Financial Misreporting and Debt Contracting G32

Subject Code :
G32
University :
others Exam Question Bank is not sponsored or endorsed by this college or university.
Country :
India

Financial Misreporting and Debt Contracting

September 2020

Abstract

This paper examines the effect of financial misreporting on bank loan contracting. In particular, this dissertation studies four of the various features in a debt contract and how it differs for firms which are required to restate their disclosed financial statement upon discovery of financial misreporting and firms that arent. Compared with firms with no restatement announcements, debt contracts of firms with restatement announcements have significantly higher spreads, smaller loan amounts, more financial covenant restrictions and a higher likelihood of collateral requirement. This paper also provides evidence that the effect financial misreporting has on bank loan contracting is magnified when a firm is closer to default. These findings consistent with banks enforcing tighter debt contract terms in response to an increase in risk and information asymmetry caused by financial misreporting.

Keywords: Financial misreporting; Restatement; Debt contracts

Abstract........................................................................................................................................ 2

Dedications and Acknowledgements............................................................................................... 3

Authors Declaration..................................................................................................................... 4

Introduction........................................................................................................................... 6

Literature Review and Hypotheses............................................................................................ 8

Research Design................................................................................................................... 15

Empirical Results.................................................................................................................. 19

Distance to default influence on the impact financial misreporting has on debt contract terms

Robustness Test............................................................................................................. 24

Conclusion........................................................................................................................... 25

Table 1 Summary Statistics.......................................................................................................... 27

Table 2 Difference between Means................................................................................................ 27

Table 3 Pairwise Correlations Matrix........................................................................................... 28

Table 4 Main Regressions............................................................................................................ 28

Table 5 Additional Analysis.......................................................................................................... 29

Table 6 Robustness test................................................................................................................ 30

References.................................................................................................................................. 31

1. Introduction

Financial information disclosed by publicly listed companies has been gaining more interest and attention from regulators and investors due to the recent controversies. Over the years, numerous companies including high-profile ones (such as Worldcom, Xerox, Carillion and Wirecard) have made restatement announcements, which in general diminished the initially reported earnings. General Accounting Office (2006) displayed concerning statistics on this phenomenon when the number restatements increased from 314 in 2002 to 523 in 2005 (affecting about 6.8% of list companies at that time). While the number of restatements has dropped in more recent years (Audit Analytics, 2019), the size of firms involved and the earnings revised is still concerning e.g. Tesco PLC, in 2014, has overstated its profit by an overwhelming amount of 250 million.

Restatement announcements are negatively perceived by the public as they sway investors confidence in the prospect of the firm and the reliability of the information disclosed in assisting investment decisions. Extensive studies can be found on the downside of restatement announcements where firms experience loss in firm value, a downward revision of earnings forecast and upward revision of cost of equity capital (Anderson and Yohn, 2002; and Hribar and Jenkins, 2004). Overall, the evidence insinuates that the impaired trust between corporations and market participants could result in a realised cost if investors choose to shy out from investing, thus hinders capital markets role of allocating resources efficiently.

Debt securities constitute by far the vital source of external capital for firms. The existence of syndicated loans and corporate bonds with maturities flexibility allows firms to raise large amounts of funds, extensively and, often simultaneously to finance investment activities. As cited in Graham et al., (2008) Federal Reserve System indicates that the flow of funds over the last decade consists of a landslide of $780 billion in net debt and only $2 million in equity. Specifically, the secondary loan market, our main focus, has grown significantly within the past decades to become one of the most innovative capital markets (Yago and McCarthy, 2004). Among the newly issued debt in 2005, more than half of it is owed to bank loans, while the remaining came from issuance of bond (Bharath et al., 2008). The rapid growth in liquidity and participation of various lenders and institutional investors are the result of the fully developed securitisation tools and ratings (Oldham, 1998).

Numerous literature can be found on the cost associated with financial misreporting, however, most focus on the equity market while scarce can be found on the debt market. The lack of attention requires a resolution following the vital role of debt and the growing cases of financial restatements. Given these absolute reasons, it is important to investigate the economic consequences of financial misreporting from the perspective of debt holders.

This paper investigates whether and how lenders react to restatement announcements, by focusing on the effect financial misreporting has on debt terms during the contracting process. I examine both price and nonprice terms of the debt contracts to provide a more multidimensional view on how it changes with financial misreporting. Based on existing literature, I predict that if restatement announcements are negatively viewed as associated lending risk and asymmetric information has increased, lenders will not only price protect themselves by charging higher borrowing cost but enforce shorter maturity period for loans and tighter debt contract terms with a greater number of covenants and collateral requirements. To conduct the research, regression analysis linking financial misreporting with the four dependent variables, which are loan spread, loan size, number of financial covenants and collateral requirement was done on a sample consist of 8,411 unique U.S based firms throughout 1996 to 2017. Bank loan contract data were extracted from DEALSCAN, while relevant accounting information of firms was obtained from the COMPUSTAT.

This research further reinforces and complements existing debt contract literature by documenting significant empirical evidence and the usage of more recent data. Consistent to Graham et al. (2008), following a restatement announcement, firms are subjected to higher interest rates and greater number of financial covenant in the debt contract. My sample which displays a significant decrease in the size of loan for firms following restatement announcements is similar to Wittenberg-Moermans (2008) research where she inversely associates loans size with information asymmetry experienced by a firm. The positive correlation between financial misreporting and collateral requirement found in this research supports Scott and Smith (1986), Berger and Udell(1990), Booth (1995) and English and Nelsen (1999) assertion that riskier borrower would pledge more asset as collateral to secure loan contract. All tested regressions are robust. This signifies that firm who engage financial misreporting will experience higher cost of debt, difficulty in raising larger amount of external capital, enforced with larger number of covenants and are more likely required to provide collateral as lenders try to protect themselves from the increasing risk and uncertainty. This research further extends the existing literature by undertaking additional analysis to test the severity of the impact financial misreporting has on debt contract terms when firms are closer to default. Significant evidence was obtained indicating firms within the reach of default were imposed with more unfavourable terms during the debt contracting process.

The remainder of the article is organised as follows: Section 2 discusses and review related literature on financial misreporting and debt contracts, as well as the building of the hypothesis. Section 3 describes the research sample data and the research methodology. Results, additional analysis and robustness tests are disclosed in the following part, Section 4. This research is then concluded in Section 5.

2. Literature review and hypotheses

2.1 Debt Security and Debt Contract

Debt is one of the three main finance pillars of an organisation apart from stocks and retained earnings. In 2006, more than 95% of newly issued external capital was some form of debt financing (bonds, debt, or other types of loans) and the balance of 5% was some form of equity (common or preferred stock). This statistic proves how frequent firm access the debt market in comparison to the stock market.

Most research on debt has agency theory (Jensen and Meckling, 1976) as its foundation. The theory provides three foundation outlook in understanding debt. First, owners/managers have incentives to engage in self-interest actions even at the expense of external capital providers. Two, in expectation of such behaviour, external capital providers will price protect their claims. Third, to counteract external capital providers' actions, owners/managers are willing to incur monitoring and bonding costs as a form of behavioural control mechanism. Conflicts on debt arent restricted between debt holders and owners/managers only. Smith and Warner (1979) highlight possible conflict debt holders might have with equity holders. Conflict arises over dividend payment when stockholders demand higher dividend payment which will restrict the available reserve to payoff debtholders claim. The level of leverage would be one of the sources of conflict. Future increase in debt increases financial distress cost and possibly leaving the firm in an overleveraged state. The difference in preferable investment course (asset substitution and underinvestment) proves to cause conflict between the two. Upon new debt issuance, firms would prefer to invest in riskier projects in hope of higher return which will transfer wealth from debt holders to shareholders. Alternatively, upon default possibility, firms may waive positive net present value investment opportunities because the limited-to-non benefit will be left for shareholders after catering for bondholders claims. All these risks will be taken into consideration by debt holders during debt contracting process.

Debt contracting process would conventionally be assisted by accounting information to reduce any agency cost incurred (Watts and Zimmerman, 1978,1986,1990). They highlight the inclusion of accounting-based covenants in debt contracts, which restrict dividend payments or following issuance of debt. Security agreements to collateralize the firms assets are often used to counter the transfer of risk conflict. Other most common covenants to be included would be capital expenditure restrictions, asset sale restrictions and sweeps, which obligate firm to serve their debt first using the proceeds. Although these provisions are competent in preventing asset substitution, issues such as underinvestment or commitment in maintaining appropriate future investment risk profile are difficult to be addressed using covenants. These costs can be reduced using price protection through higher interest rates or reduction in debts maturity. The explicit nature of debt contract and incorporation on various clauses and accounting-based covenant is the exact opposite nature of the relationship firms share with its shareholders which is mostly informal. This stimulates firms to be more concerned and creative with their choice of accounting, which may lead to the extreme resort of financial misreporting.

2.2 Financial Misreporting: Background

Financial misreporting is one of the few most commonly used terms to describe the act of fraud or misconduct in the financial world. An act which ceaselessly poses a significant threat to the existence and the efficiency of the capital markets itself. Financial reports which serve as one of the main communication channels between corporations and market participants to mould and forge trust are also a vital source of information to assist economical decision making. Its heavily relied upon characteristics makes it the most common alleged source of material to be falsified by a firm. Thus, any misreporting activities can affect both market participants trust and the capital market capability to allocating resources efficiently.

Financial misreporting can be done in various ways depending on its manipulator and his motivations, thus it is mentioned under different names and described with different characteristics across the vast literature. COSO/ACFEs Fraud Risk Management guide (2016) provides a generally accepted definition of fraud as any intentional act or omission designed to deceive others, resulting in a loss on the victims side and/or gain on the perpetrators. As it does give an overall understanding of what fraud is, its lack of depth on fraud identification and the containment of subjective elements is deemed unsatisfying from a researcher point of view. Amiram et al. (2018), with his in-depth literature review on this matter, clearly defines the common features it shares when defining misreporting as per stated within relevant laws and ruling. Misreporting is a combination of multiple elements which are (i) Misrepresentation should be in the form of misstatement, misreporting, or omission; (ii) That misrepresentation should be material; (iii) The person responsible for the misrepresentation must have done with some sault in the sense that the material misrepresentation was committed negligently, recklessly or with knowledge of its falsity; and, (iv) in private suits, the misrepresentation is causally related to a loss suffered by the plaintiff. Due to the nature of fraud which can be defined variously, reliance solely on common law principles may result in misreporting remain unruled-out as it can yield distorted results in the context of financial reporting when demonstrating fraud.

Karpoff et al. (2017) suggested a practical, objective and replicable method to ease identification of financial misreporting which is by relying on the Security and Exchange Commissions (SEC) fraud allegations presented under the 1933 and 1934 Acts. The most prominent acts to be known by researchers would be the Section 17(a) of the 1933 Securities Act covering fraudulent interstate transactions, Section 10(b) of the 1934 Securities Acts covering manipulative and deceptive devices, Section 13(b) requiring firm to keep accurate records and mandates internal control systems and section 13(a) requiring firm to timely file certain reports to the SEC. Any violations of these acts could be used against the firm in the accusation of adopting financial misreporting activities.

When violations of these acts is an act of crime, earning management on the other hand are legitimate actions done with discretionary accounting choices which fully complies with the GAAP provisions. However, earning management activities could still draw the same drawbacks as financial misreporting upon discovery. Activities such as overstating cashflows and/or accruals from its just state provide information that is misleading and harmful to the end receivers. Hill et al. (2019) support the deleterious effect earning management strategies may have on the long-term performance of the firm, while Kim and Sohn (2013) emphasized how investors account the real negative impact of earning management in predicting firms future performance. Because of this, prior literature has interchangeably use earning management literature to develop models and behavioural predictors for financial misreporting. An apparent case of financial misreporting may trigger a restatement issuance.

2.3 Financial Restatement and Lenders Belief

The Sarbanes-Oxley Act (SOX) of 2002 includes several provisions that require firms to rectify inaccurate, incomplete, or misleading disclosures (Palmrose et al., 2004). Financial restatement can either be voluntarily done by the firm or prompted by the SEC, an auditor or any combination of them that deemed its need. According to the United States General Accounting Office(2004), the number of restatements grew from an average of 43 per year for 1990 to 1997, to an outstanding number of 225 in 2001. The sudden surge in the number and size of restating companies has spark interest in researchers to have a better grasp of the event (Bardos and Zaiats, 2012). The most confabulate explanation by academics is the firms dismal performance. Managers of firms which typically experiencing high growth or improving performance due to the period of economic boom tend to set unrealistic expectations on the firms future performance. When reality sets in and proves the otherwise, they struggle to cope with the initial target sets, thus resort to financial misreporting to either at least meet the target or even beat it (Dechow and Schrand, 2004). Turner and Weirich (2006) explain that while complying to the SOX Act Section 404 which require firms to hire independent auditors to test the effectiveness of their internal controls, has led many of them to restate their disclosed account. Regardless of the reason behind a restatement, either an honest incorrect revenue recognition or other misapplication of GAAP, or caused by serious criminal activities such as corporate fraud, it will still adversely affect the firm (Graham et al. 2008).

Financial restatements are normally interpreted as manager incapability of assessing and reporting the firms prospect in a credible way, thus the need for a correction. Upon revelation of accusations of financial misreporting and restatements, firm will experience a significant fall in its shareholder value as documented by Gande and Lewis (2009), which clearly shows the negative bias of investors. This action affects investors evaluation of a firm through revisions in beliefs about the firms expected future cash flows and/or the uncertainty about the firms financial information (Graham et al., 2008). Watts and Zimmerman (1986) and Holthausen and Watts (2001) highlight the importance of financial statement information as one of the main sources of information to assist decision making such as contracting activities. A restatement will annul previous predictions made on firms prospect and most probably will further discount it instead as Kinney and Mc Daniel (1989) find that in the number earning overstatement announcement is twice the number of earning understatement announcement in their sample, similar with Graham et al. (2008) which record overstatements outnumbered understatements by nine to one.

This discount effect can be explained by the downward adjustment of expected cash flow and the upward adjustment of cost of capital. Hribar and Jenkins (2004) provide evidence supporting these claims when they find that upon restatement, analyst will negatively revise earning forecast and increase the cost of equity, especially for the highly leveraged firm and restatements initiated by auditors. Firms reputation which is put on the line may lead to a real harmful effect on its cash flow. The firms ability in generating earnings will be questioned and a revision on its credit profile risk will be done which will influence the firms term of trade with its investors, customers and suppliers (Karpoff, Lee and Martin, 2007) to be worse off. Apart from that, the loss of management credibility may indicate sub-optimal investment and operation policies, thus the need for tighter monitoring activities and management changes which are costly. Burks (2010) records a high turnover number of top management and poorer job prospect following a restatement. Karpoff, Lee and Martin (2007) further explain the substantial loss from harsh legal and regulatory sanctions that top managements have to suffer after their firm being enforced to do restatement by the SEC. This appalling state entails an increase in firm default risk, thus will be lead to a stricter debt contract terms demanded by lenders.

Misreporting does not only affect the trust investors have on the firms future performance but also the credibility of the information its disclosed. It creates uncertainty and reduces the quality of the disclosed information. As readers or financial statement users tend to skim through financial statement and disregard other disclosed note or information, a restatement on one particular accounting item would instead cause the firm to be judged as a whole. Other aspects of the firms operation and reported performance will be questioned. Palmrose, Richardson and Scholz (2004) provide evidence for this when they find that following a restatement announcement, there was a significant increase in analyst forecast dispersion. Financial statements were meant to address information asymmetry, instead, the act of restatement increase them and the risk associated with lending.

2.4 Financial Misreporting and Debt Contract

Following Jensen and Mecklings (1976) framework, lenders are expected to demand higher returns as compensation for the agency cost associated with managers incentive to act against their best interest. The component of financial accounting can be used to mitigate this agency cost when misreporting on the other hand does the exact opposite. Lending risk of a lender is significantly associated with borrowers credit risk and to be heavily accounted for when determining loan pricing (Freixas and Rochet, 1997). Greater credit risk increases lending risk, thus the higher demanded interest rate. Barry and Brown (1984) claim that the volume of available information influences systematic risk and scarce information available should raise systematic risk, thus the need for it to be reflected in the price of securities. Diamond and Verrecchia (1991) with their acute perspective argue on how firm with greater information transparency attracts greater number of investors, increasing securities liquidity which leads to reduction in firms cost of capital. From this, we can expect that restating firms, with greater credit risk will bear wider loan spread.

H1: The spread of a loan deal is positively associated with the event of a financial misreporting.

The size of a loan which serves as a reflection of trust and expectation lenders have on the ability of the firm to serve its commitment is liable to the downside caused by financial misreporting. Trust thats forged in accordance to the amount and quality of information available on the regard of firms prospects will be endangered as financial misreporting introduces asymmetry information, thus affect lenders volition in lending a during the contracting stage. Jones et al. (2005) claim that larger loans suffer from less severe information asymmetry issues as any fixed cost associated with garnering information on the borrower are lower. While loans size is typically positively correlated with the size of a firm, so does the extent of information asymmetry. Diamond and Verrechia (1991) state that larger firms obtain more benefits from less asymmetry information than smaller firm due to the tighter regulations and disclosure requirements, while Bharath et al. (2004) and Bharath et al. (2007) highlight how small borrowers suffer from greater information asymmetries and distress probability. Wittenberg-Moerman (2008) disclosed evidence in her research that larger loans are anticipated to be associated with lower information asymmetry issues. Burns and Kedia (2006) provide evidence on how restating firms are on average riskier than non-restating firm due to their higher leverage and Price-earnings ratio. The widely accepted cognizance that financial misreporting may have detrimental effects on investors judgements on a firm as information asymmetry and risk increases, enable us to deduce that in the event of financial misreporting, firm will suffer from restricted access to external capital, allowing them to borrow at a smaller amount.

H2: The size of a loan is negatively associated with the event of financial misreporting.

Smith and Warner (1979) highlight the importance of financial covenant in debt contacts in reducing associated agency cost. Because agency costs are caused by information asymmetry, most covenants are likely to be designed around the elements of firms financial reporting system. Costly Contracting Hypothesis (CCH) by Smith and Warner (1979) asserts the trade- off firms undergo by including covenant in their debt contract is to benefit from lower agency cost of debt. Debt covenants are used to reduce firms flexibility and enhancing banks incentives to monitor. Bradley and Roberts (2015) empirically documented the negative relation between firms financial health and the presence and density of covenants in debt contracts. This literature implies that distressed or risky firm would heavily use of covenants in debt contracts.

H3: The number of financial covenants in a loan deal is positively associated with the event of financial misreporting.

Further reference made to Smith and Warners (1979) research, they discussed how the difference in preferable investment course (asset substitution and underinvestment) may ignite conflict between lenders and borrowers. The transfer of wealth from debt holders to shareholders and vice versa for risk highlight the importance of collateral to address this issue. Welch (1997) and Morris and Shin (1999) claim the usage of collateral as a contractual instrument that strategically aims in restricting future borrower behaviour in a way desired by lenders. The existence of collateral requirement restrict irrational decision making and enhance efficient monitoring (Rajan and Winton, 1995) as a certain or specific part of firms asset need to be registered as collateral, reducing available resources and should investment has gone wrong, lenders have priorities over the liquidated asset collateralised. Scott and Smith (1986), Berger and Udell(1990), Booth (1995) and English and Nelsen (1999) find that riskier borrowers tend to pledge more collateral for better loan contract terms. Bharath et al. (2008) provide evidence on how in private debt markets, firms with higher accounting quality can negotiate a less restrictive collateral requirement. From this, we can expect that restating firms with greater agency cost and risk, are more likely to have their assets as collateral in a loan contract.

H4: The likelihood of a loan deal being secured is greater with the event of a financial misreporting.

3. RESEARCH DESIGN

3.1 Data

The corporate misreporting data was obtained from financial restatement database collected by the Audit Analytics. This database includes restatement announcements by 8,411 unique U.S based firm ranging from the period of 1996 to 2017. The announced restatement consist of accounting irregularities resulting in material misstatements of previously disclosed financial reports. The sample includes restatements both caused by material errors and fraud. Following Graham et al. (2008) and Chava et al. (2009), bank loan contract data are secured from DealScan, a Loan Pricing Corporation (LPC), for, information on the total of US 40,498 loan deals were issued to the U.S firms throughout the 21 years were discovered. Relevant accounting information used to measure control borrowers characteristics were obtained from the Compustat. Further sample elimination was made as firms operating in the financial industries (SIC codes between 60000 and 6999) were removed due to difficulty in interpreting variables such as leverage and market-to-book (MTB) value that were used as controls. To rule out the effects of extreme values, continues variables are winsorized at 1th and 99th percentiles, leaving a final sample of 16,243.

3.2 Multivariate analysis

The effect of misreporting on the debt structure and debt contracts terms will be examined using OLS regressions analysis. The main empirical models are as follows:

Log(Loan Spread)i,t (1)

= f1 + f2misreportingI,t + f3log(asset)I,t + f4profitabilityI,t + f5LeverageI,t + f6MTBI,t+ industry FE + Year FE + ?I,t

Log(Loan Size)i,t

(2)

= g1 + g2misreportingI,t + g3log(asset)I,t + g4profitabilityI,t + g5LeverageI,t + g6MTBI,t+ industry FE + Year FE + ?I,

Log(NumCov)i,t

(3)

= h1 + h2misreportingI,t + h3log(asset)I,t + h4profitabilityI,t + h5LeverageI,t + h6MTBI,t+ industry FE + Year FE + ?I,t

Collaterali,t (4)

= j1 + j2misreportingI,t + j3log(asset)I,t + j4profitabilityI,t + h5LeverageI,t + h6MTBI,t+ industry FE + Year FE + ?I,t

From the regression equation seen above, f1, g1, h1 and j1 are intercept terms while f2, g2, h2 and j2 are the coefficient of the test variable misreportingI,t in each regression. Fi=3,4,5,6, gi=3,4,5,6, hi=3,4,5,6 and ji=3,4,5,6 are coefficients of the four control variables (ie., log(asset), profitability, leverage and MTB) in respective equation. ?I,t is the error term. Year and industry fixed effect are controlled as time trend and heterogeneity across the industry may influence the regression result.

3.3 Independent Variable

Consistent with previous studies, this research uses a dummy variable to reflect whether the firm has made a restatement announcement as a proxy of misreporting. If the firm has made a restatement announcement in the given year, the dummy will take a value of 1 and the value of 0 otherwise.

3.4 Dependent Variables

Reference made to the hypothesis section mentioned earlier, four loan characteristics which have been identified to be greatly influenced by misreporting have been chosen as the dependent variables. Following prior literature, the four dependent variables are (1) Loan Spread, (2) Loan Size, (3) Number of financial covenant and (4) Collateral requirement

Loan Spread

The borrowing cost of a bank, loan spread (log(Loan Spread)), is calculated according to the amount the borrowers pays in basis points over LIBOR or LIBOR equivalent for each dollar drawn. This measure adds to the borrowing spread any annual fees paid to the bank. It is predicted to be inversely related to misreporting. As financial restatement increase information asymmetry, which leads to higher lending risk, the cost of borrowing should rise accordingly. Hence the coefficient of misreporting is predicted to be negative.

Loan Size

The size of the loan (log(Loan size)) should reduce in amount following the event of misreporting. Lenders would be willing to lend a higher amount to a firm greater information and financial stability. Thus the coefficient of misreporting is expected to be negative.

Number of financial covenants

The number of financial covenants in a loan deal (log(NumCov)) with misreporting which is expected to have positively related. Borrowers would be required to adhere to larger number of financial covenants in its attempt to reduce the agency cost. This would produce a positive expected coefficient of misreporting.

Collateral Requirement

The existence of collateral requirement (collateral) used dummy variable as its representation. Should lenders require asset to be registered as a collateral in the debt contract, the dummy will take a value of 1 and should there be no collateral requirement, The value of 0 will be recorded. It is highly expected that a loan deal would more likely to require asset as a collateral from firms that adhere misreporting activities. The coefficient of misreporting is predicted to be positive.

Upon restatement announcements, lenders will revise their view on firms future performance and the reliability of disclosed information, which normally be negatively-biased. Lenders dismayed trust and expectation on the firm cause greater reluctancy in lending as the perceived lending risk is higher. This explains their demand for higher loan spread, smaller size of loan, greater number of financial covenants and collateral requirement.

3.5 Control Variables

To increase the accuracy of the regression, other firm characteristics are examined and controlled for to alleviate the effects of other factors which may impact this study. By referring to previous research in examining the effect misreporting have on debt structure and contract terms, four control variables have been shortlisted. The four control variables included are (1) firm size, (2) profitability, (3) leverage and (4) market to book value.

Firm Size

The firm size is derived as the natural logarithm of firms market value of equity (log(asset)). This variable should be controlled for as larger firms are associated with better corporate governance and greater access to external financing. Larger firms are tied with This enabled them better bargaining power during contracting activity.

Profitability

Firms profitability reckoned as firms return on asset (ROA) which is calculated by dividing the incomes before extraordinary items with total assets. This variable is chosen as it represents firms financial strength in generating and/or stabilising earning. Firms with greater profitability is expected to have larger excess to earning reserve, allowing them to timely serve their obligation and reduce the risk of financial distress. Thus, allowing them to have better loan term deals.

Leverage

Leverage evaluated as firms total liability over total asset. This should be controlled for as overleveraged firm is associated with greater risk of being in the state of insolvency and default. It can be expected that firm with high amount of leverage will be forced to bear higher borrowing cost.

Market to Book value

In this research, the firms growth opportunity is represented by the firms MTB value (ie., the market value of assets divided by the book value of assets). This is crucial as MTB represent investors valuation of the firm. Firms with high MTB value are perceived to have strong earnings growth, thus allowing them to borrow at a better term.

4. Empirical Result

4.1 Summary Statistics

The sample has been cut from 34,764 to 16,243 after the introduction of data constrains by additional variables. Table 1 presents summary statistics for misreporting dummy, loan variables (i.e, loan spread, loan size, NumCov and collateral) and control variables (i.e, firm size, profitability, leverage, MTB). The misreporting dummy has a mean of 0.130 indicating over the period 1996 to 2017, a total of 2112 firms have made restatement announcements due to financial misreporting. In respect to firm variables, the distribution of firm size, profitability, leverage and MTB have mean (median) values of 4,276.940 (751.450), 0.014 (0.037), 0.297 (0.278) and 1.764 (1.446) respectively after winsoring outliers. This can be interpreted as on average firms have positive ROA, total liabilities compromising 29.7% of their total asset and can be labelled as high growth firm for having MTB value greater than 1. In terms on loan variables, mean (median) of the loan spread, loan size, NumCov and collateral are 204.157 (175.000), 443.874 (200.000), 2.446 (2.000) and 0.614 (1.000) respectively. It means that a debt contract on average pays a hefty sum of 204 points over LIBOR and have an average loan amount granted of USD 443.874 million. Each loan deal consists on average of 2 financial covenants and 61.4% of its required assets as collateral.

Table 2 discloses the univariate comparisons of these variables between firms with financial misreporting activities and firms without financial misreporting activities to give an early overview of how financial misreporting may affect the debt contract terms. The mean loan spread grows from 210 basis points over LIBOR for the non-misreporting firm to 235 basis points for the misreporting firm. The average loan size shrunk from $545.521million for the non-misreporting firm to $478.092million. In general, firms that misreport and those that do not have the same number of covenants in their debt contract which is 2. Nevertheless, the difference in coefficient records a negative value of 0.052 (t-statistic = -2.051), proving the negative effect misreporting have on the number of financial covenants in a debt contract even by a small margin. Firms with misreporting activities record on average a slightly higher possibility of being required to provide collateral in comparison to firms without misreporting activities with a difference of 7% (t-statisitc = -8.454). All these difference in mean are significant at 1% level except for number of covenants which is significant at 10% level.

As for the control variables contained in the research model, misreporting firm on average can be seen to be smaller in size, highly leverage, lower in both ROA and MTB value. This reflects the tendency for firms with less tangible assets, high volatility, financially distressed and low growth opportunity to engage with financial misreporting activities. By being significance at 1% level, all these control variables considered do have some impact on companies misreporting activities and debt contract terms.

Table 3 reports the pairwise correlations matrix among variables tested across this research. Overall, misreporting is significantly related to all the dependent variables in the 1% significance level except for the number of covenants (ie, NumCov) which is significant in the 5% significance level. It was also shown that respectively, misreporting has a negative correlation with loan size and positive correlation with loan price, NumCov and collateral. These figures provide preliminary support for all the hypothesis to be tested, however, to provide more concrete evidence against the relationship between financial misreporting and debt contract terms, multivariate analysis has been executed

4.2 Effect of financial misreporting on debt contract terms

The results of the four regression (ie., equation (1) to (4)) are presented in table 4. The top horizontal columns list four dependent variables while the left side of the vertical columns lists the independent variables, control variables and additional statistics (ie., number of observations and R²). Regressions in each column include firm characteristics, year fixed effect and industry fixed effect as its control.

4
.2
.1 Loan spread

Column 1 of Table 4 analyses the effect of misreporting on the price of debt. The estimated misreporting coefficient is 0.101 with a t-statistic of 6.22 and is in the 1% significance level. This indicates that firm with misreporting activities is charged with loan spread of one and a tenth per cent higher. This empirical result from the regression of equation (1) is consistent with what stated in hypothesis H1, where misreporting activities drives the loan spread up.

On regard of the control variables, it can be deduced that firm with smaller size (coefficient= - 0.205, t-statistics= -42.76 of log(firm size)), lower ROA (coefficient= -0.507, t-statistics= -9.08 of profitability) and heavily leveraged (coefficient= 0.371, t-statistics= 10.41 of leverage) would face a higher borrowing cost of debt. Theres no suffice evidence to significantly and statistically supports the influence growth opportunity (MTB) has on the loan spread as the coefficient (t-statistics) obtain is 0.00 (0.004).

4.2.2 Loan size

Column 2 of Table 4 records the negative relation between the size of loan and financial misreporting. Regression of equation (2) results in misreporting coefficient (t-statistic) of - 0.070 (-3.00) and it is significant at 1% level. This signifies that the size of a loan decrease with the existence of misreporting activities, further supporting the prediction in hypothesis H2.

For the firm-specific control variable aspects, theres an indication that larger sized firms (coefficient= 0.484, t-statistics= 55.38 of log(firm size)) with higher ROA (coefficient= 0.294, t-statistics= 2.51 of profitability), heavier level of leverage (coefficient= 1.150, t-statistics= - 20.58 of leverage) and lower MTB value (coefficient= -0.235, t-statistics= -20.92) are associated with larger loan size. All coefficients of the control variables are significant at 1% level except for profitability at 5%. The greater access to external debt finance is due to firm owning more tangible asset and financial stability. Although the sign of the leverage and MTB coefficient is against the initial belief, this can be explained by the facts that; 1. firm are capable of receiving more leverage only when they have good credit record, credit quality is positively related to loan size and 2. MTB ratio may signify firm is undervalued, lenders or banks who have access to inside information and closer relationship with a firm may have a better understanding of the firms' actual financial position which differs from what stated in the disclosed financial statements.

4.2.3 Number of financial covenant

Column 3 of Table 4 lays the result of equation (3)s regression. Financial misreporting is shown to have a positive effect on the number of financial covenants in a debt contract term when the resulted misreporting coefficient (t-statistic) is 0.069 (5.26). The misreporting coefficient has a positive significant value at 1% level. This provides empirical evidence in support of hypothesis H3, where financial misreporting results in higher number of financial covenants to be included in a debt contract.

Regarding control variables, the results stipulate the plausibility of firm that is smaller in size (coefficient= -0.062, t-statistics= -17.35 of log(firm size)) with higher ROA (coefficient= 0.606, t-statistics= 15.94 of profitability), greater leverage level (coefficient= 0.086, t-statistics= 3.27 of leverage) and lower growth opportunity (coefficient= -0.001, t-statistics= -0.20) will adhere higher number of financial covenant in a debt contract terms. Coefficient of MTB has an insignificant value while the other controlled variables are significant in the 1% level.

4.2.4 Collateral requirement

The outcome of the equation (4) in testing the possible effect financial misreporting have on the possibility of a debt contract to include collateral requirement is presented in table 4 column 4. A positive relation has been identified between misreporting and collateral as the coefficient (t-statistic) produced is 0.051 (4.81) and it is in the 1% significance level. This can be interpreted as firms that associate with financial misreporting are more likely to be required to provide asset collateral in a debt contract. The results are in line with hypothesis H4.

With respect of firm-specific control variables, the obtained coefficients (t-statistic) for log(asset), profitability, leverage and MTB are -0.096 (-42.96), -0.179 (-6.02), 0.176(8.49). All of the coefficients are in the 1% significance level. These indicate that small, volatile, highly leveraged firm with high MTB value are more likely to have their debt secured to a tangible asset. The positive relation between MTB and collateral requirement is due to the possible overvaluation interpretation of the MTB ratio. Banks information source is not limited only to the financial statement as it may be to other investors. This allows them to have greater chance of identifying overvalued firm, thus, increasing the chance for a debt contract to include the collateralisation feature.

4.3 Additional analysis

To further emphasize the effect financial misreporting have on debt contract terms, additional analysis with firms conditional setting of distance to default is conducted. Distance to default or default probability of a firm can take the smallest possible value but it can never be nought. Lenders recognise the massive possible loss in the event of default, thus urge its identification (Koll?r, 2014).

Distance to default is associated with various firm-specific variables such as firm size, profitability, leverage and MTB value. Firms with larger amount of asset, greater earning generating capability, non-heavily leveraged and wider opportunity growth are capable of widening the default distance. However, being nearer to default could turn these variables against the firm itself. Being within default reach increases financial distress level as firms are forced to resort desperate measures such as selling off their asset and forego long term investment opportunity to raise and/or maintain short-term liquidity. Relationship between various stakeholder will be at stake should firm engage credit extension, dividends cut,, employee redundancy in an attempt to escape this sinkhole. Short distance to default increases the uncertainty of firms actual financial health and firm announcing financial restatement due to misreporting further worsen the situation. Information asymmetry increases as previously disclosed reports are no longer reliable. This type of firm has more severe risk and information problems, thus, investors will expect greater compensation which will be reflected in stricter loan contract terms. For these reasons, we can induce that the effect financial misreporting have on debt contract terms will be amplified when the firms distance to default becomes narrower.

The distance of default will be measured according to Altmans (1968) Z-score method. A dummy variable (DtoD) will be created to reflect the default distance of the firm. DtoD is equal to 1 if the Z-score is greater than the sample mean Z-score, indicating firm is far from default and DtoD is equal to 0 if the Z-score is smaller than the sample mean Z-score, signifying firm is close to default. The additional analysis will be done by using four separate multivariate analysis comparisons between each loan-specific variables with long-distance to default firm and short-distance to default firm. Results are recorded in two separate panels (Panel A and Panel B) in table 5. Panel A shows estimation results of the effect misreporting have on the cost of debt (log(loan spread))and loan size (log(loan size))while Panel B shows estimation results of the effect misreporting have on number of covenants (log(NumCov)) and the collateral requirement (collateral).

Panel A of Table 5 discloses the regression result between misreporting with Loan Spread and Loan Size under the influence of distance to default (ie., DtoD). Column (1.1) analyses the number of financial covenants for far from default firm with misreporting while column (1.2) does the same except its for firm close to default. An observable difference can be seen as the coefficient (t-statistic) rise from 0.058 (2.41) to 0.123 (5.91). High default risk firms with financial misreporting records will suffer a higher cost of borrowing. Column (2.1) and (2.2) compare the effect misreporting have on loan size for firms that are far from default (DtoD=1) and firm thats near to default (DtoD=0), a greater negative coefficient is recorded for firm closer to default than firm far from default with values of -0.069 and 0.063. It is statistically significant at least at the 10% level. This reflects the restriction financially distressed firm with misreporting activities has in accessing external source of finance.

The result reported in panel B table 5 indicates that distance to default does magnify the effect misreporting has on the number of financial covenants and the likelihood of a collateral provision to be included in a debt contract. Column (3.1) and (3.2) compare the effect misreporting has on the number of covenants (log(NumCov)) and collateral requirement (collateral) for firms that are far from default (DtoD=1) and firm thats near to default (DtoD=0). The estimated coefficient (t-statistic) increases from 0.049 (2.37) to 0.077 (4.54) and are significant. This serves as the empirical evidence in supporting the notion that firm that is closer to default and adheres financial misreporting experience greater number of financial covenants in the loan contract term. Referring to column (4.1) and (4.2) which compares the effect misreporting have on the collateral requirement for firms that are far from default (DtoD=1) and firms that are near to default (DtoD=0), a slight but significant increase can be seen as the coefficient (t-statistic) rise from 0.046 (2.70) to 0.048 (3.59). Both values are significant at 1% level. This supports the claim that firms experiencing financial difficulty and resorting misreporting activities are more likely to be required to provide collateral in a loan contract.

The results support from this additional analysis highlight the extreme effect misreporting has on debt contract term under the pressure of being nearer to default which is consistent with the results obtained by Chava et al. (2017) and Graham et al. (2008).

4.4 Robustness test

In previous regression, various firm-specific variables have been controlled for. However, it is possible that some other loan-specific variables may correlate to the loan contract terms itself. To test the robustness of equation (1) to (4), this research introduced two extra loan-specific variables which are log(maturity) estimated as the natural logarithm of loan maturity and log(NumLenders) estimated as the natural logarithm of the number of lenders in a syndicated loan deal. Borrowers requiring to borrow at a longer-term would reduce the opportunity for lenders to renegotiate debt contract term at a more frequent rate should there be news on firms financial misreporting. This causes risk and information problems created from financial misreporting to be left unaddressed. Reduced number in lenders will apportion a larger percentage of risk to be bear by the remaining lenders. Thus, magnifying the effect of financial reporting have on the debt contract terms.

Table 6 reports the result of the regression. The results exhibit the coefficient (t-statistic) ofmisreporting with respective log(loan price), log(loan size), log(NumCov) and Collateral to be 0.091 (5.80), -0.045 (-2.27), 0.068 (5.29) and 0.045(4.44). All coefficients hold the same significance level (1% significance level) against the initial regression result displayed in table 3 with the exception of loan size (log(loan size)) which is now significant at 5% level. Nevertheless, the effect of financial misreporting on debt contract terms are robust to various specification and remain economically and statistically significant.

5. Conclusion

In this paper, I examined the negative effect of financial misreporting by investigating the differences in debt contract terms of firms with and without restatement announcements. This research is conducted on a sample of 8,411 U.S based firms from 1996 to 2017 and tested through regressions linking financial misreporting with the four dependent variables, which are loan spread, loan size, number of financial covenants and collateral requirement. The results obtained provide significant evidence in support of banks enforcing tighter debt contract terms in response to an increase in risk and information asymmetry caused by financial misreporting. Financial misreporting records a negative relation with loan size and positively related to loan spread, number of financial covenants and collateral requirement. All tested regressions are robust. This signifies that firms which engage with financial misreporting activities will suffer higher cost of debt, difficulty in raising a larger amount of external capital, larger number of financial covenants and are more likely required to provide assets as collateral. Additional analysis is conducted to test the magnitude of the negative impact financial misreporting has on debt contact terms when a firm is closer to default. Significant evidence was obtained deducing firm within defaults reach experience more unfavourable terms during debt contracting process. Bank considers restatement announcements as firms previously misreport as it attempt to conceal its actual state, has now required corrections. The intensity of uncertainty has led banks to revise their initial belief and expectation they have, which work against the favour of the firm.

This research exists within several limitations due to operational restrictions. First, the source of the restatement was disregarded. Financial restatements may be triggered by fraud-related misreporting or error-related misreporting. The severity of the possible drawback varies as banks are expected to penalise fraud more heavily in comparison to an honest mistake. Graham (2008) do find evidence in support of this notion when the loan spread for fraud-related restatements records a higher value relative to non-fraud-related restatements. Regardless of that, both are expected to negatively affect debt contract terms, as error-related misreporting diminished the credibility of the information disclosed thus increase the lending risk. The research design only accounts for four firm-specific control variables (i.e., firm size, profitability, leverage and market to book value) when realistically, various other factors may influence debt contract terms. Other firm-specific factors such as firms tangibility, cash flow volatility and asset maturity, and macroeconomics factors such as credit spread were proven to have significant relation with the cost of debt (Graham, 2008). Generally, banks will require higher compensation from firms with lower tangibility, greater cash flow volatility, longer asset maturity and when the economic condition experiences a recession. Lastly, it is important to understand that this paper documents the existence of correlations between financial misreporting and debt contract terms tested and not the causation. Evidence reported shows the positive correlations financial misreporting has on loan price, number of financial covenants and the likelihood of being secured and negative correlations with loan size. The setting of this research restricts the ability to deduce a legitimate causal relationship as confounding variables cannot be completely removed. Future research might complete the control factors and apply different approach settings to better prove the causal relations that financial misreporting does cause lenders to be more pessimistic during the contracting process, thus enforce tighter debt contract terms.

In conclusion, financial misreporting has always been the most discussed topic either by academics or in the corporate world. Managers who attempt to alter firms reported economic performance to either mislead stakeholders view or influence contractual should put more thoughts when weighing between possible gain and loss. Managers are driven to engage this misbehaviour to gain private control benefits, and maintain their reputation and credibility in stewarding an organisation, may only continue to do so when it remains discrete. Failure to remain discrete may results not only in disciplinary actions to be taken against them (Zingales, 1994) but also a direct tangible cost to the firm as shown in this and various existing research.

Order New Solution

Uploaded By : Akshita
Posted on : May 23rd, 2025
Downloads : 0
Views : 65

Financial Misreporting and Debt Contracting G32