Skip to main content

Unexpected distractions and investor attention to corporate announcements

Abstract

We investigate how unexpected distractions affect investor reactions to corporate earnings announcements. We use a daily news pressure (DNP) index as a proxy for the presence of potential investor distraction. Since breaking news captured by this index is largely unpredictable and unrelated to investors’ valuation decisions, our research design offers a unique opportunity to examine investor attention in the absence of strategic timing of announcements by managers. Using overall trading volume and Google searches as measures of investor attention, we find that investors are susceptible to distractions in their reactions to earnings announcements. We further find that DNP measures a form of distraction that affects retail but not institutional investors. Furthermore, in contrast to prior research that employs predictable measures of distraction, we find that price reactions to earnings announcements are not affected by unexpected distractions. Our results reveal that unexpected distractions reduce the attention of retail investors to earning announcements but do not necessarily lead to observable pricing effects.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2

Notes

  1. 1.

    DNP as a measure of investor distraction was initially used in an earlier version of this study titled “The Timing of Management Earnings Forecasts and Investor Inattention.” That draft was circulated as early as 2010 and is available from the authors upon request.

  2. 2.

    A tangential literature, including the work of Bushee and Friedman (2016) and deHaan et al. (2017), explores the impact of weather on investor behavior. Although seemingly related to our study, these papers differ critically in focus. In these studies, weather is not used as a measure of distraction but rather as a shock to individuals’ moods. In contrast, our goal is to understand the impact of distractions, independent of mood or sentiment, on investor attention. While DNP may exhibit a low degree of correlation with mood, an important feature of the measure is that it is not systematically related to investor mood. We offer more detail on this point in Section 3.1

  3. 3.

    The implementation of Regulation Fair Disclosure (“Reg FD”) in 2000 prompted firms to publicly announce in advance their earnings announcement dates. This generates the potential for other firms to incorporate this information into their own disclosure decisions and for investors to anticipate the announcements and adjust their attention accordingly. Consequently, it is plausible that NumEA was a more effective measure of investor distraction during the Hirshleifer et al. (2009) sample period but is now more susceptible to strategic timing.

  4. 4.

    Grossman and Stiglitz (1980) describe a long-run equilibrium where the absence of sufficient noise trading eventually leads to a reduction in costly information acquisition. However, in the case of an unexpected and temporary reduction in noise trading, informed investors may not change their information acquisition. Nonetheless, price efficiency may still be affected by such unexpected and temporary reductions in noise trading through their effect on liquidity provision.

  5. 5.

    The Vanderbilt Television News Archives contain evening news broadcasts from the major U.S. national television networks beginning August 5, 1968. We focus on ABC, CBS, and NBC broadcasts because they have retained the same format (i.e., 30 minutes aired between 5:30 pm-6:00 pm) over our sample period, 1995-2015. In contrast, CNN had varying news formats over our sample period, and Fox News is only available after 2004.

  6. 6.

    In the rare instances where the news broadcast deviates from the normal 30-minute format, typically when extraordinary events occur, we set the index to a missing value. Note, too, that our sample is limited to days when U.S. equity markets are open.

  7. 7.

    Peress and Schmidt (2020) examine the highest DNP for each year between 1995 and 2015, along with the main news event on that day. They observe that high values of the index typically coincide with major unexpected news events that seem unrelated to earnings announcements. For example, the highest DNP observation in 2012 (18.33) corresponds to the Sandy Hook Elementary School shooting on December 14, 2012.

  8. 8.

    Specifically, Peress and Schmidt (2020) show that DNP is not significantly correlated with the Baker et al. (2016) measure of economic policy uncertainty, the Aruoba et al. (2009) index of business activity, instances of Federal Open Market Committee meetings, or with releases of the Consumer Price Index or U.S. employment statistics from the Bureau of Labor Statistics. They find weak evidence of an association between DNP and the NYT sentiment index, supporting the view that DNP relates somewhat to mood. However, the economic magnitude of this relation is quite small. Even when used together, these six indices explain less than 10% of variation in DNP.

  9. 9.

    Our finding that managers disclose fewer losses on high NumEA days may initially seem counterintuitive, given the popular interpretation of high NumEA days as high distraction days. However, our results in Section 5.1 cast doubt on this view, as we find that high NumEA days are associated with more retail investor attention per unit of earnings news. To the extent that there is more attention on high NumEA days, it might benefit managers to strategically disclose losses on low NumEA days. More generally, we emphasize the systematic difference in loss incidence across levels of NumEA, as this illustrates the strategic timing concern that challenges the use of NumEA in studies of investor distraction around corporate disclosures.

  10. 10.

    In untabulated analyses, we also find that there is no significant difference in the proportion of negative management forecasts on high DNP days, relative to other disclosure days.

  11. 11.

    Da et al. (2011) justify the use of tickers, rather than full company names, by noting that tickers are “less ambiguous” and relatively more likely be used by investors interested in financial information than searches for full company names, which are often the result of searches for nonfinancial information. Nonetheless, deHaan et al. (2019) describe a specific form of measurement error that likely affects the Google SVI variable and outline several procedures to mitigate the impact of this error on statistical inferences. Following deHaan et al. (2019), we replicate our analyses using a reduced sample of firms whose tickers are less “noisy,” and thus Google SVI is less likely to suffer from measurement error. We use the two approaches suggested by deHaan et al. (2019) to identify these noisy tickers. First, we exclude all tickers with fewer than four characters, and we then exclude all tickers that appear on deHaan et al. (2019)’s list of common words. Using either of these reduced samples, we continue to find a statistically and economically significant negative association between DNP and the level of Google SVI. This reassures us that our main inferences are not attributable to measurement error in Google SVI.

  12. 12.

    Blankespoor et al. (2018) note that this approach offers low type I and high type II errors. In other words, using this approach we are unlikely to misclassify trades as retail but probably are not capturing the full extent of retail trading.

  13. 13.

    Ben-Rephael et al. (2017) provide several evidence for the claim that institutional investors are the primary users of Bloomberg terminals. First, they note that Bloomberg has approximately 325,000 terminal subscriptions that range in cost from $20,000 to $25,000 per year, a level that would be prohibitive for most non-institutional investors. Second, they examine the user profiles of terminal subscribers and find that approximately 80% of them work in financial industries. Academic users constitute less than 1% of the user base, and the remainder are largely Bloomberg employees.

  14. 14.

    Note that, in measuring institutional investor attention, we use a measure of investor demand for information rather than a measure based on trading volume. This is due to limitations on data availability. In particular, one cannot interpret the difference between overall trading volume and our measure of retail trading volume as a measure of institutional trading volume. As previously noted, ARVol offers low type I and high type II errors so that we are unlikely to misclassify trades as retail but probably are not capturing the full extent of retail trading. We believe that our approach of using retail trading volume and institutional investor searches provides the best strategy for minimizing type I error in identifying investor attention across these two groups. Moreover, recognizing that our measures are imperfect, we also bolster our analyses by including several controls and fixed effects.

  15. 15.

    Following the literature, in our IPT tests, we exclude observations with absolute CAR[0, 5] of less than 2% to reduce measurement noise due to a small denominator. Our inferences are the same if we do not apply this filter.

  16. 16.

    Our inferences are the same if, in Eqs. 2a and 2b, we use the levels of explanatory variables and not their decile ranks.

  17. 17.

    Specifically, Google search volume data is available starting in 2005, retail trading volume is available starting in 2003, and Bloomberg search volume is available starting in 2010.

  18. 18.

    We obtain the estimate -1.16 by calculating the following product: − 0.004 × 2.114 × 1.372. − 0.004 is the estimated coefficient on DNP, and 2.114 is the standard deviation of DNP; 1.372 is the mean raw (i.e., not log-transformed) value of ATVol in our sample (1.372 = e0.316)

  19. 19.

    We obtain the estimate -1.15 by calculating the following: (− 0.028 × 2.114 × 0.0539) ÷ 0.372. − 0.028 is the estimated coefficient on DNP×AbsSUE; 2.114 is the standard deviation of DNP; 0.0539 is the standard deviation of AbsSUE; 0.372 is the percentage increase in trading volume around earnings announcements, relative to the non-announcement periods.

  20. 20.

    Untabulated statistics show that the sum of the coefficients on DNP and DNP×SUE is also indistinguishable from zero. Hence Table 5 provides no evidence of different announcement- or post-announcement-period price reactions to earnings announcements that coincide with unexpected distractions.

  21. 21.

    We include Friday in all our previous estimations by inclusion of day-of-week fixed effects. The difference in Table 7 is that now we demonstrate the coefficient estimate on Friday and interact the Friday indicator with AbsSUE.

  22. 22.

    Our evidence pertains to unexpected distractions as reflected in DNP. Different relations may exist between alternative measures of distraction and market pricing.

References

  1. Admati, A. R. (1985). A noisy rational expectations equilibrium for multi-asset securities markets. Econometrica, 53(3), 629–657.

    Article  Google Scholar 

  2. Aruoba, S., Diebold, F., & Scotti, C. (2009). Real-Time Measurement of business conditions. Journal of Business and Economic Statistics, 27 (4), 417–427.

    Article  Google Scholar 

  3. Baker, S., Bloom, N., & Davis, S. (2016). Measuring economic policy uncertainty. Quarterly Journal of Economics, 131(4), 1593–1636.

    Article  Google Scholar 

  4. Bamber, L. S., Barron, O. E., & Stevens, D. E. (2011). Trading volume around earnings announcements and other financial reports: Theory, research design, empirical evidence, and directions for future research. Contemporary Accounting Research, 28(2), 431–471.

    Article  Google Scholar 

  5. Barber, B. M., & Odean, T. (2008). All that glitters: The effect of attention and news on the buying behavior of individual and institutional investors. Review of Financial Studies, 21(2), 785–818.

    Article  Google Scholar 

  6. Barth, M. E., Gee, K. H., Israeli, D., & Kasznik, R. (2021). Stock price management and share issuance: Evidence from equity warrants. The Accounting Review.

  7. Beaver, W. H. (1968). The information content of annual earnings announcements. Journal of Accounting Research, 6(1-2), 67.

    Article  Google Scholar 

  8. Ben-Rephael, A., Da, Z., & Israelsen, R. D. (2017). It depends on where you search: Institutional investor attention and underreaction to news. Review of Financial Studies, 30(9), 3009–3047.

    Article  Google Scholar 

  9. Beyer, A., Cohen, D. A., Lys, T. Z., & Walther, B. R. (2010). The financial reporting environment: Review of the recent literature. Journal of Accounting and Economics, 50(2), 296–343.

    Article  Google Scholar 

  10. Black, F. (1972). Capital market equilibrium with restricted borrowing. Journal of Business, 45(3), 444–455.

    Article  Google Scholar 

  11. Blankespoor, E., deHaan, E., & Zhu, C. (2018). Capital market effects of media synthesis and dissemination: evidence from robo-journalism. Review of Accounting Studies, 23(1), 1–36.

    Article  Google Scholar 

  12. Boehmer, E., Jones, C. M., Zhang, X., & X. Zhang. (2021). Tracking Retail Investor Activity. Journal of Finance.

  13. Bushee, B. J., & Friedman, H. L. (2016). Disclosure standards and the sensitivity of returns to mood. The Review of Financial Studies, 29(3), 787–822.

    Google Scholar 

  14. Butler, M., Kraft, A., & Weiss, I. S. (2007). The effect of reporting frequency on the timeliness of earnings: The cases of voluntary and mandatory interim reports. Journal of Accounting and Economics, 43(2-3), 181–217.

    Article  Google Scholar 

  15. Cheng, M., Hodder, L. D., & Watkins, J. (2017). Usefulness of Interest Income Sensitivity Disclosures. Working Paper.

  16. Chi, S. S., & Shanthikumar, D. M. (2017). Local bias in google search and the market response around earnings announcements. The Accounting Review, 92(4), 115–143.

    Article  Google Scholar 

  17. Da, Z., Engelberg, J., & Gao, P. (2011). In search of attention. Journal of Finance, 66(5), 1461–1499.

    Article  Google Scholar 

  18. deHaan, E., Lawrence, A., & Lijtens, R. (2019). Measurement Error in Dependent Variables in Accounting: Illustrations Using Google Ticker Search and Simulations. Working Paper.

  19. deHaan, E., Madsen, J., & Piotroski, J. D. (2017). Do Weather-Induced Moods Affect the Processing of Earnings News? Journal of Accounting Research, 55(3), 509–550.

    Article  Google Scholar 

  20. deHaan, E., Shevlin, T., & Thornock, J. (2015). Market (in)attention and the strategic scheduling and timing of earnings announcements. Journal of Accounting and Economics, 60(1), 36–55.

    Article  Google Scholar 

  21. DellaVigna, S., & Pollet, J. (2009). Investor inattention, firm reaction, and friday earnings announcements. Journal of Finance, 64(2), 709–749.

    Article  Google Scholar 

  22. Diamond, D. W., & Verrecchia, R. E. (1981). Information aggregation in a noisy rational expectations economy. Journal of Financial Economics, 9(3), 221–235.

    Article  Google Scholar 

  23. Doyle, J. T., & Magilke, M. J. (2009). The timing of earnings announcements: an examination of the strategic disclosure hypothesis. The Accounting Review, 84(1), 157–182.

    Article  Google Scholar 

  24. Drake, M. S., Gee, K. H., & Thornock, J. R. (2016). March market madness: The impact of Value-Irrelevant events on the market pricing of earnings news. Contemporary Accounting Research, 33(1), 172–203.

    Article  Google Scholar 

  25. Drake, M. S., Jennings, J., Roulstone, D. T., & Thornock, J. R. (2017). The comovement of investor attention. Management Science, 63(9), 2847–2867.

    Article  Google Scholar 

  26. Drake, M. S., Roulstone, D. T., & Thornock, J. R. (2012). Investor information demand: Evidence from google searches around earnings announcements. Journal of Accounting Research, 50(4), 1001–1040.

    Article  Google Scholar 

  27. Eisensee, T., & Stromberg, D. (2007). News droughts, news floods, and u. s. Disaster relief. Quarterly Journal of Economics, 122(2), 693–728.

    Article  Google Scholar 

  28. Gervais, S., Kaniel, R., & Mingelgrin, D. H. (2001). The high-Volume Return Premium. Journal of Finance, 56(3), 877–919.

    Article  Google Scholar 

  29. Grossman, S. J., & Stiglitz, J. E. (1980). On the impossibility of informationally efficient markets. The American Economic Review, 70(3), 393–408.

    Google Scholar 

  30. Hellwig, M. F. (1980). On the aggregation of information in competitive markets. Journal of Economic Theory, 22(3), 477–498.

    Article  Google Scholar 

  31. Hirshleifer, D., Lim, S. S., & Teoh, S. H. (2009). Driven to Distraction: Extraneous Events and Underreaction to Earnings News. Journal of Finance, 64(5), 2289–2325.

    Article  Google Scholar 

  32. Hirshleifer, D., & Teoh, S. H. (2003). Limited attention, information disclosure, and financial reporting. Journal of Accounting and Economics, 36(1-3), 337–386.

    Article  Google Scholar 

  33. Israeli, D., Kaniel, R., & Sridharan, S. A. (2021). The Real Side of the High-Volume Return Premium. Management Science.

  34. Israeli, D., Lee, C. M. C., & Sridharan, S. A. (2017). Is there a dark side to exchange traded funds? an information perspective. Review of Accounting Studies, 22(3), 1048–1083.

    Article  Google Scholar 

  35. Jagannathan, R., & Wang, Z. (1996). The conditional CAPM and the cross-section of expected returns. Journal of Finance, 51(1), 3–53.

    Article  Google Scholar 

  36. Johnson, T. L., & So, E. C. (2018). Time will tell: Information in the timing of scheduled earnings news. Journal of Financial and Quantitative Analysis, 53(6), 2431–2464.

    Article  Google Scholar 

  37. Kahneman, D. (1973). Attention and Effort. Englewood Cliffs NJ: Prentice-Hall.

    Google Scholar 

  38. Kasznik, R., & Kremer, I. (2014). The strategic timing of management earnings forecasts around scheduled releases of macroeconomic news. Working Paper.

  39. Koester, A., Lundholm, R., & Soliman, M. (2016). Attracting attention in a limited attention world: Exploring the causes and consequences of extreme positive earnings surprises. Management Science, 62(10), 2871–2896.

    Article  Google Scholar 

  40. Kyle, A. S. (1985). Continuous auctions and insider trading. Econometrica, 53(6), 1315–1335.

    Article  Google Scholar 

  41. Kyle, A. S. (1989). Informed speculation with imperfect competition. The Review of Economic Studies, 56(3), 317–355.

    Article  Google Scholar 

  42. Lintner, J. (1965). Security prices, risk, and maximal gains from diversification. Journal of Finance, 20(4), 587–615.

    Google Scholar 

  43. Livnat, J., & Mendenhall, R. (2006). Comparing the Post–Earnings announcement drift for surprises calculated from analyst and time series forecasts. Journal of Accounting Research, 44(1), 177–205.

    Article  Google Scholar 

  44. Melessa, S. (2013). Monthly Employment Reports and the Pricing of Firm-level News. Working Paper.

  45. Michaely, R., Rubin, A., & Vedrashko, A. (2016a). Are Friday announcements special? Overcoming selection bias. Journal of Financial Economics, 122(1), 65–85.

    Article  Google Scholar 

  46. Michaely, R., Rubin, A., & Vedrashko, A. (2016b). Further evidence on the strategic timing of earnings news: Joint analysis of weekdays and times of day. Journal of Accounting and Economics, 62(1), 24–45.

    Article  Google Scholar 

  47. Penman, S. (1980). An empirical investigation of voluntary disclosure of corporate earnings forecasts. Journal of Accounting Research, 18(1), 132–160.

    Article  Google Scholar 

  48. Peress, J., & Schmidt, D. (2020). Glued to the TV: Distracted retail investors and stock market liquidity. Journal of Finance, 75(2), 1083–1133.

    Article  Google Scholar 

  49. Rogers, J. L., & Buskirk, A. V. (2013). Bundled forecasts in empirical accounting research. Journal of Accounting and Economics, 55(1), 43–65.

    Article  Google Scholar 

  50. Ross, S. A. (1976). Options and efficiency. Quarterly Journal of Economics, 90(1), 75–89.

    Article  Google Scholar 

  51. Sharpe, W. F. (1964). Capital asset prices: a theory of market equilibrium under conditions of risk. Journal of Finance, 19(3), 425–442.

    Google Scholar 

  52. Twedt, B. (2016). Spreading the word: Price discovery and newswire dissemination of management earnings guidance. The Accounting Review, 91(1), 317–346.

    Article  Google Scholar 

  53. Verrecchia, R. E. (1982). Information acquisition in a noisy rational expectations economy. Econometrica, 50(6), 1415–1430.

    Article  Google Scholar 

  54. Verrecchia, R. E. (2001). Essays on disclosure. Journal of Accounting and Economics, 32(1-3), 97–180.

    Article  Google Scholar 

Download references

Acknowledgements

We appreciate helpful comments from Stephen Penman (editor), two anonymous reviewers, Henry Friedman, Darren Roulstone, Eric So, and seminar participants at Emory University, Nazarbayev University, Tel Aviv University, Tilburg University, University of Waterloo, and the 2018 Financial Accounting and Reporting Section (FARS) Midyear Meeting. We thank Mason Jiang for excellent research assistance and Christina Zhu for sharing her code on estimating retail trading volume.

Author information

Affiliations

Authors

Corresponding author

Correspondence to Ron Kasznik.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendices

Appendix : A: Daily news pressure index (DNP)

The daily news pressure (DNP) index captures the availability of news, measured as the median number of minutes across the main TV news broadcasts (ABC, CBS, and NBC) devoted to the first three news segments in a given day. To compile that, we obtain from the Vanderbilt Television News Archives a detailed list of all news segments broadcast by the three networks on their evening news program each day. We then count, for each network, the number of seconds spent on the first three segments (excluding commercials, anchor segments, and program introductions). The daily median number of seconds is then divided by 60 to derive a daily news pressure index. Because the duration of each program is 30 minutes, the news pressure index takes a value between (close to) 0 and 30.

To further illustrate the data underlying the calculation of DNP, Table 11 provides the detailed breakdown of evening news coverage on the three main news networks on August 2, 2007. The top three news segments for broadcast are in bold. The table reveals that ABC spent 13:10 minutes on “Minneapolis Bridge Collapse,” 2:50 minutes on “Infrastructure,” and 0:30 minutes on “Toy Recall” for a total of 990 seconds on its first three news segments. CBS spent 12:30 minutes, 2:50 minutes, and 1:30 minutes (total of 1,010 seconds) on its first three news segments, and NBC spent 13:10 minutes, 2:40 minutes, and 0:30 minutes (total of 980 seconds). Across the three networks, the median number of seconds spent on the first three news segments on August 2, 2007, was 990 seconds, or 16.5 minutes.

Table 11 Construction of DNP: example from August 2, 2007

In contrast, on August 8, 2007, ABC spent 300 seconds on “Heat Wave,” “Global Weather,” and “Mine Cave-in.” CBS spent 220 seconds on “Mine Cave-in,” “Minneapolis Bridge Collapse,” and “Wild Weather.” And NBC spent 640 seconds on “Heat Wave,” “Shuttle Endeavour Liftoff,” and “Utah Mine Collapse.” Thus, on August 8, 2007, the median number of seconds spent on the first three news segments across the three networks is 300 seconds or 5.0 minutes. We interpret this as an indication that the news pressure on August 2, 2007, is greater than on August 8, 2007.

Appendix : B: Variable definitions

Variable Description
S U E i,t Standardized unexpected earnings of firm i for quarter t. Measured as the difference between net income of quarter t and net income from four quarters ago, scaled by a firm i’s stock price at the end of quarter t
D N P t Average daily news pressure across days t and t + 1
A T V o l i,t Log of 1+ the ratio of average overall share turnover for firm i across days [0, 1] to the average overall daily turnover for firm i over days [-54, -5] minus an analogous amount for all firms traded on major stock exchanges in the United States, relative to firm i’s earnings announcement day t
G o o g l e i,t Log of the rank of Google search volume index (SVI) for firm i on days [0, 1] relative to the average SVI over days [-30, -1] of a firm’s earnings announcement day t
A R V o l i,t Log of 1 + the ratio of average retail share volume for firm i across days [0, 1] to the average retail share volume for firm i over days [-54, -5] minus an analogous amount for all firms traded on major stock exchanges in the United States, relative to firm i’s earnings announcement day t
A I A i,t Abnormal institutional attention from Bloomberg terminals, estimated following the procedure outlined by Ben-Rephael et al. (2017)
CAR[a, b]i,t Cumulative size and book-to-market-adjusted stock return for firm i from day a to day b relative to earnings announcement day t
I P T i,t Intraperiod timeliness of returns of firm i for quarter t. Defined as \({\sum }_{j=0}^{4}{\frac {CAR[0,j]}{CAR[0, 5]}} + 0.5\)
S i z e i,t Log of market value of equity for firm i on day t
B T M i,t Ratio of book value of equity to market value of equity for firm i on day t
A n a l y s t i,t Log of 1 + the number of analysts providing an earnings forecast for firm i during month t
I n s t O w n i,t Percentage of firm i’s shares owned by institutions at the most recent quarter-end, relative to day t
F r i d a y t Indicator equalling 1 if day t is a Friday and 0 otherwise
N C A A t Indicator equalling 1 if day t is an NCAA March Madness tournament day as defined by Drake et al. (2016) and 0 otherwise
N u m E A t Log of the number of earnings announcements made on day t
E A M F i,t Indicator equalling 1 if day t of firm i’s earnings announcement day coincides with a management forecast day
MF_Surpi,t Indicator equalling 1 if firm i’s management forecast is classified as negative or positive (i.e., guidance codes of 1 or 2 in the IBES database)
MF_Posi,t Indicator equalling 1 if firm i’s management forecast is classified as positive (i.e., a guidance code of 2 in the IBES database)
MF_Negi,t Indicator equalling 1 if firm i’s management forecast is classified as negative (i.e., a guidance code of 1 in the IBES database)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Israeli, D., Kasznik, R. & Sridharan, S.A. Unexpected distractions and investor attention to corporate announcements. Rev Account Stud (2021). https://doi.org/10.1007/s11142-021-09618-4

Download citation

Keywords

  • Investor attention
  • Earnings announcements
  • Retail trading
  • Distraction

JEL Classification

  • G12
  • G14
  • L20
  • M41