Title page 1

Contents 1

Abstract 5

1. Introduction 6

2. Literature Review 8

3. Hypotheses development 11

4. Models 13

5. Data 15

5.1. The Bank of Italy Survey of Industrial and Service Firms (INVID) 15

5.2. Balance sheet data and summary statistics 17

6. Empirical evidence 21

6.1. Ex-Ante: baseline analysis 21

6.2. Ex-Ante Balance-sheet analysis 23

6.3. Ex-ante embracing new technologies 25

6.4. Ex-post: Difference-in-differences 27

6.5. Ex-post: Propensity Score Matching robustness analysis 32

7. Conclusion 35

References 36

Appendix 1 39

Tables 16

Table 1. Distribution of cyberattacks over the years 16

Table 2. Descriptive Statistics of Selected Variables 18

Table 3. Difference in the means: Treatment vs Control Groups 20

Table 4. Panel logit regression results - baseline estimation 22

Table 5. Panel logit regression results: balance sheet determinants of cyberattacks 24

Table 6. Logit regression results: Digital technology channels for cyberattacks 26

Table 7. Difference-in-differences estimation results 29

Table 8. Difference-in-Differences estimation results with PSM 34

Figures 31

Figure 1. ATT confidence intervals of the results from the DiD estimation defined in equation (2) over year-time variation 31

Figure 2. Propensity Score Matching density function 32

Appendix Tables 40

Table 1. Description of variables 40

Table 2. Difference in mean PSM: Treatment vs Control Groups 41