A large body of research has been conducted into equity market efficiency, but the efficiency of bond markets is relatively unexplored.

The authors look at 32 different financial statement measures that have previously been demonstrated to affect equity returns. These measures are grouped into value/glamour, accruals, profitability, investments, financing, financial reporting quality, and miscellaneous. The authors investigate the efficiency of US corporate bond markets by looking at the effect that different anomalies within these groupings have on returns.

The authors also examine how key bond market factors are related to the anomalies—for example, how a higher proportion of institutional investors or uncertainty about the creditworthiness of issuers relates to market efficiency.

The authors examine a sample of US non-financial firms with publicly traded bonds, using bond returns from the two largest databases: Datastream and TRACE (the Trade Reporting and Compliance Engine). Merging these data with additional data from Compustat, they compile a sample of 460,259 bond-month observations from January 2001 to December 2011. They tabulate descriptive statistics for the bond returns and bond characteristics.

Next, they construct long, short, and long–short portfolios based on each of the 32 fundamental measures and determine which of these portfolios outperform a benchmark. To perform this calculation, they use a six-factor model on monthly returns. The six risk factors are market, size, value, momentum, default, and term structure.

Because of the practical difficulties entailed in shorting bonds, the authors also examine the efficacy of long-only strategies, both before and after transaction costs are considered. They use characteristic-adjusted returns, which may more accurately reflect actual portfolio performance relative to benchmarks.

Finally, they investigate whether particular categories of bonds are related to these return anomalies. They look at the prevalence of institutional investors in the market for a given bond by using two proxies: (1) whether there is a credit default swap (CDS) tied to the firm and (2) the proportion of trades larger than US$1 million. They also look at the amount of uncertainty around the creditworthiness of the issuer and the role of bond ratings.

In their analysis of long–short alphas, the authors find that 17 of the 32 financial statement measures are significantly related to future bond returns. In 15 of these cases, this relationship is positive. This is particularly conspicuous in the accruals category (where four out of five measures produce statistically significant, positive returns) and the investments category (four out of seven). The largest positive return of all comes from the size anomaly.

Using the more pertinent characteristic-adjusted returns, 17 anomalies produce statistically significant abnormal returns prior to transaction costs. Ten of these also return abnormal positive returns after transaction costs—including at least one measure in each of the seven categories.

In their analysis of key bond market factors, the authors find that institutional investors reduce inefficiency in the bond market. Just 6% of anomalies are significant when there are a high number of $1 million trades, compared with 28% otherwise. Moreover, anomalies are found to be more significant for bonds of firms without a CDS. The authors also find that higher uncertainty about the creditworthiness of the issuer increases market inefficiency and that anomalies are concentrated in non-investment-grade bonds.

The authors find evidence that corporate bond markets, like equity markets, contain inefficiencies related to financial statement measures. They find 10 anomalies that are associated with positive abnormal returns for long-only portfolios after transaction costs are accounted for. This finding suggests that bond investors using long-only strategies focusing on these measures—including the ratio of operating cash flow to price, working capital accruals, and net new investment accruals—could improve their returns.

The further association between uncertainty about the creditworthiness of an issuer and greater inefficiency suggests that investors should look for bonds with large earnings surprises and split ratings assigned by rating agencies. They could also consider bonds without a large institutional presence, though the authors note that these bonds, which are likely to be relatively illiquid, could incur substantial transaction costs.