Three essays: Conflict risks and Investment, Terrorism and Information, and Conservatism in Accounting for Oil Prices

Transcription

1 Three essays: Conflict risks and Investment, Terrorism and Information, and Conservatism in Accounting for Oil Prices Ao Li Supervisory Panel: Professor Neil Fargher (Chair) Associate Professor Janet Lee Dr Lijuan (Lily) Zhang

2 2 Overview Three-paper Thesis Study 1: The Influence of Conflict Risk on Investment Efficiency for Multinational Enterprises Study 2: The Impact of Terrorism on Financial Markets: Intra-day Evidence from Foreign Exchange Market Reactions to ISIS attacks Study 3: The Role of Product Market in Asymmetrically Timely Gain and Loss Recognition Evidence from the U.S. Oil and Gas Industry

3 Study 3 The Role of Product Market in Asymmetrically Timely Gain and Loss Recognition Evidence from the U.S. Oil and Gas Industry Ao Li 3

4 4 Research Question Do earnings respond asymmetrically to changes in product market prices? Specifically: Whether earnings exhibit an asymmetric pattern in timeliness of gain and loss recognition in regards to information reflected in lagged oil prices changes in O&G sector?

5 5 Contributions Interact of two strands of literature: accounting conservatism and energy economy Accounting earnings respond asymmetrically to product market prices Accounting reflects changes in product market in a delayed fashion Conservatism is of value to stakeholders who demand accounting conservatism, such as lenders and investors. Product market price shocks are able to signal potential sacrifices of earnings ahead of accounting-based measures. Provide opportunities to reduce downside risk and take early proactive actions.

6 6 Literature Review Accounting conservatism Unconditional conservatism vs. conditional conservatism (Beaver & Ryer 2005). Basu (1997): Bad news (negative stock returns) is reflected in a timelier manner than good news and earnings are more sensitive to bad news. Ball and Shivakumar (2006): Changes in operating cash flow indicators have an incremental asymmetric effect on earnings. Banker et al. (2017): Changes in sales to be a new major indicator for asymmetrical timeliness of earnings recognition.

7 7 Literature Review Energy economy Oil price shocks and macroeconomy Economic recessions are preceded to positive oil price shocks(e.g. Hamilton 1983; Gisser & Goodwin 1986; Cũnado & Perez de Gracia 2003). Affect GDP growth, inflation and currency exchange rate (Hamilton 2005; Brahmasrene et al. 2014). Oil prices and stock market At stock market-wide level, increase in oil prices negatively impact on stock market prices (e.g. Jones & Kaul 1996; Sadorsky 1999; Driesprong et al. 2008; Cuñado & Perez de Gracia 2014). O&G sector, positive association between oil prices and equity value (e.g. Faff & Brailsford 1999; Sadorsky 2001; El-Sharif et al. 2005; Nandha & Faff 2008).

8 Hypothesis Development Oil price shocks can be transmitted to gain and loss recognition via four pathways: 1. Oil prices predict stock returns (Driesprong et al. 2008; Narayan & Gupta 2004). Positively associated with equity value in O&G sector (Faff & Brailsford 1999; Sadorsky 2001; El-Sharif et al. 2005). U.S stock market s reaction to oil shocks can be completely accounted for by the impact of these shocks on real cash flow (Jones & Kaul 1996). 2. Oil prices changes have direct impact on O&G firms operating cash flows. Lower oil prices may reduce the viability of drilling, as operating cost may exceed revenue (Deloitte 2016). Lower oil prices reduce projects NPV. O&G firms may reduce or defer major expenditures. 8

9 9 Hypothesis Development 3. Sales revenue is determined by price and quantities. Higher price means higher sales revenue (steady demand for crude oil and refinery products in the market (Dunn & Holloway 2012)). Higher price suggest higher expected returns for exploration projects. 4. Use Fair values in accounting Asset write-downs and impairments based on the comparison of fair value and carrying amount Raise in prices are reflected in the value of revenues but not directly reflected in other asset values

10 10 Hypothesis Development Changes in Oil prices With high verification to recognise Good News (oil price increase) Bad News (oil price decrease) With low verification to recognise Stock returns Operating cash flows Sales Revenue Fair Value H1 Cash Flow Earnings Earnings with conditional conservatism

11 11 Hypothesis Development Earnings is thus expected to exhibit asymmetric association with oil price changes. H1a: After controlling for the asymmetric effect of stock returns, operating cash flow changes and sales changes, earnings exhibit asymmetric association with oil price changes. As sales revenues is determined by selling prices and quantities, revenues can be closely correlated with prices if quantities remain stable. H1b: After controlling for the asymmetric effect of stock returns and operating cash flow changes, earnings exhibit asymmetric association with oil price changes.

12 To test H1a (Extension of Banker et al. (2017)): Empirical Model EARN it = DR it + 2 RET it + 3 DR it RET it + β 1 DC it + β 2 ΔCF it + β 3 DC it ΔCF it + γ 1 DS it + γ 2 ΔSALES it + γ 3 DS it ΔSALES it + δ 1 DLW t + (4) δ 2 lag_wti t + δ 3 DLW t lag_wti t + ς it To test H1b (Extension of Ball & Shivakumar (2006)): EARN it = DR it + 2 RET it + 3 DR it RET it + β 1 DC it + β 2 ΔCF it + β 3 DC t ΔCF it + δ 1 DLW t + (5) δ 2 lag_wti t + δ 3 DLW t lag_wti t + ε it δ 3 is expected to be positive to indicate conditional conservatism of oil prices returns 12

13 13 Variable Definition EARN it = net income in year t lag_wti t = WTI oil price returns for the 12-month period of fiscal year t-1 DLW t = dummy variable that equals 1 if lagged WTI oil price return lag_wti t is negative, and 0 otherwise RET it = stock return for the 12-month period of fiscal year t DR it = dummy variable that equals 1 if stock return RET t is negative, and 0 otherwise ΔCF it = change in operating cash flow from year t-1 to year t DC it = dummy variable that equals 1 if cash flow change ΔCF t is negative, and 0 otherwise ΔSALES it = change in sales from year t-1 to year t DS it = dummy variable that equals 1 if cash flow change ΔCF t is negative, and 0 otherwise

14 14 Sample and Data U.S listed firms in Oil and Gas Sector Two-digit SIC =13 Sample period: Stock return data and financial report information CRSP/Compustat Merged Database Crude oil prices West Texas Intermediate (WTI), from U.S Energy Information Administration

15 15

16 16 $ (July 2008) $108 (June 2014) $46.27 (Dec 2008) $30 (Feb 2016)

17 17 Table 2 Descriptive Statistics Variable Mean S.D. MIN Q1 Median Q3 MAX EARN RET ΔCF ΔSALES lag_wti

18 18 Table 3 Correlation Matrix EARN 1 EARN RET ΔCF ΔSALES lag_wti RET * 1 ΔCF * * 1 ΔSALES * * * 1 lag_wti * * * 1

19 Results 19

20 20 Results F-Stat for the asymmetric effect of DLW, DLW Lag_WTI : ***

21 21 Additional Analysis Tangible assets write-downs WD it = DR it + 2 RET it + 3 DR it RET it + β 1 DC it + β 2 ΔCF it + β 3 DC t ΔCF it + δ 1 DLW t + (6) δ 2 lag_wti t + δ 3 DLW t lag_wti t + ε it Goodwill impairments GW it = DR it + 2 RET it + 3 DR it RET it + β 1 DC it + β 2 ΔCF it + β 3 DC t ΔCF it + δ 1 DLW t + (7) δ 2 lag_wti t + δ 3 DLW t lag_wti t + ε it

22 22

23 23 Sensitivity Analysis Main results are robust in tests of: Extending the Basu (1997) model by adding indicator of changes in oil prices Separate slope estimations for positive and negative news reflected in oil price Timing of oil price changes: current oil price returns vs. lagged oil price returns With additional control variables of economic factors and reporting factor

24 24 Summary of Results Earnings respond asymmetrically to lagged crude oil price changes after controlling for stock returns and changes in operating cash flows Support H1b bad news (negative oil price shocks) are recognised more fully and timelier than good news (positive oil price shocks) The asymmetric impact of product market price changes on earnings is subsumed in the effect of the sales changes. Goodwill impairments exhibits asymmetric association with product market price changes.

25 25 Limitations Examines a single industry Focuses on conditional conservatism O&G firms sensitivity to price changes