Academic journal article Academy of Accounting and Financial Studies Journal

Interest Rates Are Sticky: Implications for the Yield Curve

Academic journal article Academy of Accounting and Financial Studies Journal

Interest Rates Are Sticky: Implications for the Yield Curve

Article excerpt

INTRODUCTION

Interest rates have fallen somewhat in over the past few years following the unconscionable losses in the real estate mortgage market, followed by the government bailout of the partially responsible institutions, leading ultimately to a period of easy money now known as quantitative easing. Until recently (2013), the impetus for firms to take advantage of the lower rates with heavy rollover of existing debt to the lower rates has largely not materialized.

Expansion plans have been effectively put on hold due to the lack of strong consumer demand. Businesses wisely refrain for issuing new debt in uncertain markets, but this does not fully explain why firms do not call existing debt and refinance at lower rates. Refinancing requires a financial incentive because of the cost of both calling existing debt and reissuing new debt is ultimately paid for by the corporation raising their effective cost of capital.

If a reduction in interest rates is not uniform across all risk classes, then incentives to refinance would vary according to the bond rating. This paper examines the change in interest rates in two ways. First the absolute change in rates is measured to see if the reduction across bond ratings is uniform. Then a more rigorous test using a certainly equivalent method is applied to determine the comparative change in rates implied in various bond rating categories while controlling for changes in capital structure.

LITERTATURE REVIEW

The Certainty Equivalent (CE) method in capital budgeting is thought to be superior to applications of a Risk Adjusted Discount Rates (RADR), (see Sick 1986, Zhang 2010, Cheremushkin 2010 et.al.). The individual CFs are adjusted to reflect the cash that is assured with little or no risk. The Risk Free (Rf) rate is then used as the discount rate in the CE method, once the CFs have been reduced to account for risk. The 10 year Treasury serves as the proxy for the Risk free rate. To evaluate this question, estimates of past and present cost of capital are coupled with past and present Rf data to examine the stability of the CE model.

There are two assumptions of the Certainty Equivalent method. First, practitioners assume that future CFs are accurately adjusted downward to the likely cash inflow estimates accounting for minimal projected sales. Since the discount rate for CE calculations is the Rf rate, a known value, then the estimation of the CE cash flows becomes the main focus of concern. If the CF estimates are either too high (resulting in the NPV biased upward), or too low (resulting in the NPV biased to reject), errors in capital budgeting decisions are likely to occur. To correctly apply the CE method, it is of paramount importance to accurately estimate CE cash flows. One can only hope the cash flow reductions applied in the CE method represent realistic financially distressed sales at the margin otherwise the project NPV would biased. To avoid this pitfall, the present study uses established CFs from example problems commonly found in academic texts: See Brigham and Ehrhardt (2009) and Brigham and Daves (2013).

Risk adjustment under uncertainty is not new and constitutes an ongoing process of discovery in the literature. For early discussions of uncertainty reduction bias in capital budgeting see Mukherjee (1991). The issue is primarily the problem of the timing of the uncertainty. Specifically, the further into the future the CFs are predicted, the more uncertain the estimates. The RADR method offers as its' strength the security of discounting. In application a fixed cost of capital results in future CFs being increasingly reduced as they grow further from the present time frame. In reality it is the random timing of the future CFs that is unknown. Unanticipated changes in the final CF models, both good and bad are to be expected. In Perrakis (1975) the author allows an element of randomness in future estimates as an improvement to the model. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.