هزینه های مقایسه ای مذاکره در مقابل فروش اوراق قرضه رقابتی: شواهد جدیدی از وضعیت اوراق قرضه تعهد عمومی

Previous work on the interest cost implications of the municipal bond underwriting method of sale decision has produced results asserting that competitive bid sales result in lower interest costs. However, negotiated offerings still dominate the municipal market. This paper invokes financial certification theory to explain the apparent paradox. After correcting for selection bias which is predicted by the theoretical model, empirical estimates show that for a sample of state general obligation bonds, negotiated offerings have at worst no higher and perhaps lower interest costs.

The prevailing wisdom regarding the choice of underwriting method (competitive sale vs. negotiated offering) by municipal issuers has for years recommended competitive sales over negotiated offerings. One recent article summarized the prevailing wisdom: “where municipalities have a choice, they should sell their new bond issues by competitive means” (Simonsen & Robbins, 1996). Most of the empirical work in this area is significantly dated, however. After a period of initial interest in this issue, research has declined to just a handful of articles in the last 10 years. And the municipal bond market is still dominated by negotiated issues (62% of all sales in 1994 were done negotiated (The Bond Buyer, 1987–1999)). There are advantages and disadvantages with both types of underwriting method. Competitive issues may provide the appearance of a fair and impartial decision regarding which underwriter to use. However, since syndicates bidding on issues may not know if they will be awarded the bonds, there will be less incentive to conduct an exhaustive “search” for the highest level of market demand prior to the sale date. If there is more uncertainty regarding the credit quality of an issuer, it may be difficult to fully subscribe an issue. Also, the issuer has less flexibility to change the timing of a sale once the sale date is announced. If market conditions change adversely, the issuer may not be able to delay an issue. Negotiated underwriters may undertake a higher level of market search prior to issue and it has been suggested that they may be able to time the issue better. However, negotiated sales are perceived to lack competition and may result in charges of patronage to firms that are political insiders. Owing to this perception and the results of previous empirical studies, some states have considered legislation restricting issuers to only use competitive bid sales (Juarez & Bonpua, 1994). Previous empirical research has concentrated on three main hypotheses regarding differences between interest costs of competitive and negotiated issues. It has concentrated on the perception that lower competition in negotiated sales leads to rent-seeking behavior, that market search is greater under negotiation, and that market uncertainty leads to issuers seeking to maximize issue flexibility through using negotiated offerings. Most of the results have indicated that competitive bid sales result in lower interest costs for issuers. However, the paradox remains that the majority of municipal bond issues are negotiated offers. This paper offers a different explanation for the choice between negotiated and competitive underwriting. Negotiated underwriting may help to resolve information asymmetries that potentially increase interest costs for all new issues. By accurately certifying the fair price of new issues, negotiated underwriting can assist the municipal securities market in reaching a stable separating equilibrium. Since this is the case, a self-selection is created; issuers that most need to solve information problems are those issuers most likely to use negotiated offering. The problem with past empirical studies is that they have largely ignored selection biases. This paper applies for the first time a technique often used in studies of economic decisions to the question of the relative costs of negotiated offerings and competitive bid sales. This technique allows a clear interpretation of the results while controlling for self-selection problems. The results indicate that issuers rationally choose the underwriting method they use. Issues more likely to need independent certification of value are offered through negotiation. Once selectivity biases are controlled, negotiated offering is shown to produce at worst the same interest cost for issuers that use negotiation and at best lower interest costs.

The results presented for the underwriting decision suggest that municipalities make a rational choice in choosing the type of offering. These results are consistent with some previous findings (Leonard, 1996). The interest cost results suggest that the monopsony rent hypothesis can be positively rejected for our sample and the certification hypothesis is supported. Interest costs are lower for those issues that used negotiated offering, when the endogenous issuers’ choice of underwriting method is taken into account. These results contradict the prevailing wisdom, but are consistent with studies that control for selection biases using other methods (Leonard, 1996 and Smith, 1987). The certification hypothesis is supported both by the results of the negotiated/competitive decision model and the interest cost model. The results of the interest cost model indicate a positive interest cost benefit for the use of negotiated offers by issuers subject to information asymmetry. The fact that issuers that should be more likely to issue negotiated (smaller issuers with more economic uncertainty and higher debt burdens that will receive lower ratings) suggests that issuers in this sample are making choices of underwriting method in order to minimize interest costs. The interest cost model results indicate that once information asymmetries are resolved, issuers appear to be successful in minimizing these costs. Also, the market uncertainty hypothesis is supported in our model. The positive and significant coefficient on the T30RESQ variable in the probit model of the choice to issue negotiated suggests that issuers facing more interest rate uncertainty tend to increase their use of negotiated offering. Thus, issuers do seem to employing negotiated offering more frequently during periods of market uncertainty. The policy implications of these findings cannot be overstated. Therefore, any effort to restrict issuers to one type of method would impose costs on those issuers who were making an efficient choice of the other underwriting method. The drive to restrict issuers seems to be based on a fear that municipalities are being “taken” by underwriters. The research presented in this paper suggests otherwise. The analysis of method of sale has come a long way since its initial studies. This paper is just one more step in the process. Further research should be conducted on underwriting decisions, using different samples to test the various hypotheses. An interesting area of potential research is the relationships between underwriting and other certification decisions (e.g., credit ratings, reputation, bond insurance). It has been shown that these decisions are related (Kriz, 2000). It would be interesting to estimate a systems model of the issuance process in order to measure the effects of all of the relationships in a more formal manner.