Core CFO and Future Performance Rodrigo S. Verdi Sloan School of Management Massachusetts Institute of Technology 50 Memorial Drive E52-403A Cambridge, MA 02142 rverdi@mit.edu Abstract This paper investigates the market reaction to (the core components of) cash flow from operations (CFO). Specifically, we decompose CFO into its core components - collections from customers, payments to suppliers, interest payments to creditors - and we investigate the predictive power of the core component of cash flow with respect to future accounting and stock return performance. Using a sample of 85,848 firm-year observations during 1988 to 2005 we find strong evidence that the information in Core CFO is able to predict future performance. The hedge portfolio size-adjusted returns in our sample equals 11.72% a year and the returns are positive in 16 of the 18 years in the sample. The abnormal returns are robust when the strategy is implemented among firms with larger market capitalizations and in the most recent years. Overall, our findings provide evidence that the information in cash flows, especially in the core components of CFO, is a strong predictor of future performance. January 2010
1. Introduction A growing literature has shown that components of earnings are able to predict future returns. Sloan (1996) shows that investors underreact to the cash flow component of earnings but overreact to the accruals component of earnings. He then emphasizes the overreaction to accruals and documents that a trading strategy based on accrual information is able to generate future abnormal returns. Specifically, a trading strategy buying firms in the lowest decile and selling firms in the highest decile of accruals generates abnormal returns above 10% per year. Our paper takes a different approach. Instead of focusing on accruals, we investigate the market underreaction to the cash flow component of earnings. In particular, we decompose cash flow from operations into its core components and investigate the predictive power of the core component of cash flow with respect to future performance. We follow Livnat and Zarowin (1990) to decompose cash flow from operations into its core components: collections from customers, payments to suppliers, and interest payments to creditors. We then define Core CFO as the cash flow from operations resulting from the collections from customers minus the payments to suppliers and interest payments to creditors. Livnat and Zarowin (1990) show that decomposing CFO into its core components enhances the degree of association between cash flow from operations and current returns. We extend their research by investigating the association between the core components of CFO and future returns. Our empirical design can be summarized as follows. First, for each firm-year observation from 1988 to 2005, we measure Core CFO. We then follow prior research and assign firms into deciles based on the magnitude of Core CFO in a given year and 1
investigate the future performance (accounting and stock returns) in each decile portfolio for the twelve-month period subsequent to the portfolio formation date (this date is defined as four months after the fiscal year end). In addition, we measure the returns of a hedge portfolio strategy that buys (sells) firms with high (low) Core CFO. Using a sample of 85,848 firm-year observations during 1988 to 2005, we find strong evidence that the information in Core CFO is able to predict future accounting and stock returns. Firms with high (low) Core CFO in a given year have high (low) Core CFO in the following year. In terms of stock returns, a hedge portfolio based on Core CFO earned 11.72% a year with positive returns in 16 out of the 18 years in the sample. The strategy profitability is robust when the strategy is implemented among firms with increasing market capitalization cut-offs suggesting that the strategy s profitability is not driven by small firms. For instance, when we restrict the analysis to firms with market capitalization above US$25 million and US$100 million, the abnormal returns increase to 12.61% and 12.45% per year, respectively. The results are also large even when we restrict the sample to firms with market capitalization above US$500 million and US$1000 million. However, the volatility of the portfolio increases considerably likely due to smaller portfolio size. Finally, we investigate the performance of the Core CFO strategy when implemented quarterly using trailing twelve-month data and find that the magnitude of the returns are largely robust to this alternative implementation criteria. Overall, our findings provide evidence that the information in cash flows from operations, especially in the core components of CFO, is able to predict future returns. The magnitude of the returns is similar or perhaps larger than those originally documented using accruals which is striking given the focus in the literature purely on 2
accruals. Our results suggest that the information in operating cash flows is equally important, and that by focusing on the core components of cash flows, one can enhance the predictability of the CFO model. This analysis is similar in spirit to Richardson et al. (2005) which investigate the predictive power of different components of accruals when compared to aggregate accruals. The remainder of the paper proceeds as follows. Section 2 describes the sample and the research design. Section 3 presents the main results. Section 4 concludes. 2. Methodology 2.1 Sample We obtain accounting variables for each firm from the CRSP-Compustat Merged Database for the fiscal years between 1988 and 2005. We require each firm to have a valid statement of cash flows (i.e., Compustat data item #318 = 7), which restricts the sample period to post-1987. We then restrict the firms to those with ordinary shares (CRSP share code beginning with 1) listed on NYSE, AMEX, and NASDAQ (CRSP exchange code 1, 2, and 3, respectively) at the beginning of the fourth month after the fiscal year end, which is the time when we form our portfolios. Consistent with previous literature (e.g., Sloan, 1996), financial firms (i.e., SIC codes in the 6000-6999 range) are excluded because of the different nature of these firms. The final sample after computing the main variables described below consists of 85,848 firm-year observations during 1988 to 2005. 2.2 Computing Core CFO 3
4
3. Results Table 1 presents a description of our sample. The sample consists of 85,848 firmyear observations during the fiscal years of 1988 to 2005 (with stock returns from May 1988 to April 2007). Panel A reports the number of firms in each year of our sample. The number of firms per year ranges from 3,489 firms in 1988 to 5,991 firms in 1997. Panel B provides the descriptive statistics of our key variables. The mean value for Core CFO is 4.42% meaning that, on average, Core CFO is approximately 4% of sales. However, this result is driven by a few outliers. For instance, the median Core CFO in the sample equals 29.05%, i.e., approximately 29% of sales. In the subsequent analyses we address these extreme values in two ways by assigning firms into deciles using ranked values of Core CFO and by creating sub-samples based on different market capitalizations. 5
3.1. Predictive power of Core CFO Table 2 Panel A presents one-year-ahead Core CFO for decile portfolios formed using current Core CFO. The second column presents the mean value of Core CFO during the year t across the deciles. By construction Core CFO in the current year increases across the deciles. For example, the mean Core CFO increases from -280% of sales for the firms in the bottom decile to 78% among firms in the top decile of Core CFO. The next column presents one-year ahead Core CFO. In this case, future Core CFO also increases across deciles suggesting that Core CFO is persistent such that firms with higher Core CFO in a given year are also more likely to have higher Core CFO in the following year. For instance, one-year-ahead Core CFO increases from -183% in the bottom decile to 55% in the top decile of firms based on current Core CFO. 2 Finally, the last column presents median one-year-ahead Core CFO across portfolios. Once again there is a monotonic increase across portfolios suggesting that the result is not driven by outliers. Overall the findings in Panel A present evidence that Core CFO is an important predictor of future Core CFO performance. Panel B presents future stock return performance across the portfolios. There is strong evidence that Core CFO predicts returns. For example, a hedge portfolio formed by buying firms in the top Core CFO decile and selling firms in the bottom Core CFO decile generates statistically significant raw (size-adjusted) abnormal returns of 11.44% (11.72%). In addition, the abnormal returns increase almost monotonically across portfolios. For example, the abnormal returns are -6.57% and -3.37% among firms in the bottom two deciles and +4.63% and +5.15% among firms in the top two deciles. Finally, 2 Untabulated analyses show that the pair-wise correlation between Core CFO in years t and t+1 equals 0.75 at the firm-level and 0.98 at the portfolio level. 6
the last column presents the Sharpe ratio (i.e., the ratio of mean monthly raw returns to its standard deviation) across the portfolios. The Sharpe ratios increase from -0.37 to 0.29 from the bottom to the top deciles. Further, the Sharpe ratio for the hedge portfolio equals 1.36, which is substantial when compared to Sharpe ratios around one advertised by hedge funds. Figure 1 presents the hedge portfolio returns in each year for the Core CFO strategy. The strategy earns positive abnormal returns in almost every year from 1988 to 2005 (the only exceptions are 1991 and 2004). Thus, the performance of the Core CFO strategy appears to be relatively stable during the last 18 years. Overall, the results in Table 2 and Figure 1 present evidence that firms with high Core CFO in a given year are likely to have high Core CFO and high stock returns in the following year. 3.2. Core CFO strategy across different market capitalization portfolios In this section, we investigate the performance of the Core CFO strategy across portfolios with different market capitalizations. This issue is important because there are concerns that many of the abnormal returns documented in prior literature require investors to trade in illiquid firms with high transaction costs. For instance, Lesmond, Schill, and Zhou (2004), Korajczyk and Sadka (2004), Hanna and Ready (2005) investigate the profitability of the momentum strategy after taking into account transaction costs. Likewise, Chordia et al. (2008) and Ng, Rusticus and Verdi (2008) investigate the impact of transaction costs on the post-earnings-announcement drift. 7
We proceed as follows. We first replicate the Core CFO strategy using the full sample. We then investigate the profitability of the strategy if an investor restricts the sample to portfolios with increasing market capitalization. In particular, we restrict the sample to firms with market capitalization higher than US$10 million, US$25 million, US$100 million, US$500 million, and US$1 billion. Table 3 presents the size-adjusted returns across the different portfolios. As discussed in table 2, the Core CFO strategy earns size-adjusted abnormal returns of 11.72% for the full sample. When we restrict the analysis to firms with market capitalization above US$10 million, the abnormal returns increase to 12.14%. Similarly, when we restrict the analysis to firms with market capitalization above US$25 million and US$100 million, the abnormal returns equal 12.61% and 12.45% respectively. These results are statistically significant with t-statistics above 5.00. Thus, it appears that the Core CFO strategy is not driven purely by small stocks and it remains relatively stable among middle and large firms. The last two columns present the abnormal returns among firms with market capitalization above US$500 million and US$1000 million. In this case the respective returns are 8.31% and 8.48%. Although these returns are still large, we note that the t- statistics reduce sharply when the strategy is implemented among firms with market capitalization of US$500 million or higher, suggesting that the returns to the strategy are more volatile among large firms. One possible reason for this result is the smaller sample size in the extreme portfolios when the strategy is implemented among large firms. For example, the number of firms in the extreme portfolios equals 30 firms on average during 8
the sample period but untabulated analysis show that the number of firms ranges from 9 firms in 1993 to 84 firms in 1999. Figure 2 presents yearly abnormal returns for the strategy when implemented across different portfolios. Consistent with findings in Table 3, the abnormal returns are largely stable when the strategy is implemented among all firms in the portfolio, or even among firms with market capitalization above US$100 million. However, when the strategy is implemented among firms with market capitalization of US$500 million or higher the abnormal returns are much more volatile. In particular, the strategy earned large positive abnormal returns from 1999 to 2002 but also earned negative returns in the mid-1990 s. 3.3. Core CFO strategy at the quarterly level In this section we evaluate the performance of the Core CFO strategy when implemented quarterly using trailing twelve month (hereafter TTM analysis) data. For the TTM analysis, we follow similar steps to those described in Section 3. In particular, we compute the core components of CFO every quarter for a given firm using data from the last four quarters (for example, for the 2 nd quarter of 1996 we use data from the 3 rd quarter of 1995 to the 2 nd quarter of 1996). We then implement the strategy quarterly and perform similar analysis described in Table 3 and Figure 2. Table 4 presents the TTM analyses. The Core CFO strategy, when implemented quarterly, earns size-adjusted abnormal returns of 11.42%, slightly smaller than the 11.72% achieved using yearly data. These returns are statistically significant. When we restrict the analysis to firms with market capitalization above US$10 million, the 9
abnormal return equals 12.29%. Likewise, when we restrict the analysis to firms with market capitalization above US$25 million and US$100 million, the abnormal returns stay relatively stable at 11.84% and 11.70% respectively. These results are statistically significant with t-statistics above 6.00. Thus, it appears that the Core CFO strategy is robust to quarterly implementation, and it remains relatively stable among mid-sized and large firms. Finally, the last two columns present the abnormal returns among firms with market capitalization above US$500 million and US$1000 million. In this case, the magnitude of the returns reduces considerably to 7.87%, consistent with findings in Table 3. Figure 3 presents quarterly abnormal returns for the strategy when implemented across different portfolios. Consistent with findings in Table 4, the abnormal returns are largely stable when the strategy is implemented among all firms in the portfolio or even among firms with market capitalization above US$100 million. For instance, the Core CFO strategy among all firms in the sample earns positive abnormal returns in most quarters with the exception of a few years such as 1999 and 2003. However, similar to the picture in Figure 2, the strategy appears to be more volatile when implemented among firms with market capitalization of US$500 million or higher. 3.4. Out-of-sample evidence In this section we provide some out-of-sample evidence of the Core CFO strategy. Fama (1998) discusses the importance of presenting out-of-sample evidence when one presents results inconsistent with market efficiency. When we first started this project, the last year with stock returns available in CRSP was 2007, so we restricted our 10
sample to financial statements ending between 1998 and 2005. However, since then the returns for the years 2008 and 2009 became available which allows us to provide some out-of-sample evidence of the Core CFO strategy when applied to the fiscal years of 2006 and 2007. Table 5 presents size-adjusted abnormal returns at the yearly level across different market capitalization portfolios for the fiscal years of 2006 and 2007. In this case we find that the Core CFO strategy earned size-adjusted abnormal returns of 18.06% in the whole sample, although these returns are mainly driven by an extreme negative performance of firms with negative Core CFO (returns of -14.68% for the bottom decile). When we restrict the analysis to firms with market capitalization above US$10 million and US$25 million, the abnormal returns are relatively stable at 18.50% and 17.98% respectively. Overall, the results in Table 5 present evidence that the Core CFO strategy also earned positive returns in the recent years. However, we note that the returns are mainly driven by the poor performance of firms with negative Core CFO. To the extent that short-sale constraints exist among these firms, the profits from the strategy will be reduced. 4. Conclusion This paper investigates the market reaction to the cash flow from operations. Specifically, we decompose cash flow from operations into its core components - collections from customers, payments to suppliers, and interest payments to creditors - and we investigate the predictive power of the core component of CFO with respect to 11
future performance. This choice is motivated by findings in Sloan (1996) who showed that investors underreact to the cash flow components of earnings and by Richardson et al. (2005) which investigate the predictive power of different components of accruals (but not of cash flow from operations). In addition, Livnat and Zarowin (1990) show that decomposing CFO into its components enhances the degree of association between cash flow from operations and current returns. We extend their research by investing the association between the components of CFO and future returns. Using a sample of 85,848 firm-year observations during 1988 to 2005 we find strong evidence that the information in Core CFO is able to predict future accounting and stock returns. Firms with high (low) Core CFO in a given year have high (low) Core CFO in the next year. In terms of stock returns, a hedge portfolio in our sample earns size-adjusted abnormal returns of 11.72% a year and is positive in 16 out of the 18 years in the sample. In addition, the abnormal returns are robust when the strategy is implemented among firms with increasing market capitalization cut-offs. For instance, when we restrict the analysis to firms with market capitalization above US$25 million and US$100 million, the abnormal returns increase to 12.61% and 12.45% respectively. Finally, we investigate the performance of the Core CFO strategy when implemented quarterly using trailing twelve-month data and find that the magnitude of the returns are largely robust. Overall, our findings provide evidence that the information in operating cash flows, especially in the core components of CFO, is able to predict future returns. These results contribute to the academic literature by showing that the predicted power of core CFO is similar or perhaps larger than those originally documented using accruals. In 12
addition, these findings have implications for analysts and investors. Analysts can use the information in core CFO to enhance its earnings predictability whereas investors can use this information to identify firms with predictable future returns. 13
References Bernard, V. L., and J. K. Thomas. Post-Earnings-Announcement Drift: Delayed Price Response or Risk Premium? Journal of Accounting Research 27 (1989): 1 48. Bernard, V. L., and J. K. Thomas. Evidence that Stock Prices Do Not Fully Reflect the Implications of Current Earnings for Future Earnings. Journal of Accounting and Economics 13 (1990): 305-340. Bradshaw, M., S. Richardson, and R. Sloan, 2006. The relation between corporate financing activities, analysts forecasts and stock returns. Journal of Accounting and Economics 42 (1-2), 53-85 Campbell, J., J. Hilscher, and J. Szilagyi, 2008. In Search of Distress Risk. Journal of Finance, forthcoming. Chordia, T.; A. Goyal; G. Sadka; R. Sadka; and L. Shivakumar. Liquidity and Post- Earnings-Announcement-Drift. Working paper, Emory University, 2008. Dichev, I., 1998. Is the risk of bankruptcy a systematic risk? Journal of Finance 53, 1141-148. Dickinson, V., 2007. Cash Flow Patterns as a Proxy for Firm Life Cycle. Working paper, University of Florida. Hanna, J. D., and M. J. Ready. Profitable Predictability in the Cross Section of Stock Returns. Journal of Financial Economics 78 (2005): 463-505. Korajczyk, R., and R. Sadka. Are Momentum Profits Robust to Trading Costs? Journal of Finance 59 (2004): 1039-1082. Lesmond, D. A.; M. J. Schill; and C. Zhou. The Illusory Nature of Momentum Profits. Journal of Financial Economics 71 (2004): 349-380. Livnat, J. and P. Zarowin. (1990). The incremental information content of cash-flow components. Journal of Accounting and Economics 13, 25-46. Ng, J., T. Rusticus, and R. Verdi, 2008. Implications of Transaction Costs for Post- Earnings-Announcement Drift with Jeffrey Ng and Tjomme Rusticus. Journal of Accounting Research 46 (3), 661-696. Ng, J., and R. Verdi, 2008. Firm Life Cycle, Cash Flows, and Long-Term Market Performance. Working paper, MIT. 14
Richardson, S., R. Sloan, M. Soliman, and İ. Tuna. Accrual reliability, earnings persistence and stock prices. Journal of Accounting and Economics 39 (2005): 437-485. Sloan, R. Do stock prices fully reflect information in accruals and cash flows about future earnings? The Accounting Review 71 (1996): 289 316. 15
TABLE 1 - Sample Description Panel A presents the number of firms in the sample for each year. The total sample consists of 85,848 firmyears with fiscal year between 1988 and 2005. Panel B presents descriptive statistics for the variables of interest for the full sample. Panel A: Frequency of observations per year Year # Firms Percent 1988 3,489 4.1 1989 4,359 5.1 1990 4,313 5.0 1991 4,349 5.1 1992 4,531 5.3 1993 4,948 5.8 1994 5,246 6.1 1995 5,477 6.4 1996 5,927 6.9 1997 5,991 7.0 1998 5,616 6.5 1999 5,407 6.3 2000 5,213 6.1 2001 4,702 5.5 2002 4,273 5.0 2003 4,035 4.7 2004 4,034 4.7 2005 3,938 4.6 Total 85,848 100.0 Panel B: Descriptive statistics Variable N Mean STD P5 Median P95 Core CFO t (%) 85,848 4.42 200.24-31.14 29.05 75.92 Ret t+1 (%) 85,848 14.44 88.03-73.51 1.94 137.29 Size Ret t+1 (%) 85,848-0.43 84.17-85.52-10.58 110.81 16
TABLE 2 - Core CFO and Future Performance This table presents future accounting and stock return performance across the portfolios sorted on Core CFO. Panel A presents one-year-ahead Core CFO whereas Panel B presents one-year-ahead stock returns across the portfolios. T-statistics are computed using the Fama-MacBeth (1973) methodology. The Sharpe ratio for each portfolio is computed as the ratio of mean monthly stock returns to its respective standard deviation. Panel A Operating performance Core CFO Decile Core CFO t Mean Core CFO t+1 Mean Core CFO t+1 Median 1-280.21-183.03-3.77 2 9.29 11.41 11.22 3 16.27 16.64 17.02 4 21.74 21.56 22.13 5 26.74 26.62 26.83 6 31.95 31.41 31.61 7 37.98 36.62 37.53 8 45.59 42.69 44.53 9 56.44 52.33 54.91 10 78.11 54.98 70.51 Panel B Return performance Core CFO Decile Ret t+1 (%) S-Ret t+1 (%) Sharpe Ratio 1 8.41-6.57-0.37 2 11.40-3.37-0.46 3 13.84-1.20-0.15 4 12.57-2.30-0.41 5 12.83-2.06-0.50 6 15.49 0.44 0.10 7 16.61 1.59 0.30 8 16.68 1.69 0.28 9 19.67 4.63 0.40 10 19.85 5.15 0.29 10-1 11.44 11.72 1.36 t-test 5.31 5.75 17
TABLE 3 - Analyses of Portfolio Returns Partitions by Market Cap Size-adjusted returns This table presents future stock return performance across the portfolios sorted on Core CFO. The portfolios have different cut-offs with increasing minimal market capitalization requirement. The returns are size-adjusted returns. T-statistics are computed using the Fama-MacBeth (1973) methodology. Core CFO All MVE > $10m MVE > $25m MVE > $100m MVE > $500m MVE > $1000m Decile Obs Size Ret t+1 Obs Size Ret t+1 Obs Size Ret t+1 Obs Size Ret t+1 Obs Size Ret t+1 Obs Size Ret t+1 1 476-6.57 405-7.37 331-8.13 183-8.73 59-4.31 30-3.73 2 477-3.37 422-3.48 360-2.70 239-2.87 117-2.08 75-2.20 3 477-1.20 425-2.18 368-1.77 250-0.60 129-0.58 84-0.39 4 477-2.30 425-1.97 367-1.91 255-1.42 139-0.89 91-0.65 5 477-2.06 425-1.74 370-1.81 264-0.93 144-0.25 98 0.22 6 477 0.44 422-0.39 364-0.48 256-1.29 133-2.11 90-1.91 7 477 1.59 418 1.39 364 0.91 257-0.01 137-0.40 91-1.15 8 477 1.69 416 1.62 362 1.26 255 1.16 136-0.44 92-0.44 9 477 4.63 422 3.42 366 2.97 263 0.99 136 0.23 92 0.90 10 477 5.15 414 4.77 362 4.48 253 3.72 121 4.00 83 4.75 10-1 476 11.72 404 12.14 331 12.61 183 12.45 59 8.31 30 8.48 t-test 5.75 5.56 6.06 5.42 1.90 1.73 18
TABLE 4 - Analyses of Portfolio Returns TTM Analysis Size-adjusted returns This table repeats the analyses in Table 3 with portfolios formed quarterly instead of annually. The portfolios are formed at the end of every calendar quarter with accounting data from the last 12-month period. As in Table 3, the portfolios have different cut-offs with increasing minimal market capitalization requirement. The returns are size-adjusted returns. T-statistics are computed using the Fama-MacBeth (1973) methodology. Core CFO All MVE > $10m MVE > $25m MVE > $100m MVE > $500m MVE > $1000m Decile Obs Size Ret t+1 Obs Size Ret t+1 Obs Size Ret t+1 Obs Size Ret t+1 Obs Size Ret t+1 Obs Size Ret t+1 1 442-4.93 370-6.70 298-6.84 159-7.75 44-3.85 22-2.87 2 443-1.38 390-1.66 329-1.39 212-1.50 104-0.83 66-0.47 3 443-1.72 397-2.19 340-1.68 229-1.17 114-1.55 74-2.14 4 443-1.51 397-1.79 344-1.65 238-1.20 127-0.82 85-0.65 5 443-0.51 396-0.83 346-1.14 245-1.07 133-0.61 90-0.27 6 443 0.34 395-0.43 343-0.50 241-0.53 126-0.54 86-0.43 7 443 1.88 392 1.58 339 1.20 238 0.22 124-0.97 83-0.99 8 443 3.54 389 2.84 338 2.54 236 1.22 126 0.40 87-0.36 9 443 4.13 390 3.55 340 2.73 242 1.27 126 0.92 85 0.26 10 442 6.49 386 5.59 335 5.00 235 3.95 119 4.01 81 5.00 10-1 442 11.42 369 12.29 298 11.84 159 11.70 44 7.87 22 7.87 t-test 6.99 7.19 6.55 7.12 3.69 3.26 19
TABLE 5 Out of sample evidence This table presents out-of-sample evidence of the future stock return performance across the portfolios sorted on Core CFO. The sample period is the fiscal years of 2006 and 2007 with stock returns from 2007-2009. The returns are size-adjusted returns. T-statistics are computed using the Fama-MacBeth (1973) methodology. Core CFO All MVE > $10m MVE > $25m Decile Obs Size Ret t+1 Obs Size Ret t+1 Obs Size Ret t+1 1 378-14.68 368-14.61 352-13.69 2 378-1.84 373-1.76 351-1.29 3 378-1.17 375-1.33 356-0.45 4 378-4.29 369-4.20 349-3.72 5 378-0.58 371-0.20 351 0.46 6 378-2.28 375-2.21 357-1.98 7 379-5.02 372-4.77 357-4.51 8 378-1.21 375-1.21 357-1.27 9 379 4.45 373 4.38 359 4.81 10 378 3.38 370 3.89 356 4.30 10-1 378 18.06 368 18.50 351 17.98 20
FIGURE 1 - Yearly abnormal returns This figure presents yearly size-adjusted stock returns for the hedge portfolio strategy based on Core CFO. The hedge portfolio buys (sells) firms in the top (bottom) Core CFO decile. 30 10-10 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 21
FIGURE 2 - Core CFO historical performance across different market capitalization This figure presents yearly size-adjusted stock returns for the hedge portfolio strategy based on Core CFO. The hedge portfolio buys (sells) firms in the top (bottom) Core CFO decile. The portfolios are identical to the ones presents in Table 3 and have different market capitalization requirements. All Firms Market Cap above $10M Market Cap above $25M Abret 40.00 30.00 20.00 10.00 0.00-10.00-20.00 1988 1990 1992 1994 1996 1998 2000 2002 2004 Abret 40.00 30.00 20.00 10.00 0.00-10.00-20.00 1988 1990 1992 1994 1996 1998 2000 2002 2004 Abret 40.00 30.00 20.00 10.00 0.00-10.00-20.00 1988 1990 1992 1994 1996 1998 2000 2002 2004 Market Cap above $100M Market Cap above $500M Market Cap above $1000M Abret 40.00 30.00 20.00 10.00 0.00-10.00-20.00 Abret 40.00 30.00 20.00 10.00 0.00-10.00-20.00 1988 1990 1992 1994 1996 1998 2000 2002 2004 Abret 40.00 30.00 20.00 10.00 0.00-10.00-20.00 1988 1990 1992 1994 1996 1998 2000 2002 2004
FIGURE 3 - Core CFO historical performance across different market capitalization TTM Analysis This figure presents yearly size-adjusted stock returns for the hedge portfolio strategy based on Core CFO implemented quarterly instead of annually. The hedge portfolio buys (sells) firms in the top (bottom) Core CFO decile. The portfolios are identical to the ones presents in Table 4 and have different market capitalization requirements. All firms Market Cap above $10M Market Cap above $25M Abret 40.00 20.00 0.00-20.00 Abret 40.00 20.00 0.00-20.00 Abret 40.00 20.00 0.00-20.00 Market Cap above $100M Market Cap above $500M Market Cap above $1000M Abret 40.00 20.00 0.00-20.00 Abret 40.00 20.00 0.00-20.00 Abret 40.00 20.00 0.00-20.00 23