Accounting information, life cycle and debt markets

Transcription

1 Accounting information, life cycle and debt markets Attila Balogh a, Jiri Svec b and Danika Wright b* a School of Banking and Finance, UNSW Business School, Australia b Discipline of Finance, The University of Sydney, Australia ABSTRACT We examine the previously unexplored relation between risk of default and life cycle. Using credit default swap (CDS) spread data for a US sample of firms from 2007 to 2016, we find large variations in the market pricing of default across the firm life cycle. Specifically, CDS spreads are lowest for growth and mature firms, and highest for introduction and decline-stage firms. We show that life cycle provides additional information to asses credit risk not captured by age, as the life cycle measure allows firms to evolve in a non-monotonic and non-sequential manner. Date: 17 January 2018 Keywords: Credit default swaps, Corporate life cycle JEL Classification: G12, G32 *Corresponding author. danika.wright@sydney.edu.au 1

2 1. Introduction The life cycle stage of a firm is an important determinant of risk, growth opportunities and valuation. In models of default risk, firm age is commonly used as a proxy for life cycle. However, this proxy oversimplifies the typically complex path firms follow through time with respect to their likelihood of default. In this paper, we examine the relation between default risk and life cycle using a recently developed proxy based on firm accounting information. A broad literature demonstrates the influence of corporate life cycle on equity investments. Key results show that mature firms have lower growth rates (Mueller, 1972; Mueller, 1975), lower internal growth opportunities (DeAngelo et al, 2006), and a lower cost of equity (Hasan et al, 2015), relative to firms at an early life cycle stage. Life cycle is also identified as an important predictor of seasoned equity offering (DeAngelo et al, 2010) and dividend payout policy (DeAngelo et al, 2006). Recent research by Koh et al (2015) demonstrates a link between life cycle and firms response to financial distress. Our paper extends the life cycle literature to an analysis of debt market pricing using Dickinson s (2011) cash flow-based classification system. There is growing interest in the accounting research literature as to what financial statement data, and particularly cash flows, can reveal about a firm s risk of default and the pricing of its debt. The volatility of firm cash flows is shown to be positively related to corporate bond spreads by Douglas et al (2016) and to credit default swap (CDS) spreads by Tang and Yan (2010). In the present study, we use cash flow data to classify firms life cycle following the Dickinson (2011) method to determine the forward-looking risk of the firm via CDS spreads. 2

3 In this way, we examine what accounting information can reveal about a firm s life cycle and its impact on debt security pricing. CDS are tradable derivative contracts, in which the holder of a long CDS receives payout when the underlying firm defaults. 1 CDS spreads provide several important advantages over earlier studies using corporate bonds. CDS spreads do not require a benchmark risk-free yield curve, provide greater liquidity, lead credit spreads in the price discovery process, and are less contaminated by non-default risk components (Longstaff, Mithal, and Neis, 2005; Brennan, and Marsh, 2005). Using data for a large sample of firms from the US from 2007 to 2016, we assess the information content of the cash flow life cycle classification in pricing firm default risk. We find that default risk follows a distinct U-shape over the corporate life cycle, with growth and mature-stage firms demonstrating the lowest average default risk. In particular, we find that firms in the introduction and decline phase of the life cycle trade at a 27 basis points (bps) premium and a 47 bps premium in 5-year CDS contracts, relative to firms in the mature stage of the life cycle, respectively. The relative safety of mature stage firms compared to firms in other life cycles is statistically significant and persistent across time. The result are more pronounced during the global financial crisis, with firms in the introduction and decline stage of life cycle commanding a premium as high as 53 bps and 97 bps over mature phase firms, respectively. The results are also robust to the use of CDS spreads with different maturities, inclusion of alternative explanatory variables in models of default risk, and different model specifications. Our paper contributes to the accounting literature in three ways. First, we demonstrate how accounting information can be used as an input to debt market security pricing. This bridges 1 A good summary of credit protection contracts and the CDS mechanism is given in Hillscher et al (2015). 3

4 the literature discussed above in cash flow models, life cycle and firm risk. Second, we extend the literature identifying the determinants of CDS spreads by demonstrating the impact of life cycle as an additional important explanatory variable of credit spreads. Furthermore, we show that life cycle provides incremental information for pricing the risk of default not captured by the firm s age. Finally, we present evidence that firm life cycle is linked to the CDS term structure. The remainder of the paper is structured as follows. We present a review of the literature and hypothesis development in Section 2. Section 3 describes the data and methodology. In Section 4 the results are reported and analysed, and Section 5 concludes. 2. Literature Review and Hypothesis Development 2.1 Life cycle identification and classification While firm age has been widely used to identify life cycle stages, recent literature highlights the potential shortcomings of this classification method. First, firms represent portfolios of products and operate across industries, each of which may be experiencing a different level of maturity. In contrast to the Darwinian concept of monotonic aging, firms can evolve in a cyclical manner and reinvent themselves multiple times in order to maintain a prime life cycle stage. Research and development, capital expenditure, and merger & acquisition activity are all means towards the ultimate goal of maintaining firm characteristics where returns are maximized and cost of capital is optimized. Other firms that are formed with bright prospects but unable to commercialize their product or service may never reach maturity. For these firms, a decline stage ensues, followed by a potential delisting or a takeover. Accordingly, the common approach of classifying these firms as young based on their age since listing is misleading, as they may already be in a decline. Second, age since listing does not consider that firms may exist for different periods of time prior to becoming publicly listed. In addition, 4

5 the impact of challenging economic conditions or suboptimal managerial decisions are less pronounced for large firms, as they are better positioned to withstand adversity. In this regard, firms size introduces survivorship bias if mature firms are identified based on age since listing, and in turn older firms tend to be larger. Finally, reaching a prime life cycle stage is expected to vary systematically across industries, and the ability to maintain sustained growth or maturity will be influenced by the firm's scope of activities. This industry effect is likely to result in a cross-sectional bias in age across different industries. Studies in the firm life cycle literature have traditionally used ranking methods to allocate companies into different life cycle stages without necessarily defining key characteristics of a specific stage. Ranking variables used by Anthony and Ramesh (1992) and DeAngelo, DeAngelo and Stulz (2006) include dividend payouts, sales growth, capital expenditure scaled by firm value, market-to-book ratio, age and abnormal stock returns as life cycle stage proxies. A shortcoming of this approach is that ranking does not always capture essential internal and external factors that pivot firms from one stage to the next; they merely represent their relative standing compared to other companies in the given sample. Dickinson (2011) developed a theoretically more robust approach using life cycle theory as her starting point by examining firms' most likely cash flow characteristics in each phase. We adopt the classification proposed by Dickinson (2011) and use cash flow patterns from operating (CFO), investing (CFI) and financing (CFF) activities to identify differential behavior in the persistence and convergence patterns of profitability which are used to proxy the life cycle. By noting the sign of these cash flows, firms are mapped to the five life cycle stages; introduction, growth, mature, transition, and decline, defined by Gort and Klepper (1982). 5

6 In the introduction phase, CFO and CFI are both less than zero and CFF is greater than zero. Companies in the introduction stage of their development often experience negative operating cash flows due to the inconsistency of their revenues and uncertainty about their cost structure (Jovanovic, 1982). They deploy capital to develop production capacity and acquire long-term assets that are recorded as negative investment cash flows. Lacking consistent operating cash flows, these firms need sources of financing to grow and will access first debt and then capital markets as sources of funding (Barclay and Smith, 2005; Diamond, 1991; Myers, 1977, 1984); the net impact in this stage is positive financing cash flows. In the growth phase, CFI is less than zero and both CFO and CFF are greater than zero. As firms enter a growth phase, those with proven business models and a steady customer base start producing consistent positive operating cash flows (Wernerfelt, 1985). Driven by this optimism, they will continue investment activity, scale production, and work towards achieving economies of scale in order to keep competitors at bay. The sources of this new investment are positive net operating cash flows and cash from financing activities. Pecking order theory suggests that firms prefer internal to external financing and will access debt before they issue equity in order to benefit from the tax shield of interest payments, balanced with the risk of over-borrowing (Diamond 1991; Myers, 1984). In the mature phase, CFO is greater than zero and CFI and CFF are both less than zero. Firms continue to enjoy positive cash flows in the mature life cycle stage and benefit from certainty in cost structure and potential revenues (Wernerfelt, 1985). While they are likely to continue their investing activities to maintain assets, mature firms, by definition, have access to fewer positive net present value projects that would warrant external financing and hence net cash from financing activities will become negative. Firms in this stage shift focus towards retiring debt, paying dividends or engaging in share buybacks. 6

7 In the decline stage, CFO is less than zero, CFI is greater than zero and CFF can be either positive or negative. A firm in its decline stage experiences weakening growth rates and deteriorating pricing power. These symptoms can stem from product obsolescence or increased competition and result in negative operating cash flows (Wernerfelt, 1985). In order to service debt or re-purchase shares, firms may need to liquidate assets and investing cash flows will be positive for the first time since the company s inception (Barclay and Smith, 2005). In this life cycle phase providers of capital may receive payments, loans may be renegotiated, or new preferred equity instruments could be issued. Financing cash flows may be either positive or negative depending on the net impact of the two factors. The three remaining combinations of cash flows that the literature does not specifically label are collectively grouped into the transition phase. In this phase, either all cash flows are positive; all are negative; or positive operating and investing cash flows are combined with negative financing cash flows. Firms may still be generating positive operating cash flows, but it is possible that they become negative at this stage (Wernerfelt, 1985). Dickinson (2011) calls this the shake-out stage and maps to Gort and Klepper (1982) Stage 4, characterised by the non-equilibrium phase of negative net market entry by firms and potential structural changes in the industry. This classification system is applied in other recent literature to examine the firm's cost of equity capital (Hasan et al., 2015), corporate restructuring activity (Koh et al., 2015), and other financing decisions (Keasey, Martinez and Pindado, 2015; Faff et al., 2016). Additional markers for life-cycle stage have been suggested to include age, as a sign of maturity (DeAngelo, DeAngelo and Stulz, 2010); earned-to-contributed capital ratio (DeAngelo, DeAngelo and Stulz, 2006), and asset growth (Grullon, Michaely and Swaminathan, 2002). 7

8 2.2 Determinants of Credit spreads One of the first economic insights into the drivers of credit spreads is provided by Merton (1974). Using the contingent claim framework developed by Black and Scholes (1973), he shows that the value of each claim can be determined by the dynamics of firm value, which although unobservable, is a function of the firm leverage, the volatility of the underlying assets and the riskless spot rate. The Distance to Default (DtD) measure implied by the model underpins many commercially successful structural models of default, including CreditGrades and Moody s KMV. Subsequent literature shows that these structural determinants are insufficient in pricing the risk of default, with CDS spreads also driven by more systematic factors including macroeconomic variables and market volatility (Collin-Durfesne, Goldstein and Martin, 2001; Alexander and Kaeck, 2008; Ericsson, Jacob and Oviedo, 2009). More recently, Tang and Yan (2010) and Galil et al. (2014) show that the majority of spread changes are explained by firmlevel determinants but market variables still have significant explanatory power after controlling for firm-specific variables. Hadlock and Pierce (2010) suggest the use of firm size and age as useful predictors of financial constraint but indicate that this relationship appears to level off, advocating the addition of a quadratic component. The relative importance of covariates is also a function of the sampling frequency. Das, Hanouna and Sarin (2009) highlight that at the quarterly sampling frequency, market-based structural models perform similarly to accounting-based models (notably, Altman (1968) and Ohlson (1980)) based on accounting ratios, and both sources of information are complementary in pricing credit risk. Given our focus on firm s life cycle as additional previously ignored component of CDS spread which is calculated quarterly using accounting information, our model is inevitably a mix of both classes of models. 8

9 3. Data and Methodology 3.1 Data and Descriptive Statistics Our sample period covers a period from July 2007 to December The CDS spreads are provided by Markit and consist of end of the day mid-point quotes. We predominantly focus on US single name senior unsecured contracts with 5 year maturity as they offer the highest liquidity (Ericsson, Jacobs and Oviedo, 2009). Accounting information and credit ratings assigned by Standard & Poor s (S&P) are obtained from Compustat. Prices are downloaded from the Center for Research in Security Prices (CRSP). One year ahead Distance to Default calculations for individual firms are sourced from the National University of Singapore, Risk Management Institute, CRI database. 2 US Government Bond spreads and the VIX index are sourced from Bloomberg. Following existing studies we exclude all financial firms from the sample as the calculated metrics differ significantly from other sectors. All continuous data is winsorized at the 1% and 99% percentiles to minimise the effect of outliers. The final sample consists of 11,179 firm-quarter observations across 435 unique firms. Table 1 illustrates a fairly uniform distribution of forms across the sample period. Table 2 depicts the summary statistics of the variables. The dataset spans the global financial crisis characterized by high volatility of CDS spreads ranging from 19 bps to 1438 bps with a mean (median) of 187 bps (104 bps). The average age of the firms in the sample is 41 years as CDS spreads are typically written on larger, more mature firms. The majority of the firms are investment grade as rated by S&P. The mean Distance-to-Default (DtD) is 5.7 indicating that there is on average 5.7 standard deviations between the expected asset value one year ahead and the liability threshold. The average annualized expectation of volatility over the sample period implied by S&P 500 index options (VIX Index) is 21.7%. Summary 2 Available at: [Accessed 9 th January, 2018] 9

10 statistics for firm specific variables across the five life cycles reported in Table 3 show that in the introduction phase, the average CDS spread is 305 bps. CDS spreads then gradually decline as firms enter the growth phase and bottom out at 163 bps once firms reach maturity. The CDS spreads then rise again through the transition phase to reach a peak of 344 bps when firms migrate into the decline stage of the life cycle. Firm size, leverage, credit rating and DtD show a corresponding increase in the riskiness of firms in the early and late stages of the life cycle. Firms in the introduction and decline stage of the life cycle are typically smaller, more leveraged, non-investment grade and exhibit smaller distances to default than firms in the growth, mature and transition stage of the life cycle. Figure 1 confirms the U-shape relationship between the five stages of the firm s life cycle and both the mean and median 5-year CDS spreads. Hasan et. al. (2015) observes a similar pattern between cost of equity and life cycle, with the lowest cost of equity corresponding to firms in the mature stage of the life cycle. Table 4 presents correlations between CDS spread and the continuous explanatory variables. As expected, spreads are negatively related to firm size, age and DtD as riskier firms are smaller, younger and closer to default. Conversely, spreads are positively related to leverage and the overall equity market volatility. All coefficients have the expected signs. The highest correlations with the dependent variable are for distance to default, followed by leverage with coefficient values of and 0.39, respectively. 3.2 Regression Analysis Like many previous studies including Avramov, Jostova and Philipov (2007) and Ericsson et al. (2009), we undertake a linear regression analysis on the relationship between the level of CDS spreads and key variables suggested by literature. As well as previously analyzed 10

11 variables, we include five stages of the firm s life cycle in the regression. Formally, our regression equation is specified as follows.,,,,,,,,, where Spread is the 5-year corporate CDS spread. Cycle is an indicator variable denoting the five life cycles of the firm, namely, introduction, growth, mature, transition and decline. Size is measured as the log of total assets of the firm and Age is the number of years since the company was incorporated. Lev is the firm s leverage expressed as total net debt/ total assets. Rating is an indicator variable equal to one if the firm is assigned an investment grade rating by S&P. DtD is the distance to default risk measure proposed by Merton (1974) which relates the firm s net worth to its volatility. We include two macroeconomic variables commonly associated with credit risk. The 5-year US Government Bond yield and the annualized expectation of volatility from the S&P 500 index captured by the VIX index. We also include dummy variables for year and industry. 4. Empirical Results 4.1 CDS spread Level The multivariate OLS regressions for all the variables except for the year and industry dummies are shown in Table 5. Model 1 depicts the results without the addition of life cycle dummies, model 2 reports life cycle dummies but omits age, while model 3 reports all variables. Model 4 does not include year or industry dummies. The results show that both age and the firms life cycle are important determinants of credit spreads. We find that the U-shape relationship between CDS spreads and life cycle stages of firms still holds, even after controlling for firm specific characteristics and macroeconomic factors commonly quoted in 11

12 literature. In model 3, firms in the introduction phase and the decline phase trade at a 27 bps and 47 bps premium to firms in the mature stage of the life cycle, respectively. The difference for both stages is significant at the 1% level. Firms in the growth phase and the decline phase do not have significantly different spreads to firms in the mature stage of the life cycle. Without year and industry dummies (model 4), the difference between firms in the introduction phase and mature phase rises to 30 bps while the difference between the declined phase and mature phase increases to 53 bps. Both differences are still significant at the 1% level. In model 4, the growth and transition phase also commands an 8bp discount and a 13 bps premium to mature firms, respectively. Both differences are significant at the 5% level. Firms in the growth phase thus have the lowest average spreads overall. The observed signs on all control variable coefficients are as predicted by structural default probability models and are statistically significant at the 5% level or better. 4.2 Terms structure of CDS spreads Next, we examine if the life cycle is an important determinant of CDS spreads across the entire term structure of CDS spreads by extending the analysis to 1-year, 2-year and 10-year spreads. Figure 2 confirms the U-shape pattern between CDS spreads and firm life cycle with firms in the introduction and decline phase of the life cycle demanding a premium to firms in the mature phase of the life cycle. Notably, the U-shape is more pronounced for longer dated contracts indicating firms in the introduction and decline phase of the life cycle have a steeper term structure. A steeper CDS slope may be indicative of an expectation of a deterioration of firm s credit quality and subsequent increase in CDS spreads in the future. Table 6 shows that 1-year CDS spreads for firms in the introduction phase of the life cycle are 16 bps higher than spreads for firms in the mature phase of the life cycle and this difference increase to 27 bps for 10-year CDS spreads. For firms in the decline stage, 1-year spread are 42 bps higher, increasing 12

13 to 47 bps for 5-year spread and then decreasing back down to 42 bps for 10-year spreads. All coefficients are statistically significant. Interestingly, the explanatory power of the model increases with maturity, with the R 2 rising from for 1-year spreads up to for 10- year spreads. The explanatory power of the model is similar to that of Avramov et al. (2007) using corporate bond spreads. 4.3 Subsample analysis Given that the sample period spans the global financial crisis (GFC), we next split the original sample using 5-year CDS spreads into two periods from 2007 to 2010 and from 2011 to 2016 to isolate the impact of the GFC. Alexander and Kaeck (2008) indicate that CDS spreads may behave quite differently during volatile periods with the importance of specific covariates is highly dependent on the prevailing economic conditions. The results of the split sample are depicted in Table 7. Models 1 and 2 are estimated with year and industry dummies (omitted from table) while Models 3 and 4 are not. The results show that the U-shape relationship between CDS spreads and firm life cycle is much more pronounced during the GFC with firms in the introduction and decline phase exhibiting a much higher premium than in the 2011 to 2016 period. During the GFC the difference in spreads between firms in the introduction (decline) phase and mature phase of life cycle range from 53 bps to 55 bps (96 bps 107 bps), depending on the model specification. Both differences are statistically significant. Post GFC we still observe the U-shape pattern documented above but the difference between the life cycle stage stages is not as large, with firms in the introduction (decline) phase of the life cycle commanding a premium over the mature stage firms of between 18 bps and 20 bps (42 bps and 45 bps), depending on the model specification. The statistical significance of the difference is also slightly lower. In terms of control variables, all have the expected sign except age which is not 13

14 significant post-gfc crisis and size, which is positive and significant, indicating that during the crisis larger firms exhibit higher CDS spreads. While this may seem counterintuitive, the result is consistent with Clarke, Cull and Kisunko (2012) who argue that the crisis affected the availability of external finance that large more established firms relied on to a greater extent than small firms. 5. Conclusion We employ a dataset of single-name US corporate CDS spreads to investigate the relationship between CDS spread levels and their theoretical determinants. Besides regressing spreads on firm-specific and macro-based variables identified by literature, we examine the previous unexplored relation between credit spreads and firm s life cycle. Using the life cycle classification proposed by Dickinson (2011) we find a large variation in CDS spreads across the firm life cycle. We observe a distinct U-shape pattern between CDS spreads and the firm life cycle measure, with firms in the introduction and decline stages of life cycle characterised by significantly higher spreads of 30 bps and 53 bps, respectively, relative to firms in the mature stage of the life cycle. The results are statistically significant, persistent across time and CDS maturity, and robust to firm-specific and macro-level controls. Our results demonstrate that corporate life cycle provides valuable incremental information that may be applied to credit risk assessment. Importantly, the contribution of this additional source of information in assessing firm risk allows for much more complex non-monotonic dynamics of firm life cycle than simplistic measures of firm age 14

15 REFERENCES Altman, E., 1968, Financial ratios, discriminant analysis and the prediction of corporate bankruptcy, Journal of Finance vol. 23, no. 4, pp Anthony, J. H., Ramesh, K., 1992, Association between accounting performance measures and stock prices: A test of the life cycle hypothesis, Journal of Accounting and Economics vol. 15, pp Arikan, A., and Stulz, R., 2016, Corporate acquisitions, diversification, and the firm s life cycle, Journal of Finance vol. 71, no. 1, pp Avramov, D., Jostova G., and Philipov, A., 2007, Understanding changes in corporate credit spreads, Financial Analysts Journal vol. 63, no. 2, pp Barclay, M. J., Smith, C. W., 2005, The capital structure puzzle: The evidence revisited, Journal of Applied Corporate Finance vol. 17, pp Blanco, R., Brennan, S., and Marsh, I., 2005, An empirical analysis of the dynamic relationship between investment-grade bonds and credit default swaps, Journal of Finance vol. 60, no. 5, pp Chen, L., Lesmond, D., and Wei, J., 2007, Corporate yield spreads and bond liquidity, Journal of Finance vol. 62, no. 1, pp Clarke G. R. C., Cull, R., and Kisunko, G., 2012, External finance and firm survival in the aftermath of the crisis: Evidence from Eastern Europe and Central Asia, Journal of Comparative Economics vol. 40, pp Collin-Dufresne, P., Goldstein, R., and Martin, J. S., 2001, The determinants of credit spread changes, Journal of Finance vol. 56, no. 6, pp Das, S. R., Hanouna, P. and Sarin, A., 2009, Accounting-based versus market-based cross-sectional models of CDS spreads, Journal of Banking & Finance vol. 33, pp DeAngelo, H., DeAngelo, L., Stulz, R. M., 2006, Dividend policy and the earned/contributed capital mix: a test of the life-cycle theory, Journal of Financial Economics vol. 81, pp Diamond, D. W., 1991, Monitoring and reputation: The choice between bank loans and directly placed debt, Journal of Political Economy vol. 99, pp Dickinson, V., 2011, Cash flow patterns as a proxy for firm life cycle, Accounting Review vol. 86, no. 6, pp Douglas, A. V. S., Huang, A. G., and Vetzal, K. R., 2016, Cash flow volatility and corporate bond yield spreads, Review of Quantitative Finance and Accounting 46(2), pp Ericsson, J., Jacobs, K., and Oviedo, R., 2009, The determinants of credit default swap premia, Journal of Financial and Quantitative Analysis vol. 44, no. 1, pp Faff, R., Kwok, W. C., Podolski, E. J., Wong, G., 2016, Do corporate policies follow a life-cycle? Journal of Banking & Finance vol. 69, pp Friewald, N., Jankowitsch, and Subrahmanyam, M., 2012, Illiquidity or credit deterioration: A study of liquidity in the US corporate bond market during financial crises, Journal of Financial Economics vol. 105, no. 1, pp Galil, K., Shapir, O. M., Amiram, D., Ben-Zion, U., 2014, The determinants of CDS spreads, Journal of Banking & Finance vol.41, pp

16 Gkougkousi, X., 2014, Aggregate earnings and corporate bond markets, Journal of Accounting Research 52(1), pp Gort, M., Klepper, S., 1982, Time Paths in the Diffusion of Product Innovations, The Economic Journal vol. 92, pp Grullon, G., Michaely, R., Swaminathan, B., 2002, Are dividend changes a sign of firm maturity? The Journal of Business vol. 75, pp Hadlock C. J. and Pierce, J. R., 2010, New evidence on measuring financial constraints: moving beyond the KZ Index, The Review of Financial Studies vol. 23, no. 5, pp Hasan, M., Hossain, M., Cheung, A., and Habib, A., 2015, Corporate life cycle and cost of equity capital, Journal of Contemporary Accounting and Economic vol. 11, pp Hillscher, J., Pollett, J., and Wilson, M., 2015, Are credit default swaps a sideshow? Evidence that information flows from equity to CDS markets, Journal of Financial and Quantitative Analysis vol. 50, no. 3, pp Jovanovic, B., 1982, Selection and the evolution of industry, Econometrica vol. 50, pp Keasey, K., Martinez, B., Pindado, J., 2015, Young family firms: Financing decisions and the willingness to dilute control, Journal of Corporate Finance vol. 34, pp Koh, S., Durand, R., Dai, L., and Chang, M., 2015, Financial distress: Lifecycle and corporate restructuring, Journal of Corporate Finance 33, pp Leland, H., 1994, Corporate debt value, bond covenants, and optimal capital structure, Journal of Finance vol. 49, no. 4, pp Longstaff, F., Mithal, S., and Neis, E., 2005, Corporate yield spreads: Default risk or liquidity? New evidence from the credit default swap market, Journal of Finance vol. 60, no. 5, pp Lu, C. Chen, T, and Liao, H., 2010, Information uncertainty, information asymmetry and corporate bond yield spreads, Journal of Banking & Finance vol. 34, no. 9, pp Merton, R. C., 1974, On the pricing of corporate debt: the risk structure of interest rates, Journal of Finance vol. 29, no. 2, pp Myers, S. C., 1977, Determinants of corporate borrowing, Journal of Financial Economics vol. 5, pp Myers, S. C., 1984, The capital structure puzzle, Journal of Finance vol. 39, pp Ohlson, J. A., 1980, Financial ratios and the probabilistic prediction of bankruptcy. Journal of Accounting Research vol. 18, pp Pittman, J., and Fortin, S., 2004, Auditor choice and the cost of debt capital for newly public firms, Journal of Accounting and Economics vol. 37, no. 1, pp Tang, D. Y., and Yan, H., 2010, Market conditions, default risk and credit spreads, Journal of Banking and Finance 34(4), pp Wernerfelt, B., 1985, The dynamics of prices and market shares over the product life cycle, Management Science vol. 31, pp

17 Figure 1. Mean and median 5-year CDS spreads across the five life cycles 5-year CDS spread (bps) Mean Median Life Cycle 17

18 Figure 2. Mean CDS spreads across the CDS term structure and five life cycles CDS Spread (bps) year 2-year 5-year 10-year Life Cycle 18

19 Table 1. Distribution of firms between 2007 and 2016 Year Frequency Percentile , , , , , , , , , Total 11,

20 Table 2. Summary statistics for all variables Variable Mean Median Std. Dev. Min Max 5-year CDS spread ln(total Assets) Age Leverage Investment Grade Distance to Default year US Bond Vix Index

21 Table 3. Summary statistics for firm specific variables across the five firm life cycles Life Cycle Variable Mean Median Std. Dev. 1. Introduction 5-year CDS spread Observations ln(total Assets) Age Leverage Investment Grade Distance to Default Growth 5-year CDS spread Observations ln(total Assets) Age Leverage Investment Grade Distance to Default Mature 5-year CDS spread Observations ln(total Assets) Age Leverage Investment Grade Distance to Default Transition 5-year CDS spread Observations ln(total Assets) Age Leverage Investment Grade Distance to Default Decline 5-year CDS spread Observations ln(total Assets) Age Leverage Investment Grade Distance to Default

22 Table 4. Correlation Coefficients CDS spread Size Age Leverage DtD US Bond yield CDS spread 1 Size Age Leverage DtD US Bond yield VIX VIX 22

23 Table 5. Multivariate OLS regression results Variables Model 1 Model 2 Model 3 Model 4 Introduction *** *** *** ( ) ( ) ( ) Growth * * ** (3.9049) (3.8905) (3.7683) Transition * ** (5.3688) (5.4295) (5.2402) Decline *** *** *** ( ) ( ) ( ) Size ** ** (1.7650) (1.6264) (1.7732) (1.6653) Age ** ** ** (0.3310) (0.3303) (0.3264) Age-Squared *** *** *** (0.0032) (0.0032) (0.0031) Leverage *** *** *** *** ( ) ( ) ( ) ( ) Investment Grade *** *** *** *** (5.2683) (5.3207) (5.2158) (5.1815) DtD *** *** *** *** (0.6145) (0.6128) (0.6144) (0.5775) US Bond 5-year yield *** *** *** *** (5.1632) (5.1587) (5.1535) (2.5343) VIX *** *** *** *** (0.3611) (0.3630) (0.3625) (0.2622) Constant *** *** *** *** ( ) ( ) ( ) ( ) Year Dummy Yes Yes Yes No Industry Dummy Yes Yes Yes No Observations 11,179 11,179 11,179 11,179 Adjusted R-squared Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 23

24 Table 6. Multivariate OLS regression results across the term structure of CDS spreads. (1) (2) (3) (4) Variables 1-year Spread 2-year Spread 5-year Spread 10-year Spread Introduction * * *** *** (9.7699) ( ) ( ) (9.0535) Growth * * * (3.5512) (3.7718) (3.8905) (3.4628) Transition (5.0979) (5.3581) (5.4295) (4.7297) Decline ** *** *** *** ( ) ( ) ( ) ( ) Size ** *** (1.6286) (1.7243) (1.7732) (1.5598) Age *** *** ** (0.3035) (0.3237) (0.3303) (0.2898) Age-Squared *** *** *** ** (0.0029) (0.0031) (0.0032) (0.0028) Leverage *** *** *** *** ( ) ( ) ( ) ( ) Investment Grade *** *** *** *** (4.8192) (5.0857) (5.2158) (4.6410) DtD *** *** *** *** (0.5086) (0.5603) (0.6144) (0.5634) US Bond 5-year yield *** *** *** *** (5.0303) (5.2409) (5.1535) (4.4821) VIX *** *** *** *** (0.3545) (0.3688) (0.3625) (0.3176) Constant *** *** *** *** ( ) ( ) ( ) ( ) Year Dummy Yes Yes Yes Yes Industry Dummy Yes Yes Yes Yes Observations 11,179 11,179 11,179 11,179 Adjusted R-squared Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 24

25 Table 7. Pre and post-gfc subsample analysis (1) (2) (3) (4) Variables Introduction *** * *** * ( ) ( ) ( ) ( ) Growth ** ** (7.9226) (3.5478) (7.6213) (3.4262) Transition ( ) (5.2114) ( ) (5.0682) Decline ** ** *** *** ( ) ( ) ( ) ( ) Size ** *** *** *** (3.9479) (1.3658) (3.7810) (1.2648) Age *** *** (0.6513) (0.2947) (0.6402) (0.2934) Age-Squared *** ** (0.0068) (0.0028) (0.0067) (0.0028) Leverage *** *** *** *** ( ) ( ) ( ) ( ) Investment Grade *** *** *** *** ( ) (4.7864) ( ) (4.6240) DtD *** *** *** *** (1.7002) (0.6005) (1.5479) (0.5847) US Bond 5-year yield *** *** *** *** (8.0110) (4.9044) (4.9674) (3.4430) VIX *** *** *** *** (0.5325) (0.4050) (0.4154) (0.3263) Constant *** *** *** *** ( ) ( ) ( ) ( ) Year Dummy Yes Yes No No Industry Dummy Yes Yes No No Observations 4,175 7,034 4,145 7,034 Adjusted R-squared Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 25