Exclusive Use of Sam MCHUGH at EXANE S.A

EQUITIES

TELECOM OPERATORS

FROM TARGET PRICE EPS 18e EPS 19e  UNDERPERFORM NEUTRAL TELECOM ITALIA  -37%  -6%  -12% EUR0.55  TARGET PRICE EUR0.38 (DOWNSIDE 31%)

Bargain Hunters Beware

Downgrading TI to Underperform; TI is likely to remain a value trap 3 SEPTEMBER 2018 at 16:41* TI has fallen ~23% YTD, underperforming the SXKP by 12%, which itself has underperformed the Sam McHugh SXXP by 13%. A quick look at any comp sheet will show you it is trading at around 4.75x (+44) 207 039 9544 EV/EBITDA (Bloomberg) a hefty ~25% discount to peers. So as investors search for “value” one [email protected] can understand why TI often comes up. However our latest work on The Future of Fibre (published today) has led us to challenge our own long term growth assumptions for TI. In short we have San Dhillon become significantly more bearish, leading to significant cuts to our medium term estimates. We (+44) 207 039 9446 cut our price target from EUR0.60 to EUR0.38 (c30% downside) & downgrade TI to Underperform. Kohulan Paramaguru, CFA (+44) 203 430 8546 Our work identifies Italy and TI as facing the biggest risk to the fixed line outlook in Europe Alexandre Roncier (+44) 203 430 8437Exclusive Our work shows TI’suse wireline business of is over-earning relative to peers. Critically now that Open Fiber has secured project financing and Iliad has launched, we think the downside risks will begin [email protected] to be crystallised. Firstly we think the market still underestimates the negative impact Open Fiber’s

Specialist sales rollout will have on broadband ARPUs, voice telephony revenues and TI’s market share. Secondly Gareth Hollis Sam asMCHUGH the market has focused on the impact of Iliad on the wireless market, the impact on the wireline (+44) 207 039 8730 market has been overlooked. We believe falling mobile data prices will increase the attractiveness of mobile broadband to the marginal fixed broadband consumer, constraining broadband ARPU and penetration. Whilst mobile voice pricing/allowances at TIM/Vodafone have reached a level where they will start exerting structural downward pressure on TI’s fixed voice ARPU’s.

Iliad’s launch has triggered further mobile price cuts and Brazil at a turning point Elsewhere, after a brief period of stabilisation, Italian mobile prices have fallen further over the summer as the outlook continues to deteriorate. We also review the outlook for TIM Brasil who has operated in a unique sweet spot for 18 months, but now several supportive factors are beginning to turn against them, leading us to conclude the outlook is more challenging than consensus believes. We cut our Brazilian valuation by 15%. Our group level OpFCF estimates are 10% below consensus by ‘20, our FCF estimates are 30% below guidance and we expect leverage to remain >3.0x until at least 2022. In our view TI will remain a value trap for a long time.

Price (31 August 2018) EUR0.55 Performance(1) 1w 1m 3m 12m Market cap (EURbn) 14.5 Absolute(%) (8) (17) (20) (32) Free float (EURbn) 9.1 Rel. Telco Operators(%) (5) (11) (17) (22) EV (EURbn) 49.2 Rel. MSCI Europe(%) (8) (15) (20) (35) 3m avg volume (EURm) 60.1 Reuters / Bloomberg TLIT.MI / TIT IM Country / Sub Sector Italy / Incumbent Telcos

Financials 12/17 12/18e 12/19e 12/20e Valuation metrics(2) 12/17 12/18e 12/19e 12/20e EPS, Adjusted (EUR) 0.08 0.06 0.07 0.07 P/E (x) 9.5 8.6 7.8 8.1 EPS, IBES (EUR) 0.05 0.07 0.08 0.08 Net yield (%) 0.0 0.0 0.0 0.0 Net dividend (EUR) 0.00 0.00 0.00 0.00 FCF yield (%) 2.7 3.7 5.0 5.4 EV/Sales (x) 2.8 2.6 2.6 2.6 Sales (EURm) 19,828 19,143 18,787 18,494 EV/EBITDA (x) 6.4 6.1 6.0 6.0 EBITA, Adj. (EURm) 4,204 3,770 3,795 3,785 EV/EBITA (x) 13.1 13.0 12.8 12.6 Net profit, Adj.(EURm) 1,778 1,295 1,429 1,375 EV/CE (x) 1.1 0.9 0.9 0.9 ROCE (%) 5.8 5.0 5.0 4.9 Net Debt/EBITDA, Adj. (x) 3.4 3.6 3.6 3.5

Source: Exane BNPP (estimates), Thomson Reuters (consensus) (1) In listing currency, with dividend reinvested (2) Yearly average price for FY ended 12/17

* Date and time (London Time) on which the investment recommendation was finalised. It may differ from the date and time of broad dissemination on the website. See Appendix (on p45) for Analyst Certification, Important Disclosures and Non-US Research Analyst disclosures.

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Price at 31 Aug. 18 / 12m Target Price EUR0.55 / EUR0.38 -31% TELECOM ITALIA (Underperform) Reuters / Bloomberg: TLIT.MI / TIT IM Analyst: Sam McHugh (+44) 207 039 9544 Incumbent Telcos | Telecom Operators - Italy Com pany Highlights EURm 2.4 Enterprise value 49,194 Market capitalisation 14,542 Free float 9,075 1.6 3m average volume 60 1.2 Performance (*) 1m 3m 12m Absolute (17%) (20%) (32%) Rel. Sector (11%) (17%) (22%) 0.8 Rel. MSCI Europe (15%) (20%) (35%) 12m Hi/Lo (EUR) : 0.88 -38% / 0.55 +0% CAGR 2006/2018 2018/2020 EPS restated (**) (8%) 3% 0.4 Target Price CFPS (7%) (1%) Price 3.8*CFPS Relative to MSCI Europe Price (yearly avg from Dec. 07 to Dec. 17) 2.2 1.3 1.0 1.0 0.9 0.8 0.6 0.9 1.1 0.8 0.8 0.5 0.5 0.5 PER SHARE DATA (EUR) De c. 07 De c. 08 De c. 09 De c. 10 De c. 11 De c. 12 Dec. 13 Dec. 14 Dec. 15 Dec. 16 Dec. 17 Dec. 18e Dec. 19e Dec. 20e No of shares year end, basic, (m) 19 407.000 19 407.000 19 407.000 19 433.963 19 280.700 19 280.700 19 281.000 19 334.900 19 364.000 21 067.200 21 230.915 21 230.915 21 230.915 21 230.915 Avg no of shares, diluted, excl. treasury stocks (m) 19 407.000 19 407.000 19 407.000 19 420.482 19 280.700 19 280.700 20 139.957 21 026.163 21 608.906 22 157.192 21 690.300 21 230.915 21 230.915 21 230.915 EPS reported, Gaap 0.12 0.11 0.08 0.15 (0.25) (0.08) (0.03) 0.07 (0.00) 0.09 0.05 0.06 0.07 0.06 EPS company definition EPS restated, fully diluted 0.14 0.12 0.09 0.13 0.03 0.08 0.04 0.04 (0.01) 0.09 0.08 0.06 0.07 0.07 % change (14.3%) (15.7%) (23.7%) 41.8% (75.4%) 150.8% (56.3%) 3.0% NS NS (3.2%) (24.7%) 9.9% (3.6%) Book value (BVPS) (a) 1.1 1.1 1.1 1.2 0.9 0.8 0.7 0.8 0.7 0.8 0.8 0.9 0.9 1.0 Net dividend 0.080.050.050.060.040.020.000.000.000.000.000.000.000.00 STOCKM ARKET RATIOS De c. 07 De c. 08 De c. 09 De c. 10 De c. 11 De c. 12 Dec. 13 Dec. 14 Dec. 15 Dec. 16 Dec. 17 Dec. 18e Dec. 19e Dec. 20e P / E (P/ EPS restated) 15.0x 10.9x 11.3x 7.8x 28.8x 9.5x 17.4x 23.8x NC 9.5x 9.5x 8.6x 7.8x 8.1x P / E relative to MSCI Europe 136% 99% 73% 63% 244% 70% 104% 136% NC 53% 61% 60% 60% 67% P / CF 4.7x 3.0x 3.0x 2.9x 3.7x 2.4x 3.1x 5.4x 4.6x 3.5x 3.2x 2.3x 2.3x 2.3x FCF yield 6.8% 9.4% 7.0% 9.7% 7.4% 13.8% 10.9% 2.8% (4.8%) 1.7% 2.7% 3.7% 5.0% 5.4% P / BVPS 1.93x 1.22x 0.98x 0.86x 0.98x 0.95x 0.87x 1.16x 1.52x 1.01x 0.97x 0.63x 0.60x 0.57x Net yield 3.7%3.8%4.8%5.7%4.6%2.6%0.0%0.0%0.0%0.0%0.0%0.0%0.0%0.0% Payout 55.5% 41.2% 53.9% 44.1% 132.8% 24.3% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% EV / Sales 2.84x 2.28x 2.31x 2.23x 2.00x 1.81x 2.05x 2.56x 3.22x 2.72x 2.78x 2.57x 2.58x 2.58x EV / Restated EBITDA 7.5x 5.9x 5.5x 5.2x 4.9x 4.6x 5.0x 6.3x 8.2x 6.3x 6.4x 6.1x 6.0x 6.0x EV / Restated EBITA 14.7x 11.8x 10.9x 9.8x 8.8x 8.4x 9.4x 12.4x 17.4x 13.2x 13.1x 13.0x 12.8x 12.6x EV / NOPAT 23.4x 18.8x 17.3x 15.6x 12.2x 11.7x 13.0x 17.2x 24.2x 18.4x 18.2x 18.1x 17.7x 17.5x EV / OpFCF 14.1x 11.9x 10.3x 10.5x 10.4x 8.9x 10.0x 17.4x 79.5x 19.7x 18.0x 17.2x 16.3x 15.5x EV / Capital employed (incl. gross goodw ill) 1.4x 1.1x 1.0x 1.0x 1.0x1.0x1.0x1.1x1.3x1.0x1.1x0.9x0.9x0.9x ENTERPRISE VALUE (EURm) 89,021 68,690 62,799 61,557 59,914 53,472 47,946 55,330 63,493 51,674 55,179 49,194 48,457 47,646 Market cap 52,074 31,494 24,724 24,655 22,624 18,572 14,830 20,763 26,622 20,723 21,721 14,542 14,542 14,542 + Adjusted net debt 35,701 37,378 38,893 35,711 35,237 32,708 30,829 32,077 33,304 30,561 29,289 29,573 28,936 28,225 + Other liabilities and commitments 2,500 2,500 2,500 2,399 2,643 2,429 2,523 2,515 2,809 2,722 5,142 5,042 4,942 4,842 + Revalued minority interests 1,539 657 828 2,409 3,875 3,483 2,716 4,067 5,628 2,292 2,243 3,252 3,252 3,252 - Revalued investments 2,793 3,339 4,146 3,617 4,465 3,720 2,952 4,092 4,870 4,624 3,215 3,215 3,215 3,215 P & L HIGHLIGHTS (EURm) Dec. 07 Dec. 08 Dec. 09 Dec. 10 Dec. 11 Dec. 12 Dec. 13 Dec. 14 Dec. 15 Dec. 16 Dec. 17 Dec. 18e Dec. 19e Dec. 20e Sales 31,310 30,158 27,163 27,571 29,957 29,503 23,407 21,573 19,718 19,025 19,828 19,143 18,787 18,494 Restated EBITDA (b) 11,822 11,682 11,383 11,801 12,302 11,703 9,668 8,757 7,782 8,199 8,677 8,123 8,055 8,002 Depreciation (5,784) (5,874) (5,622) (5,547) (5,493) (5,340) (4,548) (4,284) (4,135) (4,291) (4,473) (4,353) (4,260) (4,216) Restated EBITA (b) (**) 6,038 5,808 5,761 6,283 6,809 6,363 5,120 4,473 3,647 3,908 4,204 3,770 3,795 3,785 Reported operating profit (loss) 5,864 5,493 5,493 5,813 (602) 1,926 2,718 4,530 2,961 3,722 3,291 3,769 3,795 3,785 Net financial income (charges) (1,749) (2,484) (2,221) (1,774) (1,982) (1,964) (2,186) (2,178) (2,515) (907) (1,495) (1,525) (1,334) (1,373) Af filiates 0 68 67 99 (39) (6) 0 (5) 1 (16) (19) (19) (19) (19) Other 36 (29) (622) (7) (13) 2 341 541 611 47 0 0 0 0 Tax (1,682) (653) (1,121) (548) (1,643) (1,235) (1,111) (928) (401) (880) (490) (690) (757) (742) Minorities (7) (1) (15) (451) (446) (350) (436) (610) (729) (158) (166) (241) (256) (276) Net attributable profit reported 2,462 2,394 1,581 3,132 (4,725) (1,627) (674) 1,350 (72) 1,808 1,121 1,294 1,429 1,375 Net attributable profit restated (c) 2,796 2,358 1,800 2,555 624 1,566 714 768 (189) 1,895 1,778 1,295 1,429 1,375 CASH FLOW HIGHLIGHTS (EURm) Dec. 07 Dec. 08 Dec. 09 Dec. 10 Dec. 11 Dec. 12 Dec. 13 Dec. 14 Dec. 15 Dec. 16 Dec. 17 Dec. 18e Dec. 19e Dec. 20e EBITDA (reported) 11,648 11,367 11,115 11,423 12,249 11,698 9,453 8,815 7,340 8,016 7,801 8,122 8,055 8,002 EBITDA adjustment (b) 174 315 268 378 53 5 215 (58) 442 183 876 1 0 0 Other items Exclusive(174) (315) (268)use (378) (131) of (1,178) (235) (135) (277) (369) (354) (1) 0 0 Change in WCR 199 (209) (472) (974) (866) 213 (239) (789) (1,628) (548) (581) (1,000) (800) (800) Operating cash flow 11,847 11,158 10,643 10,449 11,305 10,738 9,194 7,833 5,877 7,282 7,742 7,122 7,255 7,202 Capex (5,520) (5,365) (4,543) (4,583) (5,538) (4,702) (4,391) (4,659) (5,078) (4,658) (4,684) (4,253) (4,282) (4,128) Operating free cash flow (OpFCF) 6,327 5,793 6,100 5,866 5,767 6,036 4,803 3,174 799 2,624 3,058 2,868 2,973 3,074 Net financial items + tax paid (2,707) (2,779) (4,320) (3,235) (3,798) (2,991) (2,883) (2,489) (2,337) (2,234) (2,419) (2,214) (2,091) (2,115) Free cash flow 3,620 3,014 1,780 2,631 1,969 3,045 1,920 685 (1,538) 390 639 655 882 959 Net financial investments & acquisitions 7 782 0 928 411 59 84 (277) 929 692 (609) (700) 0 0 Other Sam MCHUGH858 (443) (1,428) 216 (4,197) 389 412 (1,404) (600) 584 1,461 0 0 0 Capital increase (decrease) 0 0 0 0 0 0 0 0 186 1,304 16 0 0 0 Dividends paid (2,886) (1,691) (1,060) (1,122) (1,326) (964) (537) (252) (204) (227) (235) (239) (245) (248) Increase (decrease) in net financial debt (1,600) (1,662) 708 (2,653) 3,143 (2,529) (1,879) 1,248 1,227 (2,743) (1,272) 284 (637) (711) Cash flow, group share 8,919 8,584 6,739 6,960 4,884 6,157 4,068 3,418 5,168 5,165 5,400 4,981 5,059 4,902 BALANCE SHEET HIGHLIGHTS (EURm ) De c. 07 De c. 08 De c. 09 De c. 10 Dec. 11 Dec. 12 Dec. 13 Dec. 14 Dec. 15 Dec. 16 Dec. 17 Dec. 18e Dec. 19e Dec. 20e Net operating assets 68,339 66,045 64,811 68,365 61,505 52,898 49,431 50,157 50,730 52,923 53,194 53,794 53,816 53,728 WCR (4,299) (5,094) (2,697) (4,597) 314 (1,743) (2,218) (499) (159) 506 (779) 221 1,021 1,821 Restated capital employed, incl. gross goodwill 64,040 60,951 62,114 63,768 61,819 51,155 47,213 49,658 50,571 53,429 52,415 54,015 54,837 55,549 Shareholders' funds, group share 26,985 26,126 25,952 28,819 22,791 19,378 17,061 18,145 17,610 21,207 21,557 22,657 23,888 25,071 Minorities 1,063 730 1,168 3,791 3,904 3,634 3,125 3,554 3,723 2,346 2,226 2,422 2,631 2,852 Provisions/ Other liabilities 2,640 2,290 735 1,989 2,737 1,957 1,822 2,214 2,294 2,478 2,826 2,846 2,865 2,884 Net financial debt (cash) 35,701 34,039 34,747 32,094 35,237 32,708 30,829 32,077 33,304 30,561 29,289 29,573 28,936 28,225 FINANCIAL RATIOS (%) De c. 07 De c. 08 De c. 09 De c. 10 De c. 11 De c. 12 Dec. 13 Dec. 14 Dec. 15 Dec. 16 Dec. 17 Dec. 18e Dec. 19e Dec. 20e Sales (% change) 0.1% (3.7%) (9.9%) 1.5% 8.7% (1.5%) (20.7%) (7.8%) (8.6%) (3.5%) 4.2% (3.5%) (1.9%) (1.6%) Organic sales grow th (1.3%) (2.3%) (5.6%) (3.8%) 2.7% 0.5% (5.2%) (5.4%) (4.6%) (3.5%) 4.2% (3.5%) (1.9%) (1.6%) Restated EBITA (% change) (**) (19.0%) (3.8%) (0.8%) 9.1% 8.4% (6.6%) (19.5%) (12.6%) (18.5%) 7.1% 7.6% (10.3%) 0.7% (0.3%) Restated attributable net profit (% change) (**) (14.3%) (15.7%) (23.7%) 41.9% (75.6%) 150.8% (54.4%) 7.5% NC NC (6.1%) (27.2%) 10.4% (3.8%) Personnel costs / Sales NCNCNCNCNCNCNCNCNCNCNCNCNCNC Restated EBITDA margin 37.8% 38.7% 41.9% 42.8% 41.1% 39.7% 41.3% 40.6% 39.5% 43.1% 43.8% 42.4% 42.9% 43.3% Restated EBITA margin 19.3% 19.3% 21.2% 22.8% 22.7% 21.6% 21.9% 20.7% 18.5% 20.5% 21.2% 19.7% 20.2% 20.5% Tax rate 40.9% 21.2% 33.6% 13.2% NC NC NC 39.5% NC 31.4% 27.6% 31.0% 31.0% 31.0% Net margin 9.0% 7.8% 6.7% 10.9% 3.6% 6.5% 4.9% 6.4% 2.7% 10.8% 9.8% 8.0% 9.0%8.9% Capex / Sales 17.6% 17.8% 16.7% 16.6% 18.5% 15.9% 18.8% 21.6% 25.8% 24.5% 23.6% 22.2% 22.8% 22.3% OpFCF / Sales 20.2% 19.2% 22.5% 21.3% 19.3% 20.5% 20.5% 14.7% 4.1% 13.8% 15.4% 15.0% 15.8% 16.6% WCR / Sales (13.7%) (16.9%) (9.9%) (16.7%) 1.0% (5.9%) (9.5%) (2.3%) (0.8%) 2.7% (3.9%) 1.2% 5.4% 9.8% Capital employed (excl. gdw ./intangibles) / Sales 40.4% 35.0% 44.9% 43.4% 54.3% 40.8% 47.0% 59.7% 74.6% 88.7% 79.5% 90.7% 96.8% 102.2% ROE 10.4% 9.0% 6.9% 8.9% 2.7% 8.1% 4.2% 4.2% (1.1%) 8.9% 8.2% 5.7% 6.0% 5.5% Gearing 127%139%143%110%132%142%153%148%156%130%123%118%109%101% EBITDA / Financial charges 6.8x 4.7x 5.1x 6.6x 6.2x 6.0x 4.4x 4.0x 3.1x 9.0x 5.8x 5.3x 6.0x 5.8x Adjusted financial debt / EBITDA 3.0x 3.2x 3.4x 3.0x 2.9x 2.8x 3.2x 3.7x4.3x3.7x3.4x3.6x3.6x3.5x ROCE, excl. gdw ./intangibles 30.1% 34.6% 29.7% 33.1% 30.1% 38.0% 33.5% 25.0% 17.9% 16.7% 19.2% 15.6% 15.0% 14.4% ROCE, incl. gross goodw ill 5.9% 6.0% 5.8% 6.2% 7.9% 9.0% 7.8% 6.5% 5.2% 5.3% 5.8% 5.0% 5.0% 4.9% WACC 8.1%8.2%9.5%9.8%10.5%9.5%8.1%8.9%9.6%8.9%9.0%8.0%8.0%8.0% Latest Model update: 01 Sep. 18 (a) Intangibles: EUR36,654.00m, or EUR1 per share. (b) adjusted for capital gains/losses, impairment charges, exceptional restructuring charges, capitalized R&D, pension charge replaced by service cost (c) adj.for capital gains losses, imp.charges, capitalized R&D, am. of intangibles from M&A, exceptional restructuring, (*) In listing currency, w ith div. reinvested, (**) also adjusted for am. of intangibles f rom M&A, or f or am. of gw ill for pre IFRS year

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 2

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Contents

Investment summary ______4

What’s New? ______5

Summary Conclusions ______6

Estimates, guidance, consensus and valuation ______16

The outlook for TI’s wireline business is worse than consensus thinks21

Latest on Italian Mobile ______34

The Sweet Spot for TIM Brasil is Over ______39

Investment case, valuation and risks ______44

Company profile and financial highlights ______49 Exclusive use of

Sam MCHUGH

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 3

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Investment summary

Downgrading Telecom Italia to Underperform, TP EUR0.38 Telecom Italia’s shares have already fallen by 23% YTD and 12% relative to the SXKP as Italian political drama; Iliad’s more aggressive mobile launch; the roll-back of 28 day billing and the admission at 2Q results that FY17 EBITDA had been flattered by one- offs have all weighed on the stock. This is despite the revelation in March ’18 that Elliott Advisors had built a stake in the stock and the subsequent upheaval saw Vivendi lose control of the board.

So are we at the bottom? In an accompanying report also published today, The Future of Fibre we argue that consensus is too optimistic on outlook for broadband revenue growth for the European Telecom sector, and specifically that broadband ARPUs are unlikely to grow sufficiently to offset other legacy revenue declines (voice telephony, B2B price premiums, cable TV). As part of this work we undertook a detailed review of market structure to determine which European markets were most structurally at risk of delivering disappointing fixed line revenue growth in the medium term.

Unfortunately for TI, this work shows Italy is one of the worst fixed markets in Europe from a structural growth perspective and that TI screens as the most-at-risk incumbent across all of our coverage. This is despite a clear opportunity in terms of broadband penetration growth in Italy, which does lag behind other European countries. Critically Open Fiber secured project financing in August and Iliad has launched. We believe these events will prove to be significant catalysts to crystallise earnings downgrades Exclusiveand hamper use future revenue of growth at TI. We believe these factors are either underestimated or not well understood by consensus.

This work has led us to re-visit our medium to long term assumptions for TI’s fixed line Sam MCHUGHbusiness; whilst our near term (FY18) forecasts remain broadly unchanged, we cut our Wireline service revenue estimates by -0.9% in FY19, -2.0% in FY20, -4.5% in FY21 and upwards of -7% thereafter. On top of this we note that Iliad has been more aggressive than we initially expected (FREEconomics 2 - Ho, Ho, Ho), and our monthly tracking of mobile prices shows that July and August have seen further cuts following a period of relative stability. Lastly we review the outlook for TIM Brasil, where we caution the sweet spot that the company has been in for the last 18 months is set to end, which, combined with FX headwinds, is going to be a growing headwind for the TI Group in our view.

This combination of factors leads to high teen cuts to our FCF forecasts in FY18-20, and our revised cumulative FCF estimates for FY18-20 (pre Spectrum and pre- minorities) of EUR3.1bn is some 30% below company guidance of EUR4.5bn. Our longer term FCF estimates fall by close to 30%, which drives a ~35% cut in our price target from EUR0.60 to EUR0.38, implying c30% downside. On our revised forecasts TI’s leverage will remain >3.0x even by 2022, and hence those hoping to see a transfer of value from debt to equity will be left disappointed. On our estimates TI is trading on an unlevered FCF yield of as little as 3.1% in FY19 and a FCFE yield of 3.8%, leaving it trading at a material premium to peers on 5.1% and 6.9% respectively, especially when it also has no dividend support. To us TI remains a value trap, and bargain hunters should beware. We downgrade TI from Neutral to Underperform.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 4

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

What’s New?

What work have we done? Assessing the outlook for fixed ARPH growth In this report we expand on the work presented in The Future of Fibre by focusing on the potential for fixed service revenue growth for Telecom Italia. Specifically we analyse Italy and TI through the lens of five key tests that are aimed at judging which European markets are most vulnerable to ARPH (average revenue per household) declines despite growth in fibre. Specifically we look at;

1) Legacy revenue exposure; 2) Fixed line ARPH vs broadband speeds; 3) Fixed line ARPH vs value for money from competing mobile networks; 4) Market concentration (HHI’s), and 5) Fibre overbuild risk

Catalysts for change in the fixed market Whilst the above work gives us a good understanding of the potential risks, without a change in circumstances / market dynamic it is possible that the status-quo could be sustained. As a result we go on to analyse the market to understand if there are any catalysts for change.

The latest on mobile pricing and Iliad’s launch Having written in depth on the disruptive potential of Iliad’s Italian mobile launch in ExclusiveNovember 2017, seeuse FREEconomics of 2 - Ho, Ho, Ho, we update some of the work initially presented in this note. Specifically we review the trends in mobile pricing and mobile service revenue trends following Iliad’s actual launch in May, and we go on to update the work we first presented in Mobile Millennial. This helps us update our Sam assessmentMCHUGH of the impact from Iliad, and specifically to help us understand which operator (TIM, Vodafone or Wind-Tre) will be most impacted by Iliad’s launch.

Assessing the outlook for TIM Brasil Brazil may represent only 13% of TI in terms of EV but in terms of market cap it is far more important at around a third of TI equity. In this report we update our estimates for the latest FX rates, and also review the latest market dynamics in Brazil and the outlook for TIM Brasil for the next 12 months.

What does all of this mean for our estimates, valuation and rating? We then explain how all of our work has changed our estimates on TI, and the implication on valuation and our rating.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 5

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Summary Conclusions

Italy screens as one of the most ‘at risk’ fixed line markets in Europe Across the “five tests” Italy consistently appears as one of the most ‘at risk’ countries in our review, specifically 1) Italy has above average exposure to “legacy revenues”. 2) Italians are probably getting the worst value for money from their fixed line offering in all of Europe, with an average speed of less than 40Mbps and a PPP adj. ARPH of EUR65, i.e. there is little broadband ARPU upsell potential in Italy even as delivered speeds increase. 3) When comparing the PPP adjusted fixed ARPH to the PPP adjusted cost of 1GB of data (adjusted for network quality) we find that Italy is again an outlier, with high fixed ARPH and relatively low mobile prices. This suggests mobile will be more of a cap on fixed line pricing and penetration in Italy relative to other markets. 4) Italy is behind only the UK and Spain in terms of the rate of decline in HHI in the broadband market i.e. competition is growing at one of the highest rates in Europe. 5) Italy has by far the worst fiber overbulild threat in all of Europe, with Open Fiber targeting a build to at least 70% of Italian homes, aided by subsidies for rural areas. We remain sceptical of a (value creative for TI) deal between Open Fiber and TI (see our pushbacks below).

Figure 1: Italy screened as being the second worst European country in terms of risk to the fixed line revenue outlook Exane Fixed Line Scorecard

Exclusive use of Sam MCHUGH

Source: Exane BNP Paribas estimates

1. TI in particular screens as having high legacy revenues It is crucial to highlight that although a Country may be badly positioned in our analysis, this does not mean all the fixed operators in said Country are equally at risk. The burden may mostly be carried by the incumbent, or the cable operator, with alternative providers actually benefitting from some of the woes of competitors. However Italy, due to historical quirks, is one of a few markets in Europe where there is no cable infrastructure and where the incumbent, TI, still has relatively high market share in both retail and wholesale.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 6

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Figure 2: TI has the highest exposure to legacy revenues of all incumbents under our coverage at 37% (!) of group EBITDA, perhaps understandable given they have 45% retail broadband share and 98% broadband infrastructure market share “Legacy Revenue” as a % of Domestic TSR and as a % of Incumbent Broadband Market Share Group EBITDA

25% 100% IT TI FR 90% 20% DTTEF ES 80% UK 15% Telia Proximus DE PT KPN 70% Orange 10% BT Elisa 60% Telenor SE NO 5% NL Legacy Legacy Revenue as % Domestic TSR 50%

%SUBSCRIBERS ON INCUMBENT INFRA. INCUMBENT ON %SUBSCRIBERS BE 0% 40% 0% 10% 20% 30% 40% 30% 35% 40% 45% 50% Legacy Revenue as % Group EBITDA %RETAIL SUBSCRIBER SHARE Source: Exane BNP Paribas estimates

We think there are multiple catalysts ahead to crystallise earnings downgrades Having established that Italy, and TI in particular, face the most challenging fixed line Exclusiveoutlook use in Europe, the of question then becomes – is there anything that will change the status quo? The short answer is yes. We also believe that 2018 will prove a tipping point, where the catalysts required to crystallise the earnings downgrades are beginning play out. We have put these catalysts into three buckets – the first we think is Sam MCHUGHrelatively well understood, but the 2nd and 3rd are in our view not well understood at all. Catalyst 1: Open Fiber – market share loss is relatively well understood In early August 2018 Open Fiber signed a EUR2.5bn project finance package to fund their network rollout, whilst owners Enel and the CDP are also said to be willing to grant a further EUR950m of funding. This removes one of the key investor pushbacks on Open Fiber – that they would not be able to secure financing. The company have now said that they will spend at least EUR1bn a year rolling out FTTH. Whilst the Open Fiber network only currently has ~300k subscribers (half of which are Fastweb subscribers in Milan on the old Metroweb network) they now have commercial commitments from over 3m subscribers (with the likes of Vodafone signing up) and net- additions have started to accelerate.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 7

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Figure 3: We think the market does relatively well understand that growth at Open Fiber will come at the expense of Telecom Italia Structure of Italian Broadband Market, 2017, 2021 and 2025

Total Households – 25m 2017 Users Total broadband lines – 16.5m

Network used Telecom Italia – 15.1m OF FWA

Service provider Telecom Italia – 7.5m OLOs – 7.8m OLOs

TI Revenues Retail – EUR2.1bn Wholesale – EUR2.0bn

2021 Users Total broadband lines – 19.2m

Network used Telecom Italia – 16m OF – 2.3m FWA

Service provider Telecom Italia – 8m OLOs – 7.9m OLOs – 3.3m

TI Revenues Retail – EUR2.75bn Wholesale – EUR1.9bn

2025 Users Total broadband lines – 20.3m

Network used Telecom Italia – 14.5m OF – 4.6m FWA

Service provider Telecom Italia – 8.2m OLOs – 6.3m OLOs – 5.8m

TI Revenues Retail – EUR2.85bn Wholesale – EUR1.6bn

NB: OF= Open Fiber

Source: Exane BNP Paribas estimates ExclusiveCatalyst use 2: Open Fiber of – the less well understood impact of FTTH rollout Whilst these immediate first order implications of Open Fiber’s rollout may be well understood there are several second order implications which we believe are neither well understood by the market, nor included in consensus estimates for Telecom Italia. Sam MCHUGHSpecifically we believe consensus underestimates the negative impact on both ARPU and market share for TI.

2.1 Open Fiber rollout likely to cap TI’s broadband ARPU growth

Perhaps the best place to start is to revisit one of our tests from above – namely looking at the delivered broadband speed relative to ARPH. As detailed above Italy stands out as having very high ARPH and very low relative broadband speeds – if we add on some of the retail offers of those using the Open Fiber network today we see that Vodafone Italy are selling a 1Gbps FTTH plan for just EUR30/month (or EUR33 if PPP adjusted), which is inclusive of unlimited domestic and international fixed and mobile voice calling. This clearly compares very favourably to the current legacy position in Italy where ARPH is EUR65 and the average speeds are less than 40Mbps.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 8

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Figure 4: Italian Fixed ARPH is extremely high relative to the quality of the broadband packages sold, and as Open Fiber rolls out FTTH we expect ARPH to fall significantly Adjusted ARPH (EUR/month) versus average speed of packages purchased (Mbps)

80 Spain 75 "OVER-EARNING"

Italy Norway Belgium 65 UK Portugal

55 Europe

50 Netherlands Sweden 45

Germany France 40 Vodafone Italy are selling 1Gbps FTTH for

Source: Exane BNP Paribas estimates Exclusive2.2 OFuse rollout will accelerateof the decline of TI’s fixed voice revenues (EUR3.2bn) Whilst there are not lots of precedents, if we look at the Japanese and Korean markets then it is very clear that the launch of FTTH coincides with a period of accelerating or sustained line losses and voice telephony revenues. Intuitively this makes sense – Sam MCHUGHgreater availability of higher quality broadband connectivity, especially today with the proliferation of VoIP services as well as video services (Skype, FaceTime etc), results in an acceleration in the decline in usage of legacy fixed voice.

Figure 5: As FTTH penetration accelerates, so does the loss in telephony lines and decline revenues In Japan NTT Commercialised FTTH in 2001, while in Korea KT launched FTTH in 2006 NTT Line Losses (m) vs. Japan FTTH Penetration (%) KT Telephony Rev Growth (%) vs. FTTH Penetration (%)

'03 '05 '07 '09 '11 '13 '15 '17 2002 2004 2006 2008 2010 2012 2014 2016 3

0.2 56 83

-0.0 -1 -2 -1 -0.6 -3 -1.1 -5 -5 -1.3 -1.3 -1.7 -2.0 -10 -10 -2.5 -10 -11 -2.7 -11 -12 -11 -3.0 -13 -3.1 -3.3 -13 -3.4 -3.6 -3.7 Line Losses FTTH Pen. Telephony Rev. Growth Effective FTTH/B

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 9

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Part of the reason that ARPH is so high in Italy is, we think, due to the fact that TI has a lot of legacy revenue in voice (discussed in our analysis above). Whilst TI does disclose fixed voice revenues (EUR3.2bn in 2017) they do not split these between variable voice calling and line rental, translating into a voice ARPU of close to EUR30/month (inc VAT). It is clear that TI is making a lot of money from this legacy product.

As discussed above we are sceptical of TI being able to deliver sustained broadband growth with broadband ARPU’s of over EUR32/month (inc VAT), so they cannot “shift profitability” away from voice and into the broadband bucket. As a result we expect TI’s voice line losses and voice revenue declines to accelerate in the medium to long term as Open Fiber’s rollout gains traction.

2.3 New entrant risk: The disruptive impact of Sky’s (and potentially Iliad’s) broadband launches are underestimated

One of the other key takeaways from our work on Japan is that FTTH rollout has led to a proliferation of retail competition. In the case of Italy we already have one confirmed new-entrant and one very obvious potential new entrant – Sky Italia and Iliad respectively. Sky has confirmed they will launch broadband on Open Fiber in summer 2019. Whilst Sky has confirmed their launch we think current consensus estimates for TI reflect little to zero impact of Sky launching broadband in Italy. We think this will prove a mistake, and using data from our STAMP 2018 survey we show that Sky (with close to 5m subscribers in Italy) has market leading NPS scores, better pricing power than TIM vision, and their customer base over-indexes to TI broadband subs. I.e. they have some of the right tools to build up a broadband business.

ExclusiveFigure 6:use Our analysis suggestsof that Italy will undergo one of the biggest increases in retail competition across Europe over the next 4 years, and this is before we even factor in a broadband launch at Iliad Broadband HHI (2022) vs Change in Broadband (2022 vs 2017) Sam MCHUGH

2% "HIGH HHI, CONCENTRATION GROWING" Finland 0% Germany Norway Netherlands

(2%) Sweden Portugal

(4%) France Europe (6%)

UK (8%) "JAPAN IS AT ~2,100" Italy (10%)

(12%) Belgium

Change Change in Broadband HHI 2022 vs 2017 (Exane est.) "LOW HHI, CONCENTRATION "HIGH HHI, CONCENTRATION (14%) DECLINING" Spain DECLINING"

(16%) 2,000 2,200 2,400 2,600 2,800 3,000 3,200 3,400 3,600 3,800 4,000 2022 HHI (Exane est) Source: Exane BNP Paribas estimates

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 10

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

We are currently taking a very cautious approach to forecasting growth at Sky and Iliad. In our forecasts we assume Sky reaches just 200k subscribes by 2022, or 4% of their TV customer base, and we do not currently forecast any subscribers for Iliad. We think this is a very conservative approach, and could be a source of further downside to our TI estimates. Even on these numbers if we revisit our earlier HHI analysis and roll forward our analysis to look at how HHI changes from 2017 to 2022 then we clearly see that Italy moves from being around the market average to being one of the worst markets over the next 4 years.

Catalyst 3: Iliad – the overlooked impact of Iliad’s mobile launch on the broadband market Much has been made of Iliad’s mobile launch in the Italian market in May this year, and as discussed later in this note, Iliad has probably been more aggressive on pricing than we anticipated pre-launch. Most of our discussions with investors and market commentary in relation to Iliad’s mobile launch have centred on the impact on mobile service revenue in Italy. However we believe that when we look back at their launch in 5 years’ time, we will also realise that their launch had a significant impact on the fixed line market in Italy. At the moment we think consensus totally overlooks this impact and is thus too optimistic in their outlook for fixed revenue growth in Italy.

3.1 Falling mobile data prices will serve to increase the attractiveness of mobile broadband to the marginal fixed broadband consumers – limiting fixed ARPU growth and penetration gains

Average mobile data prices in Italy have fallen 80% (on a per GB basis) over the last two years – from ~EUR4/GB to just EUR0.8, and Iliad is now selling 40GB of mobile data for just EUR6.99 (we discuss and analyse some of these trends in the below Exclusivesection on the mobile use market). As of discussed in “Test 3” above, there is a good correlation between fixed ARPH and the cost of mobile data, with Italy screening as one of the big outliers where fixed line ARPH looks very high relative to the (quality Sam adjusted)MCHUGH value for money offered by mobile plans. Taking this analysis one step further, if we add Iliad’s current mobile offer into our analysis (adjusted for the network quality of Wind-Tre) then it becomes immediately clear that the relative attractiveness of a mobile-only home broadband solution to the marginal customer is growing in Italy, and in particular that it has grown materially following Iliad’s launch, and as a result we think Iliad’s aggressive mobile offers are likely to put incremental downward pressure on both Fixed ARPH and broadband penetration in Italy.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 11

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Figure 7: Iliad’s cheap mobile plans will serve to constrain both broadband growth and penetration as mobile broadband becomes a more attractive proposition for the marginal consumer of fixed broadband Adjusted Fixed ARPH (EUR/month) versus mobile VFM

"Italy appears to overearn in Fixed given the 80 reasonable cost and quality of mobile offers - furthermore Iliad's entry into the mobile market has further reduced the cost per GB even when adjusted Spain for the poor quality of the Wind-Tre network" 70 (2) Iliad Mobile (1) Italy Today Belgium Norway

60 Our work Japan suggests that Italian fixed Europe ARPH could fall by up to 30-40%, South Korea 50 in part due to growing pressure Netherlands

PPP Adjusted ARPH Fixed Adjusted PPP from mobile as a substitute to fixed Sweden France Germany 40

(3) Italy 2025? WORSE MOBILE VFM Finland 30 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Mobile VFM Score: PPP Adj. EUR/month per GB adjusted for network quality (speed and availability) Source: Exane BNP Paribas estimates

3.2 The widespread availability of unlimited voice calls at EUR6.99/month will also accelerate fixed voice declines

ExclusiveLastly prior to the announcement use of Iliad’sof launch (2016) the cost of unlimited voice calls on TI or Vodafone’s mobile network was close to 40% more expensive than getting unlimited voice calls from TI’s fixed network. However subsequent to the news that Iliad would launch in mobile in Italy this premium has evaporated, such that today Samyou MCHUGH can get unlimited voice calls (from mobile to mobile/landlines) from Vodafone for as little as EUR11/month as part of their Vodafone Smart plan with 2GB of data, i.e. a >60% discount to the equivalent price on TI’s fixed network. We believe this will also exert structural downward pressure on TI’s fixed voice ARPU’s.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 12

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Figure 8: We think mobile voice pricing has now fallen enough to have a growing impact on fixed voice revenue at TI Monthly cost of TI fixed line with unlimited voice vs the cost of a mobile plan with unlimited calls (EUR/month)

45 Historically getting unlimited mobile 40 callling in Italy was However the gap between the cost of more expensive than it getting unlimited mobile calling has fallen 35 was on TI's fixed materially below the cost of unlimited network calling on TI's fixed network 30

20 Iliad's launch

EUR/ Month has futher 15 extended this gap 10

0 Aug-16 Aug-17 Aug-18 Iliad Cheapest TI or Vod mobile plan with Unlimited Voice Calls TI Fixed Voice ARPU (inc VAT)

Source: Exane BNP Paribas estimates

Three potential mobile headwinds In December 2017, we published our second deep dive note on Iliad’s launch in Italy, see FREEconomics 2 - Ho, Ho, Ho, in which we argued mobile prices in Italy were actually higher than first appears, implying that there was an opportunity for Iliad to disrupt the market. Given we have already written in-depth on this topic in prior notes, Exclusivein this note we focus use on updating thatof view and summarising what we think are three incremental downside risks to TI’s mobile business.

In this regard we note that our monthly pricing analysis shows that having reached Sam MCHUGHsome level of stabilisation through mid-2018, mobile prices have fallen further in July and August 2018, all but confirming MSR trends will deteriorate again in 3Q as the operators face lower prices and a full quarter of competition from Iliad. Secondly we go on to demonstrate why TI has a long term demographic problem and lastly we also provide an update on our latest expectations for the upcoming spectrum auction in Italy, where we raise our expectation for TI’s spend to EUR700m, and caution that the eventual cost could be higher again given the potential competitive intensity of bidding.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 13

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Figure 9: Price per GB in Italy has been more than halved in the last 18 months all but confirming 3Q trends will be worse than 2Q Price per GB trend per data bucket cohorts*

Average Cost per GB in a 6-30GB Plan Average Price for a 6-30GB Plan May 18 May 17 May 16 Aug 16 16 Dec Aug 17 17 Dec Aug 18 Feb 17 Feb 18 Sep 17 17 Nov Sep 16 16 Nov Mar18 Mar17 Mar16 Jun 16 Jun 17 Jun 18 Jun Oct Oct 16 Oct 17 Apr 16 Apr 17 Apr 18 Jan 18 Jan Jan 17 Jan Jul 18 Jul Jul 17 Jul Jul 16 Jul

4.5 40.0 Iliad still priced 4.0 Pricing pressures 35.0 re-accellerated a 50% discount 3.5 this summer 30.0 3.0 25.0 2.5 20.0 2.0 15.0 1.5 10.0 1.0

0.5 ILIAD ITALIA 5.0 ILIAD ITALIA

0.0 0.0

Source: Exane BNP Paribas estimates * note these are 3 month trailing average prices

The sweet spot for TIM Brasil is over Brazil may represent only 13% of TI in terms of EV but in terms of market cap it is far more important at around a third. Given the importance of Brazil it is worth taking a little time to consider what is happening in this market to understand if there is any upside or downside risk. In short we believe that TIM Brasil has been in a real sweet spot for the last 18 months as multiple factors have contributed to their success (namely a first mover advantage in 4G, Oi’s balance sheet problems and macro improvement). ExclusiveHowever use we believe consensusof estimates are extrapolating this improvement into the future, whereas we believe these three factors are now reversing and as a result we are more cautious on the outlook for TIM Brasil vs consensus – specifically our local currency OpFCF estimates are 9% below Bloomberg consensus by 2020. This together Sam MCHUGHwith currency headwinds is another source of downgrades we believe consensus will need to put through for TI group over the next 12 months.

What does this mean for our group estimates and price target? Our longer term FCF estimates fall by close to 30%. We conclude TI will not meet their medium term FCF guidance (Exane @ EUR3.1bn vs EUR4.5bn guidance), whilst our EBITDA estimates are 10% below consensus. In addition the upcoming spectrum auction in Italy is an incremental headwind and if we adjust for the impact of IFRS 15 we expect TI to remain >3.0x levered all the way through to 2022.

Our revised forecasts drive a ~37% cut in our price target from EUR0.60 to EUR0.38, some 30% downside. On our estimates TI is trading on an unlevered FCF yield of as little as 3.1% in FY19 and a FCFE yield of 3.8%, leaving it trading at a material premium to peers on 5.1% and 6.9% respectively.

Pushbacks Due to the nature of our work (it takes a long term and structural view) we do caution that this is not per se a call on the next quarter or two at TI (indeed recent price increases in broadband and the employee solidarity agreements should help TI in 3Q18 vs 2Q18 and we expect domestic Adj. EBITDA to decline just -0.5% vs -6.5% in 2Q18. However we believe any relief will prove short lived and would use any strength to add to short positions.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 14

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Doesn’t TI have loads of cost cutting?

Yes, we agree – in fact nearly all incumbent telcos are over-staffed and inefficient. Within the “addressable” domestic cost base of ~EUR7.2bn in FY17 we credit TI with a significant net EUR450m (or 6%) reduction in Opex between FY17 and FY20. However we caution that growing competitive intensity and investment in content may mean we are overly generous in our net cost cutting assumptions.

Won’t they do a deal with Open Fiber / spin off the network?

In taking a stake in TI earlier this year Elliott argued that TI should spin-off their network in order to benefit from multiple arbitrage and to try to crystallise a deal with Open Fiber. In addition TI is currently engaged in a process with the regulator which is seeing them attempt to create a more independent “NetCo” within the business (with more clearly demarked reporting, much like Openreach with BT). This in itself is not a spin- off, although is seen investors as a precursor to any such move.

However we remain sceptical that a spin-off can be achieved, and even if it is, whether it will be value creative for TI shareholders. Our fundamental issue is that Open Fiber now has secured financing and clearly do not need TI’s network in the near term to progress, and it is very unclear to us why the government (ultimately 50% owners of Open Fiber through the CDP) would do a deal that would benefit non-Italian private shareholders (read Vivendi and Elliott). Putting a value on any potential “Net Co” is fraught with difficulty given the overbuild risk, and the fact that unlike regulated utilities like water, a fibre network is not a closed utility and faces competition from mobile as a substitute.

Won’t they sell assets?

Over the last 12 months TI have discussed the potential for asset sales, with Exclusivebroadcasting companyuse Persidera of and international wholesale business Sparkle being the most often discussed.

– At 2Q18 results management stated that Persidera was the only thing “on the block” Sam MCHUGHat the moment. As per TI’s 2017 annual report, Persidera generated ~EUR40m of EBITDA. The domestic stub trades on ~5.0x EBITDA every 1.0x turn of EBITDA higher they receive is worth EUR40m and would add 0.5% to our price target. Given the sale has reportedly fallen through twice we are sceptical that anything will materialise here. – On Sparkle it's worth noting a couple of important points – firstly EBITDA in Sparkle declined around 40% in 1H18 (to around EUR50m), and EBIT was negative. In their presentation on the turnaround of TI, Elliott claimed that Sparkle was worth ~EUR1bn to TI today and could be worth EUR2bn in the future. The reality is KPN recently sold a similar business, iBasis, for what we understand was a ~4x EV/EBITDA multiple, implying Sparkle is probably worth closer to EUR400m. This would clearly be disappointing vs expectations, which may well explain why they haven't been able to sell it yet.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 15

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Estimates, guidance, consensus and valuation

Changes to estimates As discussed in this note we have reviewed our long term assumptions and estimates predominantly for TI’s wireline business. Our work has led us to cut our estimates for wireline service revenue by -0.9% in FY19, -2.0% in FY20, -4.5% in FY21 and upwards of -7% thereafter.

On top of this we note that Iliad has been more aggressive than we initially expected and our monthly tracking of mobile prices shows that July and August have seen further cuts following a period of relative stability. We cut our wireless service revenue estimates by 1-2% over our forecast period.

Lastly we update for FX, with the EUR/BRL exchange rate now at 4.84 vs 4.48 in our previously published model. This results in high-single digit cuts to our Brazilian estimates. The combination of all of these factors results in a ~30% cut to our long term FCF estimates.

Our detailed model and estimates are available on request.

Figure 10: We have cut out long term FCF estimates by upwards of 30% EBNPP Es tim ate s EXA NE - NEW EST IM A T ES C HA NGE ( %) (EURm) 2018 2019 2020 2021 2022 2018 2019 2020 2021 2022

REVENUES 19,143 18,787 18,494 18,122 17,838 -0.2% -2.1% -2.8% -4.0% -5.5% ExclusiveDomestic 15,321 use 15,170 14,870 of 14,513 14,256 0.4% -0.7% -1.6% -3.0% -5.0% Wireline 10,682 10,707 10,570 10,333 10,166 0.4% -0.9% -2.0% -4.5% -7.2% Wireless Service 4,603 4,422 4,247 4,112 4,012 -0.3% -1.3% -1.8% -0.5% -1.1% Wireless Equipment 644 644 644 644 644 3.5% 3.5% 3.5% 3.5% 3.5% Brasil 3,850 3,645 3,651 3,636 3,609 -2.5% -7.5% -7.5% -7.5% -7.5% SamWireline MCHUGH 189 191 199 200 200 -2.5% -7.5% -7.5% -7.5% -7.5% Wireless Service 3,495 3,297 3,301 3,291 3,270 -2.5% -7.5% -7.5% -7.5% -7.5% Wireless Equipment 167 156 150 145 140 -2.5% -7.5% -7.5% -7.5% -7.5%

EBITDA 8,122 8,055 8,002 7,748 7,459 -0.6% -3.4% -3.4% -6.1% -9.9% Domestic 6,677 6,646 6,575 6,315 6,023 -0.3% -2.6% -2.5% -5.9% -10.6% Brasil 1,461 1,425 1,442 1,450 1,452 -1.6% -6.8% -7.2% -6.9% -6.9%

Domestic EBITDA (adj.) 6,902 6,896 6,825 6,515 6,223 -0.4% -2.2% -1.7% -5.7% -9.8%

EBIT 3,769 3,795 3,785 3,581 3,347 -1.3% -6.2% -7.1% -12.1% -20.0% Domestic 3,243 3,232 3,171 2,929 2,665 -1.7% -6.2% -7.1% -13.2% -22.8% Brasil 543 580 630 668 698 0.7% -5.9% -6.8% -6.3% -6.4%

Interest Costs -1,524 -1,334 -1,373 -1,365 -1,376 4.9% 5.6% 10.7% 11.5% 15.7% Taxes -690-757-742-681-605 -5.2% -11.7% -15.0% -22.3% -34.4% NET INCOME 1,294 1,429 1,375 1,224 1,040 -6.2% -12.8% -16.7% -25.6% -39.9% NET INCOME (Exane) 1,295 1,429 1,375 1,224 1,040 -6.2% -12.8% -16.7% -25.6% -39.9% EPS (EURc) 6.4 7.0 6.8 6.0 5.2 -6.0% -12.4% -16.1% -24.8% -38.6%

CAPEX 4,753 4,082 3,928 3,761 3,615 8.9% 1.6% 2.9% -10.3% -1.7% Domestic 3,841 3,262 3,123 2,975 2,851 12.0% 4.1% 6.0% -11.0% 0.0% Brasil 913 820 805 786 764 -2.8% -7.5% -7.5% -7.5% -7.5%

Adjusted OpFCF 3,368 3,973 4,074 3,988 3,843 -11.4% -8.0% -8.8% -1.7% -16.5%

FCF (TI Definition) 855 1,082 1,159 1,302 1,350 -7.2% -22.5% -25.4% -28.4% -31.8%

Source: Exane BNP Paribas estimates

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 16

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Exane vs Guidance TI has medium term 2018-2020 guidance and our estimates are now below on all the critical factors, namely;

– Guidance: Low single digit domestic adjusted EBITDA growth 2017-2020 CAGR. Exane: we now expect domestic adj. EBITDA to decline -1.1% 2017-2020 CAGR

– Guidance: Group adjusted net-debt/EBITDA of ~2.7x (ex-Spectrum) in 2018 and further reducing in 2019 and 2020. Exane: Our revised estimates have leverage at 3.1x in 2018 reducing to just 2.9x by 2020. TI are also in the middle of adopting IFRS 15 which is expected to have a ~EUR300m annualised negative impact on EBITDA. Adjusting for this leaves TI at >3.0x net-debt to EBITDA through to at least 2022. Our estimates include spectrum costs of EUR700m in 2018.

– Guidance: Group cumulative equity FCF of ~EUR4.5bn 2018-20. Exane: Our estimates are ~30% below guidance at EUR3.1bn of FCF over this period (notably adjusting for minorities, finance leases and spectrum we expect TI to generate just ~EUR840m of FCF cumulatively between 2018 and 2020.

Figure 11: We expect TI’s net-debt to EBITDA (IFRS15 adj.) to remain >3.0x for the foreseeable future whilst our FCF estimates are ~30% below guidance TI Net-Debt / EBITDA TI Cumulative FCF 2018-20 (pre-spectrum, pre-minorities and pre-dividend)

3.7x 5000 3.5x 4500

3.2x 4000 3500 3.0x Exclusive use3000 of

2.7x 2500

2000 2.5x Sam MCHUGH1500 2.2x 1000

500 TI Adj. Net Debt / EBITDA 0 IFRS 15 Adjusted Net Debt / EBITDA Guidance Exane Cumulative FCF 2018-2020

Source: Exane BNP Paribas estimates

Exane vs Consensus TI does not distribute a company compiled consensus with annual estimates, so we have to rely on Bloomberg estimates. We note that these only go out as far as 2020, with outer years not having enough contributing analysts to be reasonable benchmarks. Below we also compare our TIM Brasil estimates to the consensus estimates for the locally listed subsidiary.

In aggregate our group revenue estimates are 4% below consensus by 2020, our EBITDA estimates are 7% below consensus and our OpFCF estimates are ~10% below consensus by 2020. Meanwhile our net-debt estimates are ~13% above consensus and leverage is 20% above consensus by 2020.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 17

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Within the mix we estimate that our revenue estimates for Italy and Brazil are both ~4% below consensus by 2020, but at the EBITDA level we are more bearish vs consensus on Italy (-7% by 2020) than we are vs consensus on Brazil (“just” 4% below consensus by 2020).

Figure 12: Our TI estimates are below consensus nearly across the board – with a greater difference between us and consensus in Italian EBITDA vs Brazilian numbers Exane vs Consensus

EBNPP Estimates CONSENSUS - BLOOM BERG EXANE - LIV E DELTA (%) (EURm) 2018 2019 2020 2018 2019 2020 2018 2019 2020

REV ENUES 19,322 19,258 19,253 19,143 18,787 18,494 -0.9% -2.4% -3.9% Dom e s tic 15,458 15,528 15,462 15,321 15,170 14,870 -0.9% -2.3% -3.8% Wireline 10,682 10,682 10,682 Wireless Service 4,603 4,422 4,247 Wireless Equipment 644 644 644 Br as il 3,864 3,730 3,791 3,850 3,645 3,651 -0.4% -2.3% -3.7%

EBIT DA 8,361 8,495 8,589 8,122 8,055 8,002 -2.9% -5.2% -6.8% Domes tic 6,909 7,042 7,084 6,677 6,646 6,575 -3.4% -5.6% -7.2% Domestic (adj.) 7,134 7,292 7,334 6,902 6,896 6,825 -3.3% -5.4% -6.9% Brasil 1,452 1,453 1,505 1,461 1,425 1,442 0.6% -1.9% -4.2%

CAPEX 4,397 4,103 3,971 4,753 4,082 3,928 8.1% -0.5% -1.1% Domes tic 3,490 3,265 3,164 3,841 3,262 3,123 10.1% -0.1% -1.3% Brasil 907 838 807 913 820 805 0.6% -2.1% -0.3% o/w Spectrum 700 0 0

OpFCF (Adj. EBITDA) 4,189 4,642 4,868 3,609 4,239 4,340 -13.8% -8.7% -10.9%

Net-Debt 24,800 23,620 22,146 26,375 25,738 25,027 6.4% 9.0% 13.0% Net-DebtExclusive / Adj. EBITDA 2.9x 2.7x use 2.5x 3.2x of 3.1x 3.0x 9.2% 14.5% 20.8% TIM Brasil (BRL) 2018 2019 2020 2018 2019 2020 2018 2019 2020

REV ENUES 17,077 18,053 19,044 17,015 17,641 18,343 -0.4% -2.3% -3.7% EBIT DA Sam MCHUGH6,464 7,035 7,562 6,505 6,898 7,246 0.6% -1.9% -4.2% CAPEX 4,059 4,056 4,055 4,084 3,969 4,045 0.6% -2.1% -0.3% OpFCF 2,405 2,979 3,507 2,422 2,929 3,201 0.7% -1.7% -8.7%

Source: Exane BNP Paribas estimates, Bloomberg, Based on TI Earnings and Net debt definitions

Change in price target As a result of our revised forecasts our price target falls from EUR0.60 to EUR0.38 – in terms of contributing factors updating for the latest FX rates has taken EUR0.04/share from our price target, revision to the underlying value of TIM Brazil (fair value from R$13 to R$11) reduce this by a further EUR0.04/share, whilst our revised Italian estimates reduce this further by EUR0.14/share.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 18

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Figure 13: We cut our price target from EUR0.6 to EUR0.38 Exane Change in TI Price Target Bridge

0.8

0.7 0.60 (0.04) 0.6 (0.04) (0.14)

0.5 0.38 0.4

0.3

0.2

0.1

0.0 Old Price Target Brazilian FX Underlying Brazil Italian Estimates New Price Target Valuation

Source: Exane BNP Paribas estimates

Figure 14: Our revised price target is 0.38 TI SOTP

Exclusive use of Sam MCHUGH

Source: Exane BNP Paribas estimates

Relative Valuation Our price target implies a ~4.7x EV/EBITDA multiple for TI’s Italian business, broadly similar to our target multiple for the group. A better indicator of value is we think unlevered FCF and FCF to equity. Adjusting for minority leakage we estimate that TI is currently trading on a ~3.1% unlevered FCF yield in 2019 and a 3.8% equity FCF yield in the same year. This puts the company on a significant premium vs their peer group on 5.1% and 6.9% respectively. Our price target implies TI trades on a ~3.7% unlevered FCF yield and a 7.2% equity FCF yield in 2019, with the latter being broadly in line with peers despite TI’s higher leverage and Brazilian exposure.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 19

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Figure 15: We think TI is materially more expensive than many think

Exclusive use of Sam MCHUGH

Source: Exane BNP Paribas estimates

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 20

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

The outlook for TI’s wireline business is worse than consensus thinks

Our structural test identify Italy and TI as facing the biggest risk to the fixed line outlook in all of Europe, critically we also believe that the catalysts necessary to crystallise this risk are beginning to play out. Some of these are relatively well understood i.e. the first order impact of Open Fiber’s FTTH rollout. However we think the market still underestimates 1) the disruptive impact of Sky’s (and potentially Iliad’s) broadband launches and 2) the negative impact FTTH rollout will have on broadband ARPUs and legacy voice telephony revenues. In addition we think the market has overlooked the negative impact of Iliad’s mobile launch on the fixed line market, as 1) falling mobile data prices serve to increase the attractiveness of mobile-only to the marginal broadband consumer (constraining ARPU upsell and penetration gains) and 2) the widespread availability of unlimited voice calls at a low price will accelerate declines in legacy voice telephony.

Italy screens as one of the most “at risk” fixed line markets in our structural review In this report we expand on the work presented in The Future of Fibre - Inglorious Masters by focusing on the potential for fixed service revenue growth for Italy and Telecom Italia in particular. Some of the terms and “tests” used in this analysis are taken from this note, and as a result we would redirect the reader to this note in order to get a better understanding of the methodology and rationale behind some of this analysis. Specifically we analyse Italy and TI through the lens of five key tests that are aimed at judging which European markets are most vulnerable to ARPH (average Exclusiverevenue use per household) of declines despite growth in fibre. Specifically we look at; i. Legacy revenue exposure; ii. Fixed line ARPH vs broadband speeds; Sam MCHUGHiii. Fixed line ARPH vs value for money from competing mobile networks; iv. Market concentration (HHI’s), and v. Fibre overbuild risk Across the “five tests” Italy consistently appears as one of the most ‘at risk’ countries in our review, specifically;

Test 1: Legacy revenue exposure; Italy has above average exposure to “legacy revenues”, which represent ~13% of total service revenue vs 12% for the European average, but especially when compared to income adjusted ARPH – on this basis TI has high legacy exposure with a limited opportunity to increase price to offset those legacy declines.

TI in particular screens as having high legacy revenues It is crucial to highlight that although a Country may be badly positioned in our analysis, this does not mean all the fixed operators in said Country are equally at risk. The burden may mostly be carried by the incumbent, or the cable operator, with alternative providers actually benefitting from of the woes of competitors. However Italy, due to historical quirks, is one of a few markets in Europe where there is no cable infrastructure and where the incumbent, TI, still has relatively high market share in both retail (45%) and on the wholesale level (95%).

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 21

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Figure 16: UK, Belgium and Italy may struggle to offset legacy declines via consumer broadband, while Portugal, Spain and Norway appear to over-earn on broadband. Finland and France offer the best mix. Legacy exposure in TSR (%) versus Income Adjusted ARPH (EUR/home)

19% "HIGH LEGACY EXPOSURE, Belgium 17% LIMITED BROADBAND GROWTH OPPORTUNITY"

Netherlands 15% Germany UK Italy 13% Sweden

11% Finland Europe

Portugal Spain

Legacy Legacy Revenues as % of TSR 9%

France "AVERAGE LEGACY 7% "LOW/AVERAGE LEGACY Norway EXPOSURE, BUT EXPOSURE, BROADBAND POTENTIALLY OVER- OPPORTUNITY" EARNING ON BROADBAND " 5% 20 30 40 50 60 70 80 90 Income Adjusted ARPH (EURm)

Source: Exane BNP Paribas estimates

Figure 17: TI has the highest exposure to legacy revenues of all incumbents under our coverage at 37% (!) Exclusiveof group EBITDA, perhaps use understandable of given they have 45% retail broadband share and 95% broadband infrastructure market share “Legacy Revenue” as a % of Domestic TSR and as a % of Incumbent Retail vs Wholesale Broadband Market Share Group EBITDA Sam MCHUGH 25% 100% IT TI FR 90% 20% DTTEF ES 80% UK 15% Telia Proximus DE PT KPN 70% Orange 10% BT Elisa 60% Telenor SE NO 5% NL Legacy Legacy Revenue as % Domestic TSR 50%

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 22

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Test 2: Monetisation: Fixed line ARPH vs broadband speeds; In our second test, we take our adjusted ARPH as calculated above, and compare it to the broadband speed. Crucially, we take the average speed of packages purchased, rather than the delivered speed – as this theoretically should correlate better with ARPH. On this metric it is clear that Italians are probably getting the worst value for money from their fixed line offering in all of Europe, with an average speed of less than 40Mbps and a PPP adj. ARPH of EUR65, i.e. there is little broadband ARPU upsell potential in Italy even as delivered speeds increase.

Figure 18: Italy is clearly over-earning in broadband relative to the speed of packages purchased by consumers Income Adjusted ARPH (EUR/month) versus average speed of packages purchased (Mbps)

Italy Belgium Spain 65 Norway Portugal

UK 60 "OVER-EARNING" 55 Europe

Sweden Netherlands 45

Germany France Exclusive40 use of JPN/SK have an ARPH of ~EUR55 for 500Mbps, and are selling 1Gbps for ~EUR40 on the frontbook.

Income Adjusted Fixed ARPH (EUR/m) 35 "UNDER-EARNING" Sam MCHUGHFinland 30 30 40 50 60 70 80 90 100 110 120 130 Average speed of broadband packages taken (Mbps)

Source: Exane BNP Paribas estimates

Test 3: Fixed line ARPH vs value for money from competing mobile networks; Our third test focused on looking at the value for money offered by mobile networks (by looking at the cost of 1GB of data, adjusted for the relative quality of the network) and to compare it to the PPP adjusted fixed ARPH. Looking at this data we immediately see that there is a pretty good correlation between these two factors i.e. mobile does seem to be a limiting factor on fixed ARPH. This makes some intuitive sense given mobile is a substitution product for fixed line broadband for the marginal consumer. So where does Italy sit on this chart? Unfortunately it is again an outlier with high fixed ARPH and relatively low mobile prices. This suggests mobile will be more of a cap on fixed line pricing and penetration in Italy relative to other markets.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 23

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Figure 19: Italy in particular seems to over-earn on broadband relative to the quality of mobile on offer Adjusted Fixed ARPH (EUR/month) versus mobile VFM

80 "SPAIN, ITALY , BELGIUM OVER- EARN IN BROADBAND GIVEN THE VFM OF MOBILE ON OFFER" 70 Italy Belgium Spain Norway

60 UK Japan

Europe South Korea 50 "FAIR BALANCE OF FIXED ARPH AND Netherlands

PPP Adjusted Fixed ARPH Sweden MOBILE VFM" France Germany 40

WORSE MOBILE VFM 30 Finland 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Mobile VFM Score: PPP Adj. EUR/month per GB adjusted for network quality (speed and availability) Source: Exane BNP Paribas estimates

Test 4: Market concentration (HHI’s), Italy falls behind only the UK and Spain in terms of the rate of decline in HHI in the broadband market i.e. competition is growing Exclusiveat one of the highest ratesuse in Europe. of

Figure 20: Sweden is very fragmented, while Netherlands/Belgium have high concentration, which regulators are trying to address through duopoly regulation HHI versus 3Sam year HHI CAGR (%) MCHUGH 2% "THERE IS NO EUROPEAN "HIGH HHI, CONCENTRATION

1% MARKET WITH LOWW HHI THAT GROWING" IS GROWING" Netherlands Finland 0%

Germany France (1%) Norway "JAPAN IS Belgium AT ~2,100" Europe Sweden (2%) Italy Portugal Regulators addressing NL/Belg HHI through cable reform (3%)

UK (4%)

(5%) "LOW HHI, CONCENTRATION Spain "HIGH HHI, CONCENTRATION DECLINING" DECLINING" (6%) 2,000 2,200 2,400 2,600 2,800 3,000 3,200 3,400 3,600 3,800 4,000

Source: Exane BNP Paribas estimates

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 24

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Test 5: CAPEX – Which countries have the most FTTH build/overbuild threat ahead of them?

Fibre overbuilds are emerging threats for many incumbents. We wrote at length on this risk specifically for Orange in our report I live in a house, a very big house in the country. When these overbuilds are rural in nature, like in the case of France, there is the dual threat of market share loss (as market share in rural areas for incumbents are much larger than nationwide averages) and wholesale revenue loss.

Figure 21: Italy, France and the UK are at significant threat of overbuild Fibre overbuild threat – Y-Axis = new overbuild threat, X-Axis = long-term overbuild percentage

110% NL | 57% BE | 50% 100% ES | 52% UK | 51% PT | 51% 90% FR | 52% NO | 60% 80% SE | 49% IT | 57%

(as (as total % homes) 70% DE | 51% 60% Total Total cable and other potential overbuild * BUBBLE = Incumbent % Market EBITDA 50% 0% 10% 20% 30% 40% 50% 60% 70% 80% Potential new overbuild threat as % total homes

Source: Exane BNP Paribas estimates Exclusive Lookinguse at the chart of above it is clear that Italy has by far the worst fibre overbulild threat in all of Europe, with Open Fiber targeting a build to at least 70% of Italian homes, aided by subsidies for rural areas. We remain sceptical of a (value creative for TI) deal Sam MCHUGHbetween Open Fiber and TI. There are multiple catalysts that we think will crystallise this risk Having identified Italy and TI as effectively structurally over-earning in fixed line the question becomes, what can change this? Critically we believe that 2018 will prove a tipping point, where the catalysts required to crystallise the earnings downgrades are beginning play out. Below we run through the catalysts as we see them, trying to comment as to whether we think they are well, or not well, understood by the market as we go.

Catalyst 1: Open Fiber – market share loss is relatively well understood In November 2015 we downgraded Telecom Italia from Buy to Neutral, in part due to the growing threat of overbuild from Open Fiber – we followed up on this in a later note in 2016 titled Bye, bye, Italian pie? The point here being that the Open Fiber risk has been around for a long time, and we think the market understands the first order impact of Open Fiber’s growth, i.e. Telecom Italia will face growing infrastructure competition, which will drag on TI’s wholesale revenues in the medium term as other operators move subscribers onto the Open Fiber network and off the TI network. That said there have been some critical developments in the last few months, notably that;

– In early August 2018 Open Fiber signed a EUR2.5bn project finance package to fund their network rollout, whilst owners Enel and the CDP are also said to be willing to grant a further EUR950m of funding. This removes one of the key investor pushbacks on Open Fiber – that they would not be able to secure financing. The company have now said that they will spend at least EUR1bn a year rolling out FTTH.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 25

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

– Whilst the Open Fiber network only currently has ~300k subscribers (half of which are Fastweb subscribers in Milan on the old Metroweb network) they now have commercial commitments from over 3m subscribers (with the likes of Vodafone signing up) and net-additions have started to accelerate.

Figure 22: We think the market does relatively well understand that growth at Open Fiber will come at the expense of Telecom Italia Structure of Italian Broadband Market, 2017, 2021 and 2025

Total Households – 25m 2017 Users Total broadband lines – 16.5m

Network used Telecom Italia – 15.1m OF FWA

Service provider Telecom Italia – 7.5m OLOs – 7.8m OLOs

TI Revenues Retail – EUR2.1bn Wholesale – EUR2.0bn

2021 Users Total broadband lines – 19.2m

Network used Telecom Italia – 16m OF – 2.3m FWA

Service provider Telecom Italia – 8m OLOs – 7.9m OLOs – 3.3m

TI Revenues Retail – EUR2.75bn Wholesale – EUR1.9bn

2025 Users Total broadband lines – 20.3m

Network used Telecom Italia – 14.5m OF – 4.6m FWA

Service provider Telecom Italia – 8.2m OLOs – 6.3m OLOs – 5.8m

TI Revenues Retail – EUR2.85bn Wholesale – EUR1.6bn

NB: OF= Open Fiber ExclusiveSource: Exane BNP Paribas estimates use of Catalyst 2: Open Fiber – the less well understood impact of FTTH rollout Whilst these immediate first order implications of Open Fiber’s rollout may be well Sam MCHUGHunderstood there are several second order implications which we believe are neither well understood by the market, nor included in consensus estimates for Telecom Italia. Specifically we believe consensus underestimates the negative impact on both ARPU and market share for TI.

2.1 Open Fiber rollout likely to cap TI’s broadband ARPU growth

One of the key reasons that FTTH is initially priced so low is that one of the critical factors behind the NPV of an FTTH build is penetration – and it is much easier to increase prices in the future once you have the uptake rather than the other way around. So generally across the globe we observe that FTTH is generally sold at a discount to backbook FTTC or DSL ARUs in the early years to stimulate uptake. And only when uptake is sufficient do FTTH providers look to increase prices. This augurs very poorly for TI’s ability to increase broadband ARPUs. This discrepancy is also very

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 26

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

large when we compare Italy to other European markets, specifically when we compare backbook to frontbook prices across Europe; we see that frontbook pricing in Italy is at the widest spread vs the backbook ARPUs of any country we looked at.

Figure 23: Italian Fixed ARPH is extremely high relative to the quality of the broadband packages sold, and as Open Fiber rolls out FTTH we expect ARPH to fall significantly Adjusted ARPH (EUR/month) versus average speed of packages purchased (Mbps)

80 Spain 75 "OVER-EARNING"

Italy Norway Belgium 65 UK Portugal

55 Europe

50 Netherlands Sweden 45

Germany France 40 Vodafone Italy are selling 1Gbps FTTH for

PPP Adjusted Fixed (EUR/m) ARPH Fixed Adjusted PPP EUR30/month, inclusive of unlimited domestic 35 and international calls to fixed and mobile "UNDER-EARNING" Finland 30 30 40 50 60 70 80 90 100 110 120 130 Average speed of broadband packages taken (Mbps) Source: ExaneExclusive BNP Paribas estimates use of Figure 24: Every country in Europe currently has mid-tier fibre prices that are at a discount to our estimate of the backbook ARPU – however Italian pricing has the biggest gap between the backbook and frontbook pricing Sam% discountMCHUGH between frontbook Entry level / mid-tier fibre offers versus backbook dual-play ARPUs

IT FR PT DE UK BE NK ES NO SE

-5 -5

-11 -15 -17

-23 -25 -25 -25 -28 -29 -32 -34 -34 -37

-45 -46 -48 Entry Medium -50 -50 2P FRONTBOOK DISCOUNT

Source: Exane BNP Paribas estimates

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 27

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

2.4 OF rollout will accelerate the decline of TI’s fixed voice revenues (EUR3.2bn)

Whilst there are not lots of precedents, if we look at the Japanese and Korean markets then it is very clear that the launch of FTTH coincides with a period of accelerating or sustained line losses and voice telephony revenues. Intuitively this makes sense – greater availability of higher quality broadband connectivity, especially today with the proliferation of VoIP services as well as video services (Skype, FaceTime etc) results in an acceleration in the decline in usage of legacy fixed voice.

Figure 25: As FTTH penetration accelerates, so does the loss in telephony lines and decline revenues In Japan NTT Commercialised FTTH in 2001, while in Korea KT launched FTTH in 2006 NTT Line Losses (m) vs. Japan FTTH Penetration (%) KT Telephony Rev Growth (%) vs. FTTH Penetration (%)

'03'05'07'09'11'13'15'17 2002 2004 2006 2008 2010 2012 2014 2016 3

0.2 56 83

-0.0 -1 -2 -1 -0.6 -3 -5 -1.1 -5 -1.3 -1.3 -1.7 -2.0 -10 -10 -2.5 -10 -11 -11 -11 -2.7 -12 -3.0 -13 -3.1 -13 -3.3 -3.4 -3.6 -3.7 Telephony Rev. Growth Effective FTTH/B Line Losses FTTH Pen. ExclusiveSource: Exane BNP Paribas estimates use of

Sam MCHUGHPart of the reason that ARPH is so high in Italy is, we think, due to the fact that TI has a lot of legacy revenue in voice (discussed in our analysis above). Whilst TI does disclose fixed voice revenues (EUR3.2bn in 2017) they do not split these between variable voice calling and line rental, translating into a voice ARPU of close to EUR30/month (inc VAT). It is clear that TI is making a lot of money from this legacy product.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 28

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Figure 26: We think Open Fiber’s FTTH rollout will cause a re-acceleration in TI’s voice line losses and revenue declines TI line losses vs Open Fiber Homes Passed TI Voice Revenue Change vs Open Fiber Uptake

2019 2020 2021 2022 2023 2024 2025 2026 2019 2020 2021 2022 2023 2024 2025 2026 0 60% 0% 40%

-50 55% -2% 35% -100 50% -150 -4% 30% 45% -200 40% -6% 25% -250 35% -300 -8% 20% 30% -350 -10% 15% -400 25%

-450 20% -12% 10% TI Line Losses ('000) OF Homes Passed as % of Italy TI Voice Revenue y/y OF Uptake (as % homes passed) Source: Exane BNP Paribas estimates

2.5 New entrant risk: The disruptive impact of Sky’s (and potentially Iliad’s) broadband launches are underestimated

One of the other key takeaways from our work on Japan is that FTTH rollout has led to a proliferation of retail competition. In the case of Italy we already have one confirmed new-entrant and one very obvious potential new entrant – Sky Italia and Iliad Exclusiverespectively. use of Sky has confirmed they will launch broadband launch on Open Fiber in summer 2019. Whilst Sky has confirmed their launch we think current consensus estimates for TI reflect little to zero impact of Sky launching broadband in Italy. We think this will Sam proveMCHUGH a mistake, and our analysis suggests that Sky has some of the right tools to build up a broadband business – although perhaps not to the same degree they have done in the UK (where they now have over 20% market share). First it is worth noting that Sky today has a customer base of just shy of 5m in Italy, which translates into something like 1 in 5 homes who have Sky TV, i.e. they are relatively well penetrated in Italy.

To judge the impact Sky could have we decided to rely on some of our work from our consumer survey STAMP 2018, One of the first indicators we always look at is Net Promoter Score. In Italy we immediately see that Sky has an NPS of +26, which far outscores both TI and Vodafone’s score in both fixed and mobile. I.e. Sky has a customer base that are more likely to be advocates for the brand, which augurs well for their potential to cross sell broadband into this TV customer base.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 29

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Figure 27: Sky has a resilient TV customer base in Italy, with high NPS scores relative to the Telcos Sky Italia Retail TV Customers Net Promoter Score (‘000 and as a % of Households)

4,900 30 24% 26 4,850 25 4,800 22%

4,750 20% 20 4,700 18% 15 4,650 12

4,600 16% 10 4,550 6 6 14% 5 4,500 5 12% 4,450 0 4,400 10% Sky TV TIM BB Italy BB Vodafone TIM Vodafone 2014 2015 2016 2017 2018 Mobile Mobile BB -2 Retail Customers as % Households -5

Source: Exane BNP Paribas estimates ,Exane STAPM 2018

We can go one step further by looking at the pricing power of Sky TV in Italy relative to TIM Vision and other OTT TV services (again for anyone unfamiliar with this analysis we refer you to our detailed STAMP 2018 report for a full explanation). This analysis shows us that Sky has good pricing power in TV, and is second only behind Netflix in Italy in terms of pricing power. Critically it is much better than TIM Vision.

Lastly we also have data from STAMP on customer overlap – this is a good way of Exclusive judginguse who is mostof and least exposed to Sky’s customer base, and therefore who would Sky’s broadband launch impact the most. We find that some 40% of Sky TV customers take broadband from TI, followed by 18% at Fastweb (who have aligned themselves with TI for FTTH in Italy, and not with Open Fiber), i.e. TI is Sam MCHUGHdisproportionately at threat from Sky’s launch vs other operators in Italy.

Figure 28: Sky Italia also has much better NPS scores vs TIM (and the existing broadband operators) TV Pricing Power Where do Sky TV customers get their broadband from?

30 Netflix 40% 25

10 Sky 18% 5 Mediaset 13% 0 11% 11% RELATIVE NPS -5 Amazon 6% Video -10 TIM -15

-20 -20% -10% 0% 10% 20% 30% 40% VFM SCORE

Source: Exane BNP Paribas estimates

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 30

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

We take a very cautious approach to forecasting growth at Sky and Iliad. In our forecasts we assume Sky reaches just 200k subscribes by 2022, or 4% of their TV customer base, and we do not currently forecast any subscribers for Iliad. We think this is a very conservative approach, and could be a source of further downside to our TI estimates. Even on these numbers if we revisit our earlier HHI analysis and roll forward our analysis to look at how HHI changes from 2017 to 2022 then we clearly see that Italy moves from being around the market average to being one of the worst markets over the next 4 years.

Figure 29: Our analysis suggest that Italy will undergo one of the biggest increases in retail competition across Europe over the next 4 years, and this is before we even factor in a broadband launch at Iliad Broadband HHI (2022) vs Change in Broadband (2022 vs 2017)

2% "HIGH HHI, CONCENTRATION GROWING" Finland 0% Germany Norway Netherlands

(2%) Sweden Portugal

(4%) France Europe (6%)

UK (8%) "JAPAN IS AT ~2,100" Italy (10%)

(12%) Belgium

Change Change in Broadband HHI 2022 vs 2017 (Exane est.) "LOW HHI, CONCENTRATION "HIGH HHI, CONCENTRATION (14%) DECLINING" Spain DECLINING"

(16%) 2,000 2,200 2,400 2,600 2,800 3,000 3,200 3,400 3,600 3,800 4,000 2022 HHI (Exane est) ExclusiveSource: Exane BNP Paribasuse estimates of Catalyst 3: Iliad – the overlooked impact of Iliad’s mobile launch on the broadband market Much has been made of Iliad’s mobile launch in the Italian market in May this year, and Sam MCHUGHas discussed later in this note, Iliad has probably been more aggressive on pricing than we anticipated pre-launch. Most of our discussions with investors and market commentary in relation to Iliad’s mobile launch have centred on the impact on mobile service revenue in Italy. However we believe that when we look back at their launch in 5 years’ time, we will also realise that their launch had a significant impact on the fixed line market in Italy. At the moment we think consensus totally overlooks this impact and is thus too optimistic in their outlook for fixed revenue growth in Italy.

3.3 Falling mobile data prices will serve to increase the attractiveness of mobile broadband to the marginal fixed broadband consumers – limiting fixed ARPU growth and penetration gains

As discussed in “Test 3” above, there is a good correlation between fixed ARPH and the cost of mobile data, with Italy screening as one of the big outliers where fixed line ARPH looks very high relative to the (quality adjusted) value for money offered by mobile plans.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 31

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Taking this analysis one step further, if we add Iliad’s current mobile offer into our analysis (adjusted for the network quality of Wind-Tre) then it becomes immediately clear that the relative attractiveness of a mobile-only home broadband solution to the marginal customer is growing in Italy, and in particular that it has grown materially following Iliad’s launch.

Furthermore the only other market in Europe with similar levels of mobile cost per GB is Finland, which is notable for two reasons;

i. In Finland ARPH is just ~EUR32 vs Italy at EUR65 today, i.e. taking this analysis to the extreme would imply there is significant downside to fixed ARPH in Italy

ii. In Finland broadband penetration is languishing just over 60%, vs Italy at ~67%, both lagging behind other European markets like France (95%), Germany (80%), Netherlands (94%) and the UK (90%).

In summary we think Iliad’s aggressive mobile offers are likely to put incremental downward pressure on both Fixed ARPH and broadband penetration in Italy. Now it’s worth saying that we are broadly sceptical of mass adoption of mobile as a substitute to fixed line broadband, and within our TI estimates we still assume broadband penetration continues to grow towards 80% by the mid-2020’s in Italy.

Figure 30: Taking our earlier analysis one step further we think that Iliad’s aggressive offers on mobile data will serve to constrain both broadband growth and penetration as mobile broadband becomes a more attractive proposition for the marginal consumer of fixed broadband Adjusted Fixed ARPH (EUR/month) versus mobile VFM

60 Our work Japan Sam MCHUGHsuggests that Italian fixed Europe ARPH could fall by up to 30-40%, South Korea 50 in part due to growing pressure Netherlands

PPP Adjusted ARPH Fixed Adjusted PPP from mobile as a substitute to fixed Sweden France Germany 40

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 32

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

3.4 The widespread availability of unlimited voice calls at EUR6.99/month will also accelerate fixed voice declines

As part of Iliad’s launch alongside giving consumers 30GB / 40GB of data for just EUR5.99 / EUR6.99 per month, the company also included unlimited voice calls. This is notable because Iliad is the first operator to really give away unlimited voice at very low prices (when Fastweb launched their aggressive offers at the back end of 2017 voice allowances were relatively small on their low data allowance offers). Now, as discussed above, TI’s fixed voice ARPU is EUR30/month – some EUR15.5 of this relates to the line rental costs, implying the average TIM voice subscriber is paying EUR15/month for calling packages from TIM – indeed an Italian consumer can take a voice line with unlimited calls for about EUR30/month.

Now if we compare this to the cost of a mobile plan with unlimited mobile data 2 years ago you would have had to pay in the region of EUR40-45/month for unlimited calls on a mobile plan. However over the last 2 years the entry level price of a mobile data plan with unlimited calls from either TI or Vodafone has fallen to just EUR11/month. This is far below the EUR30/month cost of a fixed line with unlimited calls at TI. To make matters worse Iliad is giving consumers unlimited fixed and mobile calls as part of their plans which are priced at just EUR5.99 or EUR6.99 a month.

In summary prior to Iliad’s launch the cost of unlimited voice calls on TI or Vodafone’s mobile network was close to 40% more expensive than getting unlimited voice calls from TI’s fixed network. However subsequent to the news in 2016 that Iliad would launch in mobile in Italy this premium has evaporated, such that today you can get unlimited voice calls (from mobile to mobile/landlines) from Vodafone for as little as EUR11/month as part of their Vodafone Smart plan with 2GB of data, i.e. a >60% discount to the equivalent price on TI’s fixed network. We believe this will also exert Exclusivestructural downward use pressure on TI’sof fixed voice ARPU’s.

Figure 31: We think mobile voice pricing has now fallen enough to have a growing impact on fixed voice revenue at TI Sam MCHUGHMonthly cost of TI fixed line with unlimited voice vs the cost of a mobile plan with unlimited calls (EUR/month)

20 Iliad's launch

EUR/ Month has futher 15 extended this gap 10

0 Aug-16 Aug-17 Aug-18 Iliad Cheapest TI or Vod mobile plan with Unlimited Voice Calls TI Fixed Voice ARPU (inc VAT)

Source: Exane BNP Paribas estimates

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 33

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Latest on Italian Mobile

In December 2017, we published our second deep dive note on Iliad’s launch in Italy, see FREEconomics 2 - Ho, Ho, Ho, in which we argued mobile prices in Italy were actually higher than first appears, implying that there was an opportunity for Iliad to disrupt the market. Given we have already written in-depth on this topic in prior notes, this section is focused on updating that view and providing a view on three downside risks to TI’s mobile business. These are 1) mobile prices have fallen further in July and August, 2) TI’s long term demographic problem and 3) the upcoming spectrum auction in Italy.

Despite much scepticism prior to launch, Iliad is disrupting the Italian market and prices have fallen further in July and August With little to no brand equity in Italy we were always of the view that they would have to focus on volume in the early stages of their launch. Iliad finally launched on May 29th its first mobile offer with 30Go of 4G data and unlimited voice for EUR5.99/month. Whilst the offer was only dedicated to one million SIMs initially, it got extended three times since then; 200k more on July 18th, 100k on July 26th and lastly 50k more on August 7th. In parallel, Iliad launched a second offer for 500k clients at EUR6.99/month but this time offering 40Go of 4G data and unlimited voice on July 26th.

Figure 32: Iliad’s (delayed) rapid scaling in Italy

Exclusive use of Sam MCHUGH

Source: Exane BNP Paribas estimates

MSR trends have already deteriorated in the market As expected MSR trends in 2Q18 already showed signs of deteriorating following Iliad’s May launch, with the overall market declining by -6.2% vs just a -1.9% decline in 1Q18. Within the mix TI has actually been more resilient than both Vodafone and Wind-Tre – this is consistent with our view and forecasts, in that all of our work in FREEconomics 2 - Ho, Ho, Ho showed that TIM has the best brand perception and best network in Italy, and as such would be least and last affected by Iliad’s entry.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 34

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Figure 33: Iliad has clearly already impacted the Italian mobile market after their May launch, with MSR growth falling to -6.2% in 2Q18 vs -1.9% in 1Q18 QoQ change in Mobile Service Revenue Growth

0% Q1 2017 Q2 2017 Q3 2017 Q4 2017 Q1 2018 Q2 2018

(2%)

(4%)

(6%)

(8%)

(10%) Telecom Italia Vodafone WIND-Hutchison Market (ex-Iliad)

Source: Exane BNP Paribas estimates

But investors should expect worse to come We have long tracked mobile pricing across Europe on a monthly basis, recording as much data as possible in our mobile pricing database. As we wrote in Freeconomics, mobile data prices have fallen precipitously following the announcement of Iliad’s launch; the average price of a 6 to 30GB package (we think representative of demand in the next 2-3 years) then we note those prices have more than halved since the beginning of 2017. However prices did stabilise somewhat between January and June Exclusive2018. use of Critically over the summer (July and August 2018) we have seen prices fall even further, such that 1GB of data which used to cost EUR4/month for a subscription Sam MCHUGHbetween 6 and 30GB, are now offered around at EUR0.8/month per GB. This is still ~5x above Iliad’s currently offered rate of 40GB for EUR6.99/month, i.e. investors should expect worse to come.

Figure 34: Price per GB in Italy has more than halved in the last 18 months Price per GB trend per data bucket cohorts*

Average Cost per GB in a 6-30GB Plan Average Price for a 6-30GB Plan May 16 May 17 May 18 Aug 16 Aug 17 Aug 18 Dec 16 Dec 17 Dec Feb Feb 17 Feb 18 Sep 16 16 Nov Sep 17 17 Nov Mar16 Mar17 Mar18 Jun 16 Jun 17 Jun 18 Jun Apr 16 Apr 16 Oct 17 Apr 17 Oct 18 Apr Jan 17 Jan Jan 18 Jan Jul 16 Jul Jul 17 Jul Jul 18 Jul

4.5 40.0 Iliad still priced 4.0 Pricing pressures 35.0 re-accellerated a 50% discount 3.5 this summer 30.0 3.0 25.0 2.5 20.0 2.0 15.0 1.5 10.0 1.0

0.5 ILIAD ITALIA 5.0 ILIAD ITALIA

0.0 0.0 Source: Exane BNP Paribas estimates * note these are 3 month trailing average prices

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 35

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Network quality remains clearly superior at Telecom Italia relative to Wind/Tre (and Iliad’s roaming) or Vodafone… but trying to maintain a premium pricing as unit prices are heavily cut down is tough. Nonetheless, we expect Telecom Italia to be the last and least impacted by Iliad’s launch and mobile service revenue trends to remain more robust and eventually improve ahead of competition. However given in 3Q the operators will face a full quarter of TI’s offers in the market, and that prices have come down further, we expect to see another step down in MSR trends.

Figure 35: TI MSR growth has been more robust… and we expect this trend to continue despite pressures MSR Growth Rate YoY%

2Q13 4Q13 2Q14 4Q14 2Q15 4Q15 2Q16 4Q16 2Q17 4Q17 2Q18 4Q18E 2Q19E 4Q19E 2Q20E 4Q20E 2Q21E 4Q21E 2Q22E 4Q22E 5%

(5%)

(10%)

(15%)

(20%) Exclusive(25%) use of Italy Hutchison-WIND Telecom Italia Vodafone Sam Source:MCHUGH Exane BNP Paribas estimates The “Mobile Millennial” demographic problem Our second observation on the mobile market is that while prices coming down is clearly bad for mobile trends, there is a longer term issue for Telecom Italia. In our earlier work, see The Mobile Millennial, we identified TI as being one of the companies in our coverage who over-index to Baby Boomers.

Today this is in itself not a bad thing, they tend to have higher disposable income and tend to be happy customers less likely to switch. Incumbents have benefitted greatly from this generation. However over-reliance is also a problem if these high spending subscribers are not replaced over time. And this is where TI have a small problem, as Baby Boomers today represent a disproportionate share of their revenue, and with Wind-Tre (and likely Iliad too) over-indexing to Millennials, TI faces a growing headwind every year to maintain their mobile market share.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 36

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Figure 36: Telecom Italia did a better job for Millennial last year but Fastweb and Iliad should catch up Higher Price = More expensive. Scores are standardised where (Q,P) = (1,1) indicates average quality and price perception Mobile Network Quality vs. Price Perception | Millenials Telecom Italia Mobile Subs Distribution

1.5 Generat i ons | TIM Today 2018S 2017S 25.6% M I L L E N N I A L 1.3 1982 - 1995 Vodafone

Fastweb 1.0 TIM PRICE G E N X Three 25.3% Wind 1965 - 1981

0.8

31.3% B O O M E R 0.5 0.5 0.8 1.0 1.3 1.5 1946 - 1964 QUALITY Source: Exane BNP Paribas estimates

Figure 37: Telecom Italia over-index to Baby Boomers, and under-index to Millennials Current Consumer Share by Provider & Demographic Current Consumer Market Share ITALY | Share by Demographic ITALY | Consumer Share Now TIM Vodafone Three Wind MVNO Exclusive use of 8%

32% 20% 21% 18% TIM Sam MCHUGH12% 10% Vodaf one 21% Thr ee 12% 31% 26% Wind 24% MVNO 11%

31% 34% 30% 30% Millennial Gen X Boomer Source: Exane BNP Paribas estimates

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 37

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Upcoming Italian Spectrum Auction The last point of note on the mobile market is the upcoming spectrum auction that will see the Italian government auction off 700MHz, 3.6GHz and 26.5GHz spectrum in 2H2018. The authorities are hoping to raise EUR2.5bn for the sale.

In short we have increased the expected cost for TI to EUR700m FY2018;

– Firstly if we look at spectrum ownership in Italy then TI does not have a spectrum advantage vs peers, and as a result we don’t think TI can sit out the auction. – Secondly now that Hutchison is taking over the Wind-Tre partnership we would expect them to be more aggressive than would otherwise have been the case. – Third – we understand at least 2 x 10MHz of the 700MHz spectrum will be reserved for a new entrant. – Reports indicate that seven telcos have registered for the auction, including Fastweb (MVNO on TI’s network) and Open Fiber (who have ambitions to build a 5G network and roll out FWA). Finally we note that the 700MHz may not be available to use until 2022 and as such it is possible the payment will come later.

Exclusive use of Sam MCHUGH

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 38

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

The Sweet Spot for TIM Brasil is Over

Brazil may represent only 13% of TI’s EV, but in terms of market cap it is far more important at around a third. Given the importance of Brazil it is worth taking a little time to consider what is happening in this market to understand if there is any upside or downside risk. In short we believe that TIM Brasil has been in a real sweet spot for the last 18 months as multiple factors have contributed to their success. Whilst consensus estimates seem to be extrapolating current trends into the future, we believe all of these factors are now reversing and as a result we are more cautious on the outlook for TIM Brasil vs consensus – specifically our OpFCF estimates are 9% below Bloomberg consensus by 2020 and our implied price target stands at R$11 vs TIM’s current spot price of R$12 and our previous implied valuation of R$13 – i.e. we believe there is downside risk to consensus estimates, but it does appear that the market is somewhat pricing this in.

Why was TIM in such a sweet spot? To understand where TIM Brasil is going, we first have to understand where they have come from. If we look at the last 18 months in Brazil then TIM has enjoyed a strong reversal in fortunes with MSR growth, which has gone from close to -7% in Q2 2016 to around +6% over the last 12 months. We believe that there are three significant factors that explain this change in fortunes;

1) An early mover advantage in 4G. To their credit TIM were early in expanding their 4G coverage in Brazil, going from ~55% coverage in 2016 to close to 75% today. Critically this has seen them create “white space” between themselves and their rivals – this superior network coverage has seen them outperform their rivals in terms of post- Exclusive paiduse net-adds, transla ofting into superior MSR growth.

Figure 38: As TIM accelerated ahead of others in terms of 4G network coverage their MSR trends accelerated ahead of rivals Sam MCHUGHMobile Service Revenue Growth 4G network availability

8% 80% 6% Between 2016 and 75% today TIM has 4% opened up a gap 70% vs competitors in 2% terms of 4G 65%

0% 60% Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 2016 2016 2016 2017 2017 2017 2017 2018 2018 (2%) 55%

(4%) 50%

45% (6%) 40% (8%) Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 TIM Brazil Brazil Market 2016 2016 2016 2017 2017 2017 2017 2018 2018 Vivo TIM Claro Oi

Source: Exane BNP Paribas estimates, Opensignal

2) Oi’s bankruptcy problems effectively made this a ”3.5” and not a 4 player market. In June 2016, Brazilian incumbent Oi entered bankruptcy protection and begun trying to restructure their balance sheet and business. Over this period of time their market share has continue to fall, and ultimately they have been a big market share donor to their rival operators.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 39

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

3) Improving macro conditions. Lastly in Emerging Market telcos, market growth tends to be more closely correlated to GDP growth (telco spend is a greater proportion of PPP adjusted household income and is thus more sensitive to changing economic conditions).

Figure 39: The whole Brazilian mobile market has been supported by improving macro conditions / GDP growth Brazil GDP growth vs Mobile Service Revenue Growth

0% Q4 2015 Q1 2016 Q2 2016 Q3 2016 Q4 2016 Q1 2017 Q2 2017 Q3 2017 Q4 2017

(2%)

(4%)

(6%)

(8%) GDP Growth (%y/y) Brazil MSR Growth

Source: Exane BNP Paribas estimates

What has changed? Unfortunately for TIM we believe we are at a turning point where these 3 factors are Exclusivebeginning use to unwind. of

1) AMX (Claro) has fought back and looking forward we expect Vivo (Telefonica) to fight back too, so life will get tougher for TIM. Having been a Sam MCHUGHlaggard behind both TIM and Vivo, AMX was quietly working behind the scenes to improve its network, and over the last two years has made significant investments in “4.5G” technology (carrier aggregation, 4x4 MIMO, etc.), and in mid-2017 it finally launched its commercial “4.5G” offers.

If we look at Open Signal (crowd-sourced speed test data) then we can clearly see that in 2016 Vivo was the market leader in terms of 4G coverage and 4G speeds. It has not stood still, and in the latest speed test in mid-2018 Vivo has modestly improved both speeds and coverage on 4G. However, as discussed above TIM has clearly improved materially in terms of 4G availability (a proxy for coverage), whilst on the other hand Claro has made the biggest gains – becoming the clear leader in terms of speeds (average speeds on 4G are now close to 30Mbps vs Vivo a shade over 20Mbps) and having narrowed the coverage gap (it is now in-line with Vivo). Looking forward we do expect to see more network sharing in Brazil, which should only serve to reduce network quality differentiation, with some offset coming from lower capex.

This is now having a clear impact on market dynamics – in 2Q18 AMX grew its Brazilian Mobile Service Revenue by +12%, whilst TIM also reported broadly stable MSR growth of +5.7%. On the flip side we expect Vivo’s mobile service revenue growth to slow to 2.1% vs +4.3% over the course of 2017.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 40

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Figure 40: America Movil has delivered the biggest improvement in network quality over the last two years and is now outgrowing Telefonica and Telecom Italia 4G Availability vs 4G Speed Mobile Service Revenue Growth Y/Y (mid 2016 vs mid 2018)

80% 15% 75% TIM 2018 10% 70% 5% 65% Vivo 2018 Claro 2018 0% 60% Oi 2018 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 2016 2016 2016 2016 2017 2017 2017 2017 2018 2018 (5%) 55% TIM 2016 Vivo 2016

Average 4G Availability Claro 2016 50% (10%)

45% Oi 2016 (15%)

40% (20%) 5.0 10.0 15.0 20.0 25.0 30.0 Telefónica TIM Brazil Average 4G Speed (Mbps) América Móvil Oi

Source: Exane BNP Paribas estimates

To illustrate this shift further let’s look at total mobile service revenue share. Firstly, it is clear how dominant Vivo is, with over 43% mobile service revenue market share in Brazil, followed by TIM at ~25%, Claro on 20% and Oi with the remaining 12%. However, if we look at the change in market share over time, then we can immediately Exclusivesee Vivo is useno longer gaining of share, and we expect it to lose ~1% of MSR share y/y in 2Q18, as Claro is now winning >1% share of MSR y/y.

Figure 41: TEF (Vivo) is the clear market leader in terms of MSR share, but has begun to lose share in recentSam quarters as AMXMCHUGH (Claro) has been resurgent Mobile Service Revenue Market Share Y/Y Change in Mobile Service Revenue Share

50% 3.0%

45% 2.5%

40% 2.0%

35% 1.5%

30% 1.0% 25% 0.5% 20% 0.0% 15% Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 (0.5%) 2016 2016 2016 2016 2017 2017 2017 2017 2018 2018 10% (1.0%) 5% (1.5%) 0% Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 (2.0%) 2016 2016 2016 2016 2017 2017 2017 2017 2018 2018 Telefónica TIM Brazil (2.5%) Telefónica TIM Brazil América Móvil Oi América Móvil Oi

Source: Exane BNP Paribas estimates

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 41

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

This is consistent with our STAMP survey data which showed Claro fighting back. In our 2018 STAMP consumer survey we found that Vivo is perceived as having a premium service with a good value proposition (below the average line in the figure below) though its edge over Claro in terms of quality has again narrowed year over year (shifted further to the left), while TIM has improved alongside the fair value-for- money line. Clearly differentiation between the networks has narrowed, with Claro and TIM both closing the gap to Vivo. Oi remains, as ever, the laggard.

Figure 42: Vivo still enjoys the best consumer perception, but Claro and TIM closed the gap in our 2018 survey Higher Price = More expensive. Scores are standardised where (Q, P) = (1,1) indicates average quality and price perception 2017 Mobile Network Quality vs. Price Perception 2018 Mobile Network Quality vs. Price Perception

1.5 1.5

1.3 1.3 Vivo Vivo

Oi Claro Oi Claro 1.0 1.0 PRICE PRICE TIM TIM

0.8 0.8

0.5 0.5 0.5 0.8 1.0 1.3 1.5 0.5 0.8 1.0 1.3 1.5 QUALITY QUALITY ExclusiveSource: Exane STAMP 2018. Respondents forceuse ranked providers of by price and quality, these are subsequently converted into a standardised score.

Our pricing power data (VFM score vs relative NPS) shows that Vivo’s edge in 2018 has narrowed, especially vs Claro which in 2018 screened as having pricing power for Sam MCHUGHthe first time. Figure 43: Vivo’s pricing power is eroding as Claro in particular has caught up Pricing Power & Retail Churn Predictor 2018 STAMP Pricing Power Screen STAMP Churn Predictor 2018

40 2018S 2017S 2018S 30 2017S 22% 20 20% Vivo 10 Claro 17% 17%

0 13%

-10 RELATIVE NPS TIM

-20 Oi

-30

-40 -30% -20% -10% 0% 10% 20% 30% VFM SCORE Oi TIM MARKET Claro Vivo

Source: Exane STAMP 2018

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 42

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Given Vivo is now lagging behind both AMX and TIM, we would expect them to attempt to address this situation at some point soon, especially with Oi also coming back to the market more aggressively in 2H18 and into 2019 (see below). As a result we believe that the unique first mover advantage that TIM enjoyed will be short lived, and competition will increase going into 2019.

1) Oi set for a modest improvement relative to the last 4 years. Brazilian incumbent Oi has finally emerged from bankruptcy protection and is expected to firm up plans for a BRL4bn capital raise at some point in September 2018. Whilst they are not out of the woods, it is clear that the worst is behind them. In addition numerous news sources1 have suggested that the new shareholders are focused on improving the operations of the company, rather than thinking of selling their stakes. Having been a big share donor to everyone else (their postpaid wireless market share has fallen from 15% in 2012 to close to less than 10% today) we expect the rate of share loss to ameliorate from here (although we still forecast them losing a further 160bps of share over the next 3 years). This will just be one of the contributing factors to a more competitive mobile market.

2) Economic and political uncertainty reaching a critical juncture. With a presidential election scheduled for October 2018, their remains significant political uncertainty in Brazil (especially when Lula, the ex-president is leading in the opinion polls, whilst also serving a jail sentence for corruption). On top of this, the country is facing somewhat of a fiscal cliff-edge (the fiscal deficit is still above 5% of GDP); the economic outlook for Brazil is uncertain at best. Whilst it is of course very difficult to call, we do think that level of macro improvement seen through 2017 and early 2018 is unlikely to be sustained into 2019. Indeed we have already seen GDP growth slow to +1.2% in Q1 2018 and it is expected to have slowed further to +1% in 2Q 2018. We expect this to drag on MSR revenue trends as we progress through 2018 and into 2019. In addition low interest rates are clearly very favourable for the government debt situation, suggesting the currency is unlikely to receive much support from a change in Exclusive rates.use of

Figure 44: The exchange rate has been hurting the value of the Brazilian assets of TI, whilst the macro- sensitive prepaid market is coming under more pressure – this is making Postpaid more important Sam BRL/EURMCHUGH exchange rate and y/y change Prepaid MSR as % total Brazilian MSR and y/y change in Prepaid MSR

45% 0% 6.0 40.0% 40% -1%

5.0 30.0% 35% -2%

20.0% 30% -3% 4.0 25% -4% 10.0% 3.0 20% -5% 0.0% 15% -6% 2.0 -10.0% 10% -7%

5% -8% 1.0 -20.0% 0% -9% 0.0 -30.0% Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 2016 2016 2016 2016 2017 2017 2017 2017 2018 2018 Prepaid MSR as % total MSR (LHS) LTM Change in Prepaid MSR(RHS) BRL %y/y: BRL

Source: Exane BNP Paribas estimates

1 https://www.theepochtimes.com/brazils-oi-owners-to-focus-on-improving-business-before-selling-stakes_2632646.html

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 43

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Investment case, valuation and risks

Telecom Italia (Underperform, Target Price EUR0.38) Investment case The entry of new rivals (Iliad and Open Fiber) will prove a trigger to crystallising significant earnings downgrades at TI, as growing retail and infrastructure competition accelerates the decline in legacy revenues, while new services will be insufficient to offset these declines. In Brazil TI has operated in a unique sweet spot for the last 18 months, but we believe the tide is turning, leaving us below consensus in Brazil. All together our estimates imply TI will miss their medium term cumulative FCF guidance targets by 30% and leverage, as measured by Net-Debt to EBITDA will not budge from ~3.0x EBITDA.

Valuation methodology We value TI’s ordinary shares using an SOTP analysis of the different units, using a DCF for the domestic and Brazilian businesses and market value for Inwit.

Risks To the upside: Main upside risks include: 1) positive earnings surprise as management delivers ahead of the current cost-cutting plan; 2) renewed prospects of consolidation in Brazil; 3) a takeover of the company by another industrial player. 4) A value creative deal with Open Fiber.

To the downside: Main downside risks include: 1) further evidence of the credibility of Open Fiber’s plans; 2) greater impact from Iliad's entry / a more aggressive response from Vodafone, 3) the launch of an Iliad broadband offer.

Exclusive use of

Sam MCHUGH

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 44

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

DISCLOSURE APPENDIX

Analyst Certification I, Sam McHugh, (authors of or contributors to the report) hereby certify that all of the views expressed in this report accurately reflect my personal view(s) about the company or companies and securities discussed in this report. No part of my compensation was, is, or will be, directly, or indirectly, related to the specific recommendations or views expressed in this research report.

Non-US Research Analyst Disclosure The research analysts named below were involved in preparing this research report. Research analysts at Exane Limited and Exane SA are not associated persons of Exane Inc. and thus are not registered or qualified in the U.S. as research analysts with the Financial Industry Regulatory Authority (FINRA) or the New York Stock Exchange (NYSE). These non-U.S. analysts are not subject to the NASD Rule 2241 and NYSE Rule 472 restrictions on communications with a subject company, public appearances and trading securities held by a research analyst account.

Sam McHugh Exane Limited

Exane SA is regulated by the Autorité des Marchés Financiers (AMF) in France, Exane Limited is authorised and regulated by the Financial Conduct Authority in the United Kingdom, and Exane Inc. is regulated by FINRA and the U.S. Securities and Exchange Commission in the United States.

Research Analyst Compensation The research analyst(s) responsible for the preparation of this report receive(s) compensation based upon various factors including overall firm revenues, which may include investment banking activities.

Research Analyst-Specific Disclosures The research analyst(s) responsible for the preparation of this report (or members of their household) may have a relationship with the companies covered by this research report, as described in the numbered disclosures below. The table immediately below indicates which, if any, of these disclosures apply to the research analyst(s) responsible for preparation of this research report.

Research Analyst(s) Companies Disclosures NONE

1 – The research analyst(s) responsible for the preparation of this report or a member of his/her household has/have a financial interest in the securities of the subject company/ies, as indicated in the previous table. 2 – The research analyst(s) responsible for the preparation of this report or a member of his/her household serve(s) as an officer, director or advisory board member of the subject company/ies indicated in the previous table. 3 – The research analyst(s) responsible for the preparation of this report received compensation from the subject company/ies indicated in the previous table in the past twelve months.

Exane-Specific Regulatory Disclosures Exane SA, Exane Limited and Exane Inc. (collectively, “Exane”) may have a relationship with the companies covered by this research report, as described in the numbered disclosures below. The table immediately below indicates which, if any, of these disclosures apply to Exane’s relationship with the subject company/ies.

Companies Disclosures NONE Exclusive use of

1 – Exane beneficially owns 1% or more of any class of common equity securities of the subject company/ies. 2 – Exane managed or co-managed an offering of Equity securities for the subject company/ies in the past 12 months. 3 – Exane received compensation for investment banking services from the subject company/ies in the past 12 months (the only investment banking services for Exane with regards to the subject company/ies are those when Exane is distributor or underwriter for Equity securities offerings managed-or co-managed by BNP Paribas, whenSam BNP Paribas managed orMCHUGH co-managed an offering of Equity securities for the subject company/ies). 4 – Exane expects to receive or intends to seek compensation for investment banking services from the subject company/ies in the next 3 months. 5 – Exane SA and/or Exane Limited are/is a market maker and/or liquidity provider in the securities of the subject company/ies. 6 – Exane Inc. received compensation for products and services other than investment banking services from the subject company/ies in the past 12 months. 7 – Exane Inc. had an investment banking services client relationship with the subject company/ies in the past 12 months. 8 – Exane Inc. had a non-investment banking, securities-related client relationship with the subject company/ies in the past 12 months. 9 – Exane Inc. had a non-securities-related services relationship with the subject company/ies in the past 12 months. 10 – Exane Inc. is a market maker in the securities of the subject company/ies. 11 – Exane beneficially owns at least 0.5% long or short position of the subject company/ies. 12 – Sections of this report, with the research summary, target price and rating removed, have been presented to the subject company/ies prior to its distribution, for the sole purpose of verifying the accuracy of factual statements. 13 – Following the presentation of sections of this report to this subject company, some conclusions were amended.

Commitment to transparency on potential conflicts of interest: BNP Paribas While BNP Paribas (“BNPP”) holds a material ownership interest in the various Exane entities, Exane and BNPP have entered into an agreement to maintain the independence of Exane's research reports from BNPP. These research reports are published under the brand name “Exane BNP Paribas”. Nevertheless, for the sake of transparency, we separately identify potential conflicts of interest with BNPP regarding the company/(ies) covered by this research document.

BNP Paribas-related disclosures BNPP may have a relationship with the companies covered by this research report, as described in the numbered disclosures below. The table immediately below indicates which, if any, of these disclosures apply to BNPP’s relationship with the subject company/ies.

Companies Disclosures Telecom Italia 4; 5

1 – BNPP beneficially owns 1% or more of any class of common equity securities of the subject company/ies 2 – BNPP managed or co-managed an offering of Equity securities for the subject company/ies in the past 12 months 3 – BNPP acted as Advisor in a Public Offer involving the subject Company/ies in the past 12 months. 4 – BNPP received compensation for investment banking services from the subject company/ies in the past 12 months 5 – BNPP expects to receive or intends to seek compensation for investment banking services from the subject company/ies in the next 3 months 6 – A member of senior BNPP management is a member of the Board of the subject company 7 – BNPP beneficially owns at least 0.5% long or short position of the subject company/ies.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 45

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Explanation of Research Ratings Stock Rating Exane’s Ratings are relative ratings defined against the performance of the MSCI Index sectors.

Outperform (O/P): The stock is expected to outperform the stock’s MSCI sector over a 12-month investment horizon. Neutral: The stock is expected to perform in line with the performance of the stock’s MSCI sector over a 12-month investment horizon. Underperform (U/P): The stock is expected to underperform the stock’s MSCI sector over a 12-month investment horizon. Under review: The rating of the stock has been placed under review after significant news. Any possible change will be confirmed as soon as possible in the form of a new broadly disseminated report Restricted (RS): The stock is covered by Exane but there is no Rating and no Target Price because Exane is involved in an equity capital market transaction relating to the subject company. Not Rated (NR): The stock is covered by Exane but there is no Rating and no Target Price at this time. Not Covered (NC): Exane does not cover this company.

Distribution of Exane BNP Paribas’ equity recommendations As at 02/07/2018 Exane BNP Paribas covered 578 companies. The companies that, for regulatory reasons, are not accorded a rating by Exane BNP Paribas are excluded from these statistics. For regulatory reasons, our ratings of Outperform, Neutral and Underperform correspond respectively to Buy, Hold and Sell; the underlying signification is, however, different as our ratings are relative to the sector.

38% of the companies covered by Exane BNP Paribas were rated Outperform. During the last 12 months, Exane acted as underwriter and/or distributor for BNP Paribas on 11% of the companies with this rating for which BNP Paribas acted as manager or co-manager in an offering of equity securities. BNP Paribas provided investment banking services to 74% of the companies accorded this rating*.

44% of the companies covered by Exane BNP Paribas were rated Neutral. During the last 12 months, Exane acted as underwriter and/or distributor for BNP Paribas on 6% of the companies with this rating for which BNP Paribas acted as manager or co-manager in an offering of equity securities. BNP Paribas provided investment banking services to 72% of the companies accorded this rating*.

18% of the companies covered by Exane BNP Paribas were rated Underperform. During the last 12 months, Exane acted as underwriter and/or distributor for BNP Paribas on 7% of the companies with this rating for which BNP Paribas acted as manager or co-manager in an offering of equity securities. BNP Paribas provided investment banking services to 73% of the companies accorded this rating*.

* Exane is independent from BNP Paribas. Nevertheless, in order to maintain absolute transparency, we include in this category transactions carried out by BNP Paribas independently from Exane. For the purpose of clarity, we have excluded fixed income transactions carried out by BNP Paribas.

Exclusive use of Sam MCHUGH

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 46

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

Price and Ratings Chart

Telecom Italia

Historical closing price & target price (as of 31/08/2018)

Source: Exane BNP Paribas

Historical rating & target price changes Date & Time of Rating Target Price Closing Price Key changes Report and analyst covering the stock dissemination (UTC) 3 Aug. 2018 04:29 = EUR0.60 EUR0.64 TELECOM OPERATORS: Valuation Station - Magic Roundabout (S. McHugh) 23 Jul. 2018 15:42 = EUR0.60 EUR0.61 TP down TELCO OPERATORS: TI & TEF - Shifting Tides in Brazil (S. McHugh) 3 Jul. 2018 05:29 = EUR0.70 EUR0.65 TELECOM OPERATORS: Valuation Station - Is this the start of something? (S. McHugh) 5 Jun. 2018 05:21 = EUR0.70 EUR0.69 TP down TELECOM OPERATORS: Valuation Station - A 'blip'? (S. McHugh) 4 Jun. 2018 05:21 = EUR0.72 EUR0.70 TELECOM OPERATORS: Question Time (S. McHugh) 30 May 2018 05:17 = EUR0.72 EUR0.70 TELECOM OPERATORS: European Phonebook - Hibernation (S. McHugh) 2 May 2018 04:43 = EUR0.72 EUR0.84 TELECOM OPERATORS: Valuation Station - Every Dog Has Its Day (S. McHugh) 9 Apr. 2018 05:09 = EUR0.72 EUR0.85 TELECOM OPERATORS: How has STAMP changed our estimates? (S. McHugh) 4 Apr. 2018 04:47 = EUR0.72 EUR0.76 TP up TELECOM OPERATORS: Valuation Station - The Never-ending Story (S. McHugh) 19 Mar. 2018 05:42 = EUR0.69 EUR0.81 TELECOM OPERATORS: Question Time (S. McHugh) 14 Mar. 2018 06:00 = EUR0.69 EUR0.78 TELECOM OPERATORS: Euro Phonebook - We are unable to connect your call (S. McHugh) 7 Mar. 2018 06:25 = EUR0.69 EUR0.79 TELECOM OPERATORS: STAMP 2018: Brand Royale (S. McHugh) 7 Mar. 2018 06:22 = EUR0.69 EUR0.79 TELECOM OPERATORS: 1, 3 and 5-year view (S. McHugh) 5 Mar. 2018 06:51Exclusive = EUR0.69 EUR0.73 useTELECOM OPERATORS: of Valuation Station - The Pretender (S. McHugh) 2 Feb. 2018 11:02 = EUR0.69 EUR0.71 TELECOM OPERATORS: Valuation Station - New Year, Same Old Telcos (S. McHugh) 5 Jan. 2018 06:01 = EUR0.69 EUR0.74 TELECOM OPERATORS: Valuation Station - Flourish or Flounder? (S. McHugh) 30 Dec. 2017 23:00 = EUR0.69 EUR0.72 TP down (S. McHugh) 6 Dec. 2017 06:30 = EUR0.69 EUR0.73 TP down TELCO OPERATORS: FREEconomics 2 - Ho, Ho, Ho (S. McHugh) 4 Dec. 2017 05:58 = EUR0.80 EUR0.71 TELECOM OPERATORS: Valuation Station - Love the Laggard (S. McHugh) 20 Nov. 2017 06:09 = EUR0.80 EUR0.68 TELECOM OPERATORS: European Phonebook - Digital Times (S. McHugh) 13 Nov. 2017 14:48 = EUR0.80 EUR0.67 TP down TELECOM ITALIA: Non material data changes (S. McHugh) 10 Nov. 2017 17:39 Sam = EUR0.80 MCHUGH EUR0.71 TP down TELECOM ITALIA: A Waiting Game (S. McHugh) 3 Nov. 2017 05:45 = EUR0.85 EUR0.76 TELECOM OPERATORS: Valuation Station - No News is Good News? (S. McHugh) 3 Oct. 2017 04:50 = EUR0.85 EUR0.78 TELECOM OPERATORS: Valuation Station - Treading Water (S. McHugh) 13 Sep. 2017 15:40 = EUR0.85 EUR0.78 EUROPEAN TELECOMS: The Mobile Millennial (S. McHugh) 11 Sep. 2017 16:03 = EUR0.85 EUR0.79 THE FUTURE OF TV: WARNING: This report is strong on Content (S. McHugh) 4 Sep. 2017 15:47 = EUR0.85 EUR0.78 TELECOM OPERATORS: Valuation Station - Down with the Trumpets (S. McHugh) 4 Sep. 2017 05:20 = EUR0.85 EUR0.78 TELECOM OPERATORS: European Phonebook - Travelling South for the ... (S. McHugh) The initiation report for this Company was published prior to 2005. Due to system constraints we are unable to provide historical data as detailed in the table above, for the period prior to 1 August 2005.

The latest company-specific disclosures, valuation methodologies and investment case risks for all other companies covered by this document are available on http://cube.exane.com/compliance.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 47

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

LONDON PARIS FRANKFURT GENEVA Exane Limited Exane S.A. Branch of Exane S.A. Branch of Exane S.A. 1 Hanover Street 6 Rue Ménars Europa-Allee 12, 3rd floor Rue du Rhône 80 London W1S 1YZ 75002 Paris 60327 Frankfurt 1204 Geneva UK France Germany Switzerland Tel: (+44) 207 039 9400 Tel: (+33) 1 44 95 40 00 Tel: (+49) 69 42 72 97 300 Tel: (+41) 22 718 65 65 Fax: (+44) 207 039 9440 Fax: (+33) 1 44 95 40 01 Fax: (+49) 69 42 72 97 301 Fax: (+41) 22 718 65 00

MADRID MILAN NEW YORK SAN FRANCISCO Branch of Exane S.A. Branch of Exane S.A. Exane Inc. Exane Inc. Calle Génova, 27 Via dei Bossi 4 640 Fifth Avenue 201 Mission Street 7th Floor 20121 Milan 15th Floor San Francisco, CA 94105 Madrid 28004 Italy New York, NY 10019 USA ExclusiveSpain Tel: (+39) use 02 89 63 17 13 of USA Tel: (+1) 212 634 4975 Tel: (+34) 91 114 83 00 Fax: (+39) 02 89 63 17 01 Tel: (+1) 212 634 4990 Fax: (+34) 91 114 83 01 Fax: (+1) 212 634 5171

SINGAPORE STOCKHOLM BranchSam of Exane Limited MCHUGHBranch of Exane Limited 20 Collyer Quay Nybrokajen 5 #07-02 Tung Centre 111 48 Stockholm Singapore 049319 Sweden Tel: (+65) 6212 9059 Tel: (+46) 8 5629 3500 Fax: (+65) 6212 9082 Fax: (+46) 8 611 1802

All Exane research documents are available to all clients simultaneously on the Exane website (http://cube.exane.com). Most published research is also available via third-party aggregators such as Bloomberg, Thomson Reuters, Factset and Capital IQ. Exane is not responsible for the redistribution of research by third-party aggregators.

Important notice: Please refer to our complete disclosure notice and conflict of interest policy available on http://cube.exane.com/compliance

This research is produced by one or more of EXANE SA, EXANE Limited and Exane Inc (collectively referred to as “EXANE") . EXANE SA is authorized by the Autorité de Contrôle Prudentiel et de Résolution and regulated by the Autorité des Marchés Financiers ("AMF"). EXANE Limited is authorized and regulated by the Financial Conduct Authority (“FCA”). Exane Inc is registered and regulated by the Financial Industry Regulatory Authority ("FINRA"). In accordance with the requirements of Financial Conduct Authority COBS 12.2.3R and associated guidance, of article 313-20 of the AMF Règlement Général, and of FINRA Rule 2241, Exane’s policy for managing conflicts of interest in relation to investment research is published on Exane’s web site (cube.exane.com). Exane also follows the guidelines described in the code of conduct of the Association Francaise des Entreprises d'Investissement ("AFEI") on managing conflicts of interest in the field of investment research. This code of conduct is available on Exane’s web site (cube.exane.com).

This research is solely for the private information of the recipients. All information contained in this research report has been compiled from sources believed to be reliable. However, no representation or warranty, express or implied, is made with respect to the completeness or accuracy of its contents, and it is not to be relied upon as such. Opinions contained in this research report represent Exane's current opinions on the date of the report only. Exane is not soliciting an action based upon it, and under no circumstances is it to be used or considered as an offer to sell, or a solicitation of any offer to buy.

While Exane endeavours to update its research reports from time to time, there may be legal and/or other reasons why Exane cannot do so and, accordingly, Exane disclaims any obligation to do so.

This report is provided solely for the information of professional investors who are expected to make their own investment decisions without undue reliance on this report and Exane accepts no liability whatsoever for any direct or consequential loss arising from any use of this report or its contents.

This report may not be reproduced, distributed or published by any recipient for any purpose. Any United States person wishing to obtain further information or to effect a transaction in any security discussed in this report should do so only through Exane Inc., which has distributed this report in the United States and, subject to the above, accepts responsibility for its contents.

BNP PARIBAS has acquired an interest in VERNER INVESTISSEMENTS the parent company of EXANE. VERNER INVESTISSEMENTS is controlled by the management of EXANE. BNP PARIBAS’s voting rights as a shareholder of VERNER INVESTISSEMENTS will be limited to 40% of overall voting rights of VERNER INVESTISSEMENTS.

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 48

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

TELECOM ITALIA (Underperform) Price at 31 Aug. 18 / Target Price Incumbent Telcos | Telecom Operators - Italy EUR0.55 / EUR0.38 -31%

Company description Peer group YTD performance Telecom Italia is the Italian telecom incumbent. The company operates in Italy in the Price YTD performance in EUR (%) Stock w ireline (voice, broadband and TV) and w ireless (essentially prepaid) markets as (31 Aug. 18) Abs. Rel. Sector w ell as in the media sector. The Group also ow ned a 67% stake in TIM Brazil w hich Altice NV (=) EUR 2.59 40.6 59 provides telecom services in Brazil (mainly mobile but w ith fixed services in the Sao Elisa Comm. (+) EUR 36.9 16.2 32 Paulo region as w ell) and has recently (2015) listed its mobile tow er business Inw it in w hich it has kept a 60% stake. Telia Company (=) SEK 40.6 5.1 19 DeutscheTelekom (=) EUR 13.9 (1.4) 12

Orange (=) EUR 14.0 (1.8) 11

Altice USA (+) USD 17.9 (3.3) 10

Telenor (=) NOK 158.1 (7.3) 5

Bouygues (=) EUR 38.0 (9.7) 2

Management Telefónica D. (=) EUR 3.58 (10.0) 2

Fulvio Conti, Chairman Telefónica (=) EUR 7.0 (12.1) (0)

Amos Genish, CEO BT Group (-) p 217 (17.3) (6)

Piergiorgio Peluso, CFO Telenet (=) EUR 47.1 (18.9) (8)

KPN (+) EUR 2.20 (20.9) (10)

Liberty Global (-) USD 26.8 (22.4) (12)

Telecom Italia (-) EUR 0.55 (23.8) (14)

Ownership structure Proximus (-) EUR 19.8 (25.4) (15)

Vivendi 23.9% Vodafone Group (+) p 165 (27.5) (18)

Elliott 8.8% Iliad (=) EUR 111.3 (44.1) (37) Cassa Depositi e Prestiti 4.9% Other Shareholders 62.4%

Sector calendar 04 Sep. 18 Iliad: H1 2018 Results (07:30 CET) 2017 Revenue by country 02 Oct. 18 Bouygues: Capital Markets Day 16 Oct. 18 Com He m Holding: Q3 2018 Results (07:30 CET) 18 Oct. 18 Tele2 B: Q3 2018 Results (07:00 CET) Elisa Comm.: Q3 2018 Results (07:30 CET) 19 Oct. 18 Telia Company: Q3 2018 Results (07:00 CET) 23% DNA Oyj: Q3 2018 Results (07:30 CET) Exclusive use of 22 Oct. 18 Telekom Austria: Q3 2018 Results (07:00 CET) 23 Oct. 18 Millicom: Q3 2018 Results 77% Italy 23% Brasil 24 Oct. 18 Orange Belgium: Q3 2018 Results (07:00 CET) Telenor: Q3 2018 Results (07:00 CET) KPN: Q3 2018 Results (07:30 CET) 25 Oct. 18 Orange : Q3 2018 Results (07:30 CET) Sam MCHUGH77% 26 Oct. 18 Proximus: Q3 2018 Results (07:00 CET) 31 Oct. 18 TDC: Q3 2018 Results Telefónica: Q3 2018 Results Telenet: Q3 2018 Results 07 Nov. 18 Telefónica D.: Q3 2018 Results 08 Nov. 18 DeutscheTelekom: Q3 2018 Results 2017 Revenue by activity Telecom Italia: Q3 2018 Results 13 Nov. 18 Vodafone Group: H1 2018 Results

53% Wireline 47% 53% 47% Wireless

Analyst Sam McHugh (+44) 207 039 9544 [email protected]

Exane BNP Paribas Research Telecom Italia 3 SEPTEMBER 2018 page 49

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)

LONDON FRANKFURT MADRID NEW YORK STOCKHOLM (+44) 207 039 9400 (+49) 69 42 72 97 300 (+34) 91 114 83 00 (+1) 212 634 4990 (+46) 8 5629 3500 GENEVA MILAN SAN FRANCISCO SINGAPORE PARIS (+33) 1 44 95 40 00 (+41) 22 718 65 65 (+39) 02 89 63 17 13 (+1) 212 634 4975 (+65) 6212 9059

Price at 31 Aug. 18 / 12m Target Price EUR0.55 / EUR0.38 -31% TELECOM ITALIA (Underperform) Reuters / Bloomberg: TLIT.MI / TIT IM Analyst: Sam McHugh (+44) 207 039 9544 Incumbent Telcos | Telecom Operators - Italy Com pany Highlights EURm 2.4 Enterprise value 49,194 Market capitalisation 14,542 Free float 9,075 1.6 3m average volume 60 1.2 Performance (*) 1m 3m 12m Absolute (17%) (20%) (32%) Rel. Sector (11%) (17%) (22%) 0.8 Rel. MSCI Europe (15%) (20%) (35%) 12m Hi/Lo (EUR) : 0.88 -38% / 0.55 +0% CAGR 2006/2018 2018/2020 EPS restated (**) (8%) 3% 0.4 Target Price CFPS (7%) (1%) Price 3.8*CFPS Relative to MSCI Europe Price (yearly avg from Dec. 07 to Dec. 17) 2.2 1.3 1.0 1.0 0.9 0.8 0.6 0.9 1.1 0.8 0.8 0.5 0.5 0.5 PER SHARE DATA (EUR) De c. 07 De c. 08 De c. 09 De c. 10 De c. 11 De c. 12 Dec. 13 Dec. 14 Dec. 15 Dec. 16 Dec. 17 Dec. 18e Dec. 19e Dec. 20e No of shares year end, basic, (m) 19 407.000 19 407.000 19 407.000 19 433.963 19 280.700 19 280.700 19 281.000 19 334.900 19 364.000 21 067.200 21 230.915 21 230.915 21 230.915 21 230.915 Avg no of shares, diluted, excl. treasury stocks (m) 19 407.000 19 407.000 19 407.000 19 420.482 19 280.700 19 280.700 20 139.957 21 026.163 21 608.906 22 157.192 21 690.300 21 230.915 21 230.915 21 230.915 EPS reported, Gaap 0.12 0.11 0.08 0.15 (0.25) (0.08) (0.03) 0.07 (0.00) 0.09 0.05 0.06 0.07 0.06 EPS company definition EPS restated, fully diluted 0.14 0.12 0.09 0.13 0.03 0.08 0.04 0.04 (0.01) 0.09 0.08 0.06 0.07 0.07 % change (14.3%) (15.7%) (23.7%) 41.8% (75.4%) 150.8% (56.3%) 3.0% NS NS (3.2%) (24.7%) 9.9% (3.6%) Book value (BVPS) (a) 1.1 1.1 1.1 1.2 0.9 0.8 0.7 0.8 0.7 0.8 0.8 0.9 0.9 1.0 Net dividend 0.080.050.050.060.040.020.000.000.000.000.000.000.000.00 STOCKM ARKET RATIOS De c. 07 De c. 08 De c. 09 De c. 10 De c. 11 De c. 12 Dec. 13 Dec. 14 Dec. 15 Dec. 16 Dec. 17 Dec. 18e Dec. 19e Dec. 20e P / E (P/ EPS restated) 15.0x 10.9x 11.3x 7.8x 28.8x 9.5x 17.4x 23.8x NC 9.5x 9.5x 8.6x 7.8x 8.1x P / E relative to MSCI Europe 136% 99% 73% 63% 244% 70% 104% 136% NC 53% 61% 60% 60% 67% P / CF 4.7x 3.0x 3.0x 2.9x 3.7x 2.4x 3.1x 5.4x 4.6x 3.5x 3.2x 2.3x 2.3x 2.3x FCF yield 6.8% 9.4% 7.0% 9.7% 7.4% 13.8% 10.9% 2.8% (4.8%) 1.7% 2.7% 3.7% 5.0% 5.4% P / BVPS 1.93x 1.22x 0.98x 0.86x 0.98x 0.95x 0.87x 1.16x 1.52x 1.01x 0.97x 0.63x 0.60x 0.57x Net yield 3.7%3.8%4.8%5.7%4.6%2.6%0.0%0.0%0.0%0.0%0.0%0.0%0.0%0.0% Payout 55.5% 41.2% 53.9% 44.1% 132.8% 24.3% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% EV / Sales 2.84x 2.28x 2.31x 2.23x 2.00x 1.81x 2.05x 2.56x 3.22x 2.72x 2.78x 2.57x 2.58x 2.58x EV / Restated EBITDA 7.5x 5.9x 5.5x 5.2x 4.9x 4.6x 5.0x 6.3x 8.2x 6.3x 6.4x 6.1x 6.0x 6.0x EV / Restated EBITA 14.7x 11.8x 10.9x 9.8x 8.8x 8.4x 9.4x 12.4x 17.4x 13.2x 13.1x 13.0x 12.8x 12.6x EV / NOPAT 23.4x 18.8x 17.3x 15.6x 12.2x 11.7x 13.0x 17.2x 24.2x 18.4x 18.2x 18.1x 17.7x 17.5x EV / OpFCF 14.1x 11.9x 10.3x 10.5x 10.4x 8.9x 10.0x 17.4x 79.5x 19.7x 18.0x 17.2x 16.3x 15.5x EV / Capital employed (incl. gross goodw ill) 1.4x 1.1x 1.0x 1.0x 1.0x1.0x1.0x1.1x1.3x1.0x1.1x0.9x0.9x0.9x ENTERPRISE VALUE (EURm) 89,021 68,690 62,799 61,557 59,914 53,472 47,946 55,330 63,493 51,674 55,179 49,194 48,457 47,646 Market cap 52,074 31,494 24,724 24,655 22,624 18,572 14,830 20,763 26,622 20,723 21,721 14,542 14,542 14,542 + Adjusted net debt 35,701 37,378 38,893 35,711 35,237 32,708 30,829 32,077 33,304 30,561 29,289 29,573 28,936 28,225 + Other liabilities and commitments 2,500 2,500 2,500 2,399 2,643 2,429 2,523 2,515 2,809 2,722 5,142 5,042 4,942 4,842 + Revalued minority interests 1,539 657 828 2,409 3,875 3,483 2,716 4,067 5,628 2,292 2,243 3,252 3,252 3,252 - Revalued investments 2,793 3,339 4,146 3,617 4,465 3,720 2,952 4,092 4,870 4,624 3,215 3,215 3,215 3,215 P & L HIGHLIGHTS (EURm) Dec. 07 Dec. 08 Dec. 09 Dec. 10 Dec. 11 Dec. 12 Dec. 13 Dec. 14 Dec. 15 Dec. 16 Dec. 17 Dec. 18e Dec. 19e Dec. 20e Sales 31,310 30,158 27,163 27,571 29,957 29,503 23,407 21,573 19,718 19,025 19,828 19,143 18,787 18,494 Restated EBITDA (b) 11,822 11,682 11,383 11,801 12,302 11,703 9,668 8,757 7,782 8,199 8,677 8,123 8,055 8,002 Depreciation (5,784) (5,874) (5,622) (5,547) (5,493) (5,340) (4,548) (4,284) (4,135) (4,291) (4,473) (4,353) (4,260) (4,216) Restated EBITA (b) (**) Exclusive6,038 5,808 5,761 use 6,283 6,809 6,363of 5,120 4,473 3,647 3,908 4,204 3,770 3,795 3,785 Reported operating profit (loss) 5,864 5,493 5,493 5,813 (602) 1,926 2,718 4,530 2,961 3,722 3,291 3,769 3,795 3,785 Net financial income (charges) (1,749) (2,484) (2,221) (1,774) (1,982) (1,964) (2,186) (2,178) (2,515) (907) (1,495) (1,525) (1,334) (1,373) Af filiates 0 68 67 99 (39) (6) 0 (5) 1 (16) (19) (19) (19) (19) Other 36 (29) (622) (7) (13) 2 341 541 611 47 0 0 0 0 Tax (1,682) (653) (1,121) (548) (1,643) (1,235) (1,111) (928) (401) (880) (490) (690) (757) (742) Minorities (7) (1) (15) (451) (446) (350) (436) (610) (729) (158) (166) (241) (256) (276) Net attributable profit reported 2,462 2,394 1,581 3,132 (4,725) (1,627) (674) 1,350 (72) 1,808 1,121 1,294 1,429 1,375 Net attributable profit restatedSam (c) 2,796 MCHUGH 2,358 1,800 2,555 624 1,566 714 768 (189) 1,895 1,778 1,295 1,429 1,375 CASH FLOW HIGHLIGHTS (EURm) Dec. 07 Dec. 08 Dec. 09 Dec. 10 Dec. 11 Dec. 12 Dec. 13 Dec. 14 Dec. 15 Dec. 16 Dec. 17 Dec. 18e Dec. 19e Dec. 20e EBITDA (reported) 11,648 11,367 11,115 11,423 12,249 11,698 9,453 8,815 7,340 8,016 7,801 8,122 8,055 8,002 EBITDA adjustment (b) 174 315 268 378 53 5 215 (58) 442 183 876 1 0 0 Other items (174) (315) (268) (378) (131) (1,178) (235) (135) (277) (369) (354) (1) 0 0 Change in WCR 199 (209) (472) (974) (866) 213 (239) (789) (1,628) (548) (581) (1,000) (800) (800) Operating cash flow 11,847 11,158 10,643 10,449 11,305 10,738 9,194 7,833 5,877 7,282 7,742 7,122 7,255 7,202 Capex (5,520) (5,365) (4,543) (4,583) (5,538) (4,702) (4,391) (4,659) (5,078) (4,658) (4,684) (4,253) (4,282) (4,128) Operating free cash flow (OpFCF) 6,327 5,793 6,100 5,866 5,767 6,036 4,803 3,174 799 2,624 3,058 2,868 2,973 3,074 Net financial items + tax paid (2,707) (2,779) (4,320) (3,235) (3,798) (2,991) (2,883) (2,489) (2,337) (2,234) (2,419) (2,214) (2,091) (2,115) Free cash flow 3,620 3,014 1,780 2,631 1,969 3,045 1,920 685 (1,538) 390 639 655 882 959 Net financial investments & acquisitions 7 782 0 928 411 59 84 (277) 929 692 (609) (700) 0 0 Other 858 (443) (1,428) 216 (4,197) 389 412 (1,404) (600) 584 1,461 0 0 0 Capital increase (decrease) 0 0 0 0 0 0 0 0 186 1,304 16 0 0 0 Dividends paid (2,886) (1,691) (1,060) (1,122) (1,326) (964) (537) (252) (204) (227) (235) (239) (245) (248) Increase (decrease) in net financial debt (1,600) (1,662) 708 (2,653) 3,143 (2,529) (1,879) 1,248 1,227 (2,743) (1,272) 284 (637) (711) Cash flow, group share 8,919 8,584 6,739 6,960 4,884 6,157 4,068 3,418 5,168 5,165 5,400 4,981 5,059 4,902 BALANCE SHEET HIGHLIGHTS (EURm ) De c. 07 De c. 08 De c. 09 De c. 10 Dec. 11 Dec. 12 Dec. 13 Dec. 14 Dec. 15 Dec. 16 Dec. 17 Dec. 18e Dec. 19e Dec. 20e Net operating assets 68,339 66,045 64,811 68,365 61,505 52,898 49,431 50,157 50,730 52,923 53,194 53,794 53,816 53,728 WCR (4,299) (5,094) (2,697) (4,597) 314 (1,743) (2,218) (499) (159) 506 (779) 221 1,021 1,821 Restated capital employed, incl. gross goodwill 64,040 60,951 62,114 63,768 61,819 51,155 47,213 49,658 50,571 53,429 52,415 54,015 54,837 55,549 Shareholders' funds, group share 26,985 26,126 25,952 28,819 22,791 19,378 17,061 18,145 17,610 21,207 21,557 22,657 23,888 25,071 Minorities 1,063 730 1,168 3,791 3,904 3,634 3,125 3,554 3,723 2,346 2,226 2,422 2,631 2,852 Provisions/ Other liabilities 2,640 2,290 735 1,989 2,737 1,957 1,822 2,214 2,294 2,478 2,826 2,846 2,865 2,884 Net financial debt (cash) 35,701 34,039 34,747 32,094 35,237 32,708 30,829 32,077 33,304 30,561 29,289 29,573 28,936 28,225 FINANCIAL RATIOS (%) De c. 07 De c. 08 De c. 09 De c. 10 De c. 11 De c. 12 Dec. 13 Dec. 14 Dec. 15 Dec. 16 Dec. 17 Dec. 18e Dec. 19e Dec. 20e Sales (% change) 0.1% (3.7%) (9.9%) 1.5% 8.7% (1.5%) (20.7%) (7.8%) (8.6%) (3.5%) 4.2% (3.5%) (1.9%) (1.6%) Organic sales grow th (1.3%) (2.3%) (5.6%) (3.8%) 2.7% 0.5% (5.2%) (5.4%) (4.6%) (3.5%) 4.2% (3.5%) (1.9%) (1.6%) Restated EBITA (% change) (**) (19.0%) (3.8%) (0.8%) 9.1% 8.4% (6.6%) (19.5%) (12.6%) (18.5%) 7.1% 7.6% (10.3%) 0.7% (0.3%) Restated attributable net profit (% change) (**) (14.3%) (15.7%) (23.7%) 41.9% (75.6%) 150.8% (54.4%) 7.5% NC NC (6.1%) (27.2%) 10.4% (3.8%) Personnel costs / Sales NCNCNCNCNCNCNCNCNCNCNCNCNCNC Restated EBITDA margin 37.8% 38.7% 41.9% 42.8% 41.1% 39.7% 41.3% 40.6% 39.5% 43.1% 43.8% 42.4% 42.9% 43.3% Restated EBITA margin 19.3% 19.3% 21.2% 22.8% 22.7% 21.6% 21.9% 20.7% 18.5% 20.5% 21.2% 19.7% 20.2% 20.5% Tax rate 40.9% 21.2% 33.6% 13.2% NC NC NC 39.5% NC 31.4% 27.6% 31.0% 31.0% 31.0% Net margin 9.0% 7.8% 6.7% 10.9% 3.6% 6.5% 4.9% 6.4% 2.7% 10.8% 9.8% 8.0% 9.0%8.9% Capex / Sales 17.6% 17.8% 16.7% 16.6% 18.5% 15.9% 18.8% 21.6% 25.8% 24.5% 23.6% 22.2% 22.8% 22.3% OpFCF / Sales 20.2% 19.2% 22.5% 21.3% 19.3% 20.5% 20.5% 14.7% 4.1% 13.8% 15.4% 15.0% 15.8% 16.6% WCR / Sales (13.7%) (16.9%) (9.9%) (16.7%) 1.0% (5.9%) (9.5%) (2.3%) (0.8%) 2.7% (3.9%) 1.2% 5.4% 9.8% Capital employed (excl. gdw ./intangibles) / Sales 40.4% 35.0% 44.9% 43.4% 54.3% 40.8% 47.0% 59.7% 74.6% 88.7% 79.5% 90.7% 96.8% 102.2% ROE 10.4% 9.0% 6.9% 8.9% 2.7% 8.1% 4.2% 4.2% (1.1%) 8.9% 8.2% 5.7% 6.0% 5.5% Gearing 127%139%143%110%132%142%153%148%156%130%123%118%109%101% EBITDA / Financial charges 6.8x 4.7x 5.1x 6.6x 6.2x 6.0x 4.4x 4.0x 3.1x 9.0x 5.8x 5.3x 6.0x 5.8x Adjusted financial debt / EBITDA 3.0x 3.2x 3.4x 3.0x 2.9x 2.8x 3.2x 3.7x4.3x3.7x3.4x3.6x3.6x3.5x ROCE, excl. gdw ./intangibles 30.1% 34.6% 29.7% 33.1% 30.1% 38.0% 33.5% 25.0% 17.9% 16.7% 19.2% 15.6% 15.0% 14.4% ROCE, incl. gross goodw ill 5.9% 6.0% 5.8% 6.2% 7.9% 9.0% 7.8% 6.5% 5.2% 5.3% 5.8% 5.0% 5.0% 4.9% WACC 8.1%8.2%9.5%9.8%10.5%9.5%8.1%8.9%9.6%8.9%9.0%8.0%8.0%8.0% Latest Model update: 01 Sep. 18 (a) Intangibles: EUR36,654.00m, or EUR1 per share. (b) adjusted for capital gains/losses, impairment charges, exceptional restructuring charges, capitalized R&D, pension charge replaced by service cost (c) adj.for capital gains losses, imp.charges, capitalized R&D, am. of intangibles from M&A, exceptional restructuring, (*) In listing currency, w ith div. reinvested, (**) also adjusted for am. of intangibles f rom M&A, or f or am. of gw ill for pre IFRS year

For the exclusive use of Sam MCHUGH at EXANE S.A. (04-Sep-2018)