 billion in Q4 2010, with a reported face value of $1 trillion–$1.4 trillion. The authors use multiple data sources to collect holdings and ratings information.

The authors convincingly demonstrate that insurance companies have an inherent bias toward higher-yielding securities, irrespective of future downgrade prospects. The findings are both interesting and important. As the largest purchaser of corporate bonds, insurance companies have an integral role in fixed-income markets as well as in the broader economy. And although the period covered corresponds to the greatest (that is, worst) financial crisis in generations, which severely affected many insurance companies, the fact that the reaching-for-yield behavior returned after the crisis demonstrates a lack of understanding of the effect these decisions have on the economy. Investors should certainly pay attention to these large investors.