Mota-Engil Construction

Transcription

1 BPI EQUITY RESEARCH Mota-Engil Construction The African Dawn (Price Target raised from 3.30 to 5.00; Neutral Recommendation maintained) 4 Listing the African unit: Mota Engil (ME) announced the intention to list ME Africa - MEA (64% group EBITDA13 F ) in an European market and increase its capital to boost the fire power to tackle the huge opportunities in the African continent. As part of the operation, ME shareholders should receive shares equivalent to 20% of MEA pre k increase. The company also announced the intention to sell its treasury shares (5.4%) to reinforce its capital structure and, in our view, to raise funds for the EGF privatization (Portuguese public waste company). 4 YE14 Price Target increased to 5.00/sh (+50%, + 1.7/sh) driven by MEA (+ 2.3/ sh) after upgrading estimates and changing our valuation methodology to DCF (6.7x EBITDA14 F ) to better reflect the future prospects of the unit. We were previously valuing it through an avg. 4.2x EV/EBITDA14 F multiple (building unit) that fails to reflect the current context of MEA vs. South-African listed peers. Maintaining the previous method would have added only 0.5/sh to the SoP, resulting in just 1% upgrade (vs.50%) as we are also downgrading Ascendi (- 0.7/sh). If the IPO goes ahead, added visibility could drive another re-rating. The premium that African sectors trade vs. European counterparts points to an EBITDA multiple range for the construction sector of between 6.8x-9.0x (avg. 7.5x vs. 6.7x in our DCF), valuing MEA at 1.3bn - 1.7bn (EV) and ME fair value (w/10% disc.) at /sh. 4 Keeping a Neutral recommendation as our valuation upgrade was preceded by a strong stock rally. Earnings delivery should be strong (30% NP CAGR in F ) but leverage is still high (3.2x ND/EBITDA13 F inc. factoring&leasing and 37% of Debt to be refinanced in the next 2 years). Execution, construction cycle and country related risks are significant in this story, but ME is also an experienced contractor and a good vehicle to play African growth. Neutral High-Risk 23rd December 2013 Portugal Mota-Engil vs. PSI20 vs. MSCI Small Cap Index Source: Bloomberg. Stock data Price (19 th Dec.): 4.37 Price Target: 5.00 No. of shares (mn): Market Cap ( mn): 893 Reuters/Bloomberg: MOTA.LS/EGL PL Free-Float: 27% NET DEBT/EBITDA14 F : 2.2 ROE13: 11% EPS Growth ('12-'15 F ): 31% Avg. Daily Turnover [ '000]: Major Shareholders: Mota Family (67.78%); Treasury shares (5.42%) Estimates F 2014 F 2015 F 2016 F EPS Adj. ( ) P/E adj Dividend Yield 10.6% 7.0% 2.5% 2.5% 2.5% 2.5% EV/EBITDA FCF yield (1) 7.1% 6.8% 4.7% 7.8% 10.0% 12.2% (1) maintenance capex only. Historical Recommendation Date Recommendation 18-Sep-12 Neutral Analysts Bruno Silva, CFA bruno.miguel.silva@bpi.pt Phone Filipe Leite, CFA filipe.martins.leite@bpi.pt Phone Available on our website: BPI Online, and Bloomberg at NH BPD

2 BPI vs. Consensus Stock Momentum Company: Mota-Engil SGPS SA Sector: DJES Cns&Mat Price Performance Forward P/E and EV/EBITDA Valuation monitor Relative Valuation EV/EBITDA BPI Consensus 5.8 n.s. n.s. Sector P/E BPI Consensus Sector PBV BPI Consensus n.s. n.s. n.s. 1 Y 3 M YTD 5.0 0% 100% 200% 300% DJES Cns&M at M ota Engil SGPS SA Market Price Rating ( ) Forward P/E EV/EBITDA 0 Feb-04 Jun-07 Sep-10 Dec-13 Market Recommendations Sector Dividend yield BPI 2.5% 2.5% 2.5% Consensus 2.5% 2.5% 3.7% Sector 3.1% 3.3% 3.5% Price Target Consensus Price 0.5 Nov-12 Apr-13 Aug-13 Dec-13 P&L and B\S monitor BPI estimates/consensus Revenues -2% -3% -1% EBITDA 5% 4% 6% EBIT 0% -3% -1% Net Profit -9% -13% -4% Profitability monitor EBITDA Margin BPI 15.6% 15.6% 15.4% Consensus 14.9% 16.3% 18.0% EBIT margin BPI 9.9% 9.9% 9.8% Consensus 10.0% 11.1% 12.2% Net Profit margin BPI 2.3% 2.6% 3.2% Consensus 2.6% 3.2% 4.1% Key leverage ratios Net Debt/EV BPI 44% 42% 39% Consensus 56% 54% 52% Net Debt/EBITDA BPI Consensus Fair Value Comparison ( ) CAGR Net Profit EBIT EBITDA 2.0 Revenues 0.0 P/E PBV Consensus BPI 0% 10% 20% 30% Current M arket Price BPI Consensus EBITDA Consensus ( mn) EPS Consensus ( ) FY FY15 FY FY FY FY Nov-12 Apr-13 Aug-13 Dec-13 Nov-12 Apr-13 Aug-13 Dec-13 Source: Factset, Bloomberg and BPI Equity Research. 2

3 Mota-Engil at a Glance 2013 F EBITDA mg by business EBITDA weight (Africa 61% of FY15 F EBITDA)!"!" # $$ $ #!"!" Source: ME and BPI Equity Research. Debt maturity profile at 9M13 ( mn) Net debt as of 9M13 ( mn) Year 2 Years 3 Years 4 Years 5 Years >5 Years Gross Debt < 1 Year Cash Gross Debt > 1 Year Net Debt (ex Leasing & Fact) Leasing & Fact Net Debt (incl Leasing & Fact) Source: ME. Source: ME. Construction Backlog evolution ( bn) 9M13 Construction backlog ( 3.7bn) % 75% 70% 65% 60% 55% 50% Source: ME. Orderbook mn % of international orderbook Source: ME. Europe FY13 F Sales African FY13 F Sales 3

4 INVESTMENT CASE STRONG EARNINGS DELIVERY AND ANNOUNCEMENT OF IPO OF MOTA ENGIL AFRICA Since our last update (10th Sep 13) two main evens must be highlighted: (1) the strong Q3 earnings delivery and more importantly (2) the announcement of the IPO of MEA. Q3 earnings brought reassuring message on sustainability of African margins above 20% and better than expected performance in Portugal Regarding results in Q3 ME presented a 35% increase in EBITDA driven mainly by the international units (Africa +46% and Latam +59%) and a good (and unexpected) strong delivery in European market where 3Q EBITDA increased by 16% yoy (46% or 10mn above our quarterly estimate) helped by a top performance in Construction sector margins in Portugal. The Net debt of the company was reduced by 5% or 53mn qoq with a good WK recovery (that lead us to revise upwards our estimate for YE WK evolution, /sh or +3% in our FV). The Leasing and Factoring was also reduced by 17mn qoq (-6%) putting the total Net Debt reduction in the quarter (incl. Leasing and Factoring) at 70mn (-5%) qoq. Another positive note from the conference call was above-expected guidance given by the management not only for the international markets but also for the domestic construction market where the company is pointing to a flat top line performance during In the beginning of October (accounted in 9M numbers), the company also announced that sold a building in Lisbon (the Báltico Center Building in the region of "Parque das Nações") for 43mn, constituting the largest real estate deal made in Portugal since ME - 3Q Results Sales ( mn) 3Q13 3Q12 yoy Europe % Construction % E&S % Africa % America % Holding n.s. Total % EBITDA ( mn) 3Q13 3Q12 yoy Europe % Construction % E&S % Africa % America % Holding % Total % Source: ME and BPI Equity Research. African order intake "dealing" with USD 5bn of potential awards Moreover, in the end of September the CEO of ME announced, in an interview, that ME would post a very solid set of results for the second half of 2013 (already confirmed in Q3) with Africa and Latin America accounting for more than 60% of revenues (BPIF: 62%). The upbeat message went on as the company is also bidding for contracts worth more than USD 5bn in Africa with new markets including Zambia and Ghana. Confidence boost now reflected in our estimates We upgraded our domestic market estimates but still remain on the conservative side with a 4% top line contraction in domestic construction division. For the African business, we have simultaneously cut our point estimate for FY13 but followed the upbeat message of the company in what regards the sustainability of the current margins, considering the visibility of the profitability of the current orderbook and sentiment in the region. In consolidated terms, we are upgrading EBITDA in by an average 5% and Net debt being cut by a similar range. We estimate topline and EBITDA to expand at 10% CAGR13-16 and bottom line at 30% in the same period. Net cashflow should also follow the trend with ME amassing 208mn debt reduction in despite the expected operating growth and consequent investment required in this stage. We are keeping a fixed shareholder remuneration in line with previous years at 0.11/sh with an implied 2.5% DY. 4

5 RAISING OUR TARGET BY 50% MOSTLY ON MEA Material change to valuation method of MEA: + 2.3/share vs. +E 0.5/sh keeping previous method We have produced a material change to our valuation methodology of the African business (MEA). To start with, the upgrade in estimates of the African business would not be fully reflected in our valuation if we had kept the previous method as it used a multiple on FY14 EBITDA. For ME Angola we previously used the deal multiple implicit in the sale of 49% of the Angolan unit in 2011 which is static and thus would not reflect the changes in the operating context namely new contracts and improved growth estimates since the deal was done. In addition, that multiple was probably discounting tax advantages given to ME in Angola that we can now perceive from the earnings reported by the area in the last couple of years and lastly, the deal perimeter of the unit sold was not capturing the full business generated in Angola such as the services unit and other revenues from property and services provided by the parent company. For MEA ex-angola, we were applying a construction peer multiple of South African listed construction companies adjusted for the ST growth differential between consensus growth estimates for the peer group vs. our estimates for MEA, consequently partly foregoing the enhanced growth in the medium and long term as well as the impact of the superior return on capital employed. The upgrade of the estimates for MEA and the update of the multiple for MEA ex- Angola (would now be 5.0x vs. 4.4x before) would have resulted in a 14.1% increase of MEA unit EV equivalent to 14.3% at the equity level all else constant, 0.5/ share. Enhanced prospects and visibility, limited comparability of peer group and IPO plan consequences among others justifying changing the method The enhanced prospects of MEA, the recognition of the limitations of using a static deal multiple to value Angola and more importantly the increasing deficit of comparability of South African builders vs. MEA performance and prospects as well as geographic and business focus urged a change to our valuation method. In addition, the announcement of the possible IPO of MEA has the double effect of increasing visibility of the unit (higher expectations on the valuation benchmark and more information to be released with the operation) and a consequent reduction of the discount likely assigned now by the market to the subsidiary. DCF with implicit 6.7x multiple vs. "new" peer group average of 7.5x (range x) As such, we have implemented a DCF to value MEA and have also redefined the peer group by collecting data related to the Africa vs. Europe premium/discount observed in relevant sectors in order to apply it to an European construction peer group multiple. With that, we are more able to get an idea of the premium that African companies with a more comparable geographic reach with MEA's should have vs. the European counterparts. By using EV/EBITDA14F and EV/EBIT14 F we observed an average premium of Africa vs. Europe (average of retail, beverages, cement and telcos) of 17% and 11%. With all the shortcoming that these multiples still have, we reach very different conclusions about MEA potential valuation. Focusing on EV/EBITDA, the average of the peer group would point to 7.5x (vs the 3.9x and 5.0x multiple we would be using under the previous method). Our DCF of MEA points to a valuation of the unit of 1.25bn vs 0.77bn before (+63% at EV level) with an implicit EV/EBITDA14 F multiple of 6.7x, below the 7.5x average of 5

6 the "redefined" peer group ( x range). This implies, all else constant, adding 2.3/share to our valuation or + 1.8/sh vs. what we would get under the previous method with our new estimates. Cutting concessions valuation: - 0.7/share On the negative side, we have cut the valuation of the motorway concession business by the equivalent to 0.73/share. We were previously assuming the reported BV of Ascendi (the holding 60% controlled by ME that groups all concessions) and adding the NPV of the individual concessions as well as ME's shareholders loans. Now we are assuming the investment carried out by shareholders subtracted by the negative NPV of the availability concessions in Portugal estimated at 47mn for ME's stake while adjusting Lusoponte and foreign concession's reported equity investment by the impact of the cost of capital evolution since inception. We have changed the method due to the volatility shown in the BV of Ascendi caused by the swaps mark-to-market and consequent limited use of that metric as a good departing point for our valuation. Target up 50% to 5.00/share The above changes together with other relatively minor adjustments, our YE14 Price Target for ME has climbed by 50% to 5.00 per share including a 10% small cap discount. Still a risky sector and a risky investment case The negative points of the investment case remain. Subdued performance of Central European units, a high level of debt particularly considering debt from subsidiaries consolidated through equity (Ascendi in particular) despite their non-recourse nature, the cyclical nature of the construction sector and execution risks and the perceived higher risk of the key markets where ME is more active in terms of FX (even if a lot of contracts are done in strong currencies) and political risk in areas with traditional instability (Latam and Africa). High level of sensitivity to valuation inputs High risk rating As such, although we believe the momentum is very positive for the company and the stock, we advise investors to carefully consider the above risks and particularly to be comfortable with the valuation sensitivity to key inputs of the valuation, namely forecasts (volumes and margins), terminal values used in the valuation as well as the paramount impact of discount rates. The country risk premium has been an area of debate. We are using for Angola 6.7% and for the remaining African business 6.5%, which in fact is more conservative than the 4.0%-5.0% suggested by USD denominated sovereign yields in comparable African countries (Angola does not have publicly trading USD or Eur sovereign bonds). Despite that evidence, we have applied a 150bps additional premium to reflect the risk of political instability in Angola and other African countries and also recognized the yield differential to German bunds (our valuation is euro denominated). Finally, for Latam we assume an average spread of 1.6% (in euros) and 2.35% for Portugal. If we add 100bps to the country risk premium of Portugal, Latam and Africa to value the construction units, our valuation would drop to 4.2/share vs 5.0/sh, which is definitely considerable. In this case, we would be using a WACC in euros for Africa of 14.4% (vs 13.6% in base case) and 10.6% for the area Portugal+Latam (vs 9.8%). The decision to go ahead or not with the IPO (including a partial spin-off of MEA and a capital increase) is also an important risk. The decision and further details 6

7 will be known on the 27 th December when ME AGM should meet to vote it. The prospectus of the IPO, if it goes through, should bring more visibility into MEA potential growth and profitability and consequently be as ource of further valuation upgrades. On the flip side, the "decision" about MEA's capital structure will influence valuation, capital increase proceeds and ME parent leverage and liquidity levels. The plan also includes an authorization for management to sell treasury stock of 5.4%, which could be a source of overhang risk. Keeping a Neutral recommendation We are keeping a Neutral recommendation on Mota Engil as our valuation upgrade was preceded by a strong performance of the stock reflecting lower country risk in Portugal, upgrade of market growth prospects reflected in consensus estimates particularly since early 2013 and a general appetite for equity rsik. The IPO plan of the African unit played an important role over the last weeks justifying a second stage in the rally. Earnings delivery should be strong (30% NP CAGR in ) but leverage is still significant with a consolidated ND/EBITDA14 and 15 at 2.5x and 2.2x in our base case without considering potential proceeds from the MEA operation. If we also include factoring a leases, the ratio would increase to 3.2x and 3.0x respectively. If we also add the proportional debt from Ascendi and Indaqua, the ratio would climb to 4.2x in FY13. ME Share price evolution ( ) Q13 Results Source: Bloomberg and BPI Equity Research. 3Q13 Results and announcement of MEA IPO 2Q13 Results Kendal II reduce its stake from 4.6% to 1% Jan-13 Apr-13 Aug-13 Dec-13 Stock Price BPI PT LISTING THE AFRICAN UNIT Simultaneous spin-off, IPO and capital increase of MEA in 2Q14 - roadshow in Feb. Mota Engil (ME) announced with the presentation of 3Q13 results (21 Nov) the intention to propose to the 27 th December shareholders meeting: (1) a partial spin-off of Mota Engil Africa (MEA), whereby the current shareholders would be assigned shares equivalent to 20% of MEA as it is. (2) the floating of MEA in an European market, probably the UK and ME would keep a controlling stake above 50%; and (3) a capital increase at MEA level reserved to ME shareholders. MEA - Events calendar AGM 27 Dec13 Roadshow Feb13 IPO 2Q14 The latter has been raising questions from investors. It is our understanding that the possible decision to reserve the capital increase to shareholders is only relevant to the extent of the undervaluation (or overreaction) of MEA embedded in ME stock price. What we mean by that is that if the current ME price is under or overvaluing implicitly MEA, then an investor would be, all else constant, rationally deciding to go long or short ME shares to capture the upside or downside to be unveiled once the market explicitly values MEA in the floating process. The operation will likely use a bookbuilding process where potential investors will place their bids. Current shareholders of ME have the option to exercise the call, ie, subscribe the capital increase at the set price or give up the rights and let those who bid to subscribe the new shares on their place. In other words, the bookbuilding should theoretically be a competitive process to price MEA at fair market value leaving it in a range of indifference for current ME shareholders and other investors regarding MEA. So, the relevance today is whether ME stock price already reflects that fair value of MEA. MEA business scope Assuming the current free-float of ME holding (c.27%) as a minimum to provide the newco with a sufficient liquidity to sustain market interest, ME could end up with a stake between 51% and 70%-73%. The newco will be listed in an European market, most likely London. The IPO may take place in 2Q14 preceded by a marketing roadshow at the beginning of 2014, perhaps in February. With this move ME intends to both 7

8 provide MEA with fire power to face the opportunities ahead and capture the interest shown by the investment community in the company's African unit. The board is also proposing to sell the group's 5.4% treasury stock, in order to further reinforce the balance sheet with c 48mn proceeds at the current stock price level. MEA backlog evolution ( bn) MEA the growth engine of ME - EBITDA CAGR12-15 F of 20% MEA is a focused Sub- Saharan Africa region player, leveraging on: (1) The huge potential of the region, recipient of the largest slice of the Programme for Infrastructure Development (PIDA) in Africa, expected to invest USD 68bn by 2020; (2) ME's unique footprint in Africa, with a long-standing presence in Portuguese-speaking countries - Angola, Mozambique and Cape Verde. ME holds 100% of MEA, which in turn holds 51% of ME's Construction operations in Angola (57% of FY13 F revenues of MEA), 100% of the remaining construction operations in the African continent (38%) and 100% of the Environment & Waste unit, mostly waste treatment in Angola (5%). Source: Mota-Engil. MEA recorded FY12 revenues of 729mn, up 19% yoy (+36% in 9M13) and represents 42% of total 9M group sales with a backlog of 1.5bn (more than 25 months of sales). EBITDA of 163mn in 9M already represents 61% of total group EBITDA after jumping by an impressive 63% yoy ( 137mn in FY12). Mota Engil Africa - Estimates breakdown CAGR Sales ( mn) FY11 FY12 yoy FY13 F yoy FY14 F yoy FY15 F yoy Construction % % % % 19% Angola % % 615 8% 649 6% 10% Africa ex Angola % % % % 37% E&S % 54 80% 66 22% 76 15% 36% Total % % % % 20% Note: BPI estimates for the historical breakdown of Africa construction revenues. EBITDA CAGR mn FY11 FY12 yoy FY13 F yoy FY14 F yoy FY15 F yoy Construction % % 220 5% 234 7% 24% Angola % % 129 3% 130 1% 14% Africa ex Angola % % 91 8% % 42% E&S % 23 84% 28 22% 33 15% 37% Total % % 248 7% 267 7% 25% Note: BPI estimates for the historical breakdown of Africa construction EBITDA. EBITDA mg % FY11 FY12 yoy FY13 F yoy FY14 F yoy FY15 F yoy Construction 19% 18% -1.0pp 22% 4.3pp 21% -1.0pp 20% -1.0pp Angola 19% 18% -1.0pp 22% 4.3pp 21% -1.0pp 20% -1.0pp Africa ex Angola 19% 18% -1.0pp 22% 4.3pp 21% -1.0pp 20% -1.0pp E&S 59% 42% -17.4pp 43% 1.0pp 43% 0.0pp 43% 0.0pp Total 21% 19% -2.2pp 23% 4.4pp 22% -0.8pp 21% -0.9pp 8

9 Everyone is intimate to the fact that Africa has been one of the fastest growing regions in the world. According to the PIDA, the African population is estimated to reach approximately 1770mn in 2040, which represents an increase of 71% from Departing from 2010, GDP/per capita is expected to increase by 48% until 2020, 139% until 2030 and 260% until 2040 which reflect a total value of around USD 4709 in 2020, USD 7636 in 2030 and USD in Forecast for Africa Population (mn) Urban Population (mn) GDP (2005 PPP $ bn) GDP/per capita ME: Unique path in Africa Source: Mota-Engil. The priority action plan (included in the PIDA) comprises 51 projects to be completed by 2020, divided into four different sectors, namely transport, energy, TWR (Trans-boundary Water Resources) and ICT (Information and communications technology) as well as six different regions including Continental, North Africa, West Africa, Central Africa, Southern Africa and East Africa. These 51 projects require an initial investment of around USD 67.9bn (to be made until 2020) and ME, as one of the main construction players on the continent, is preparing itself (with the IPO) to grab its share of the pie, leveraging on its unique footprint and proven track record. MEA advantages are known but also underestimated. Africa is a complex competitive environment gathering a vast and diverse collection of risks. ME has been in Angola since 1946 and has a JV in Angola where also participates Sonangol, the national oil champion. The track record spreads over the sub-saharan region with several high-profile projects delivered to African public clients and multinationals in different sectors such as mining (Vale, Paladin, Xstrada/Glencore, Rio Tinto) and other infrastructures related sectors. If local connections and experience are valuable, then ME and its subsidiary MEA are probably a good way to gain exposure to the infrastructure growth of the region. The somewhat constrained financial chest limits in the same proportion the capacity to go after every opportunity, but on the other hand, keeps the company focused on quality projects with higher margins. Source: Mota Engil. Africa - Priority action plan - By Sector Investment Projects (USD bn) Transport Energy TWR ICT Total Source:Mota-Engil. Africa - Priority action plan - By Region Investment Projects (USD bn) Continental 7 3 North Africa West Africa Central Africa Aouthern Africa East Africa Total Source:Mota-Engil. 9

10 VALUATION OF MEA Valuing growth in Africa is challenging given the typical volatility of the investment environment and narrow visibility on MEA pipeline. The lack of a good peer group also limited our visibility into the potential market value perception of the African construction business. Almost all African listed construction players (most of them from South Africa) have their activity spread out over different regions including Asia and Australia. On average, these companies' non-african activity represented 43% of FY12 revenues and 41% of FY12 revenues did not derive from construction activities, such as mining or concessions that have very different capital structure requirements. For these reasons, we have changed our approach and focused our valuation analysis of MEA on two scenarios: (1) Average premium/discount that African players in other sectors are trading vs. the corresponding sector players in Europe in order to assess the "African" trading premium and (2) In-house DCF valuation for MEA. The latter is the methodology which we are now using in our ME SoP. We had until now valued the African businesses using the implicit multiple paid by ME's partner in Angola for 49% of ME Angola and the average trading multiple (EV/EBITDA) of the listed construction players in Africa to value the remaining MEA. We have changed our methodology as the DCF can better capture the specifics of the business and CF generation in the explicit period. We are only now adopting the DCF due to (1) the higher growth reported and anticipated by ME in Africa and also its wider geographic focus in the region and different business exposure, setting MEA apart from the until now used peer group of mostly South African listed players and (2) the recognition of the shortcomings from using the multiple implicit in the ME Angola transaction due to the narrower perimeter of the acquisition vs. the overall operation in Angola, tax advantages perceived in the reported net profit (Angola minorities in the P&L) in the last two years and substantially higher than expected growth (Construction of the Calueque dam - USD 164mn, Sonangols Expansion Project of a gas stations - USD 107mn, Project IMOLAP - USD 100mn, Financial City Project - USD 73mn, Solar Village Project, to Sonangol Holdings - USD 26mn). In the absence of an adequate peer group in Africa we chose to look for the African premium across sectors instead Until now we have been valuing the construction activity of Africa at c.4x EV/ EBITDA14 F and the E&S activity in Africa as part of the Portuguese operations (5.7x EV/EBITDA14 F ). In the specific case of MEA construction activity, we were applying a 3.9x multiple for the Angolan activity (in line with the implicit multiple from the sale of 49% of ME Angola to Sonangol and other Angolan shareholders) while for the remaining African construction activity, we applied a multiple in line with listed African peers (4.4x at the time of our last valuation). looking at other sectors As stated above, the lack of an adequate peer group of the listed construction sector in Africa limited the validity of using multiples to value MEA. Looking at other sectors (with a more consistent peer group), we collected more substantial evidence of the "African" trading premium vs. European counterparts, most likely justified by the expected superior growth of African markets despite their considerably higher risk on several fronts, particularly from the perspective of the European investor. 10

11 Average "African" trading premium on EBITDA and EBIT points to a ME fair value of /share The table below shows that on average African players trade at a 17% premium using EV/EBITDA14 F (11% at EV/EBIT14 F ). By applying this average premium (and min. and max. of the range) to the European pure construction players multiple, we reach an EV/EBITDA14 F multiple average of 7.5x (with a range between 9.0x and 6.8x). If we apply this multiple to value MEA, our fair value for ME (including 10% discount) would be 5.6/sh (for the 7.5x avg multiple) and between 5.0/ sh and 6.8/sh (for the 6.8x - 9.0x multiple range) and MEA alone would be worth 1.39bn (EV) in the case of 7.5x avg. multiple and between 1.26bn and 1.66bn using the 6.8x - 9.0x multiple range. Using the average EV/EBITDA multiple of the South-African listed construction stocks (4.5x), our fair value of ME (including 10% discount) would be 3.2/share and MEA would be worth 832mn. If we use the EV/EBIT multiple instead (11.4x avg, 11% premium vs Europe) to value MEA, we would reach a 1.63bn EV (implicit 8.8x EV/EBITDA14 F ) and 6.8/sh fair value of ME (including 10% discount). African Multiples comparison CARG12-15 EV/EBITDA14 EV/EBIT14 PER14 EV/SALES14 Sales EBITDA Telcoms AVERAGE Africa % 4% AVERAGE Europe % -3% Prem. (Disc) 5% 3% -3% 31% Breweries AVERAGE Africa % 11% AVERAGE Europe % 8% Prem. (Disc) 8% 10% 42% 10% Retail AVERAGE Africa % 15% AVERAGE Europe % 5% Prem. (Disc) 40% 29% 41% 63% Cement AVERAGE Africa % 6% AVERAGE Europe % 9% Prem. (Disc) 14% 4% 13% 171% AVERAGE Europe Const (1) Avg Prem. (Disc) all sectors (2) 17% 11% 23% 69% Min prem(disc) (3) 5% 3% -3% 10% Max prem(disc) (4) 40% 29% 42% 171% African const Multiples assuming Avg Prem. (Disc) all sectors (1) *(1+(2)) Min prem(disc) (1) *(1+(3)) Max prem(disc) (1) *(1+(4)) Source: BPI Equity Research and Bloomberg. 11

12 Valuing MEA through DCF As justified above, we are now adopting the DCF to value MEA. Assuming the IPO goes ahead, we believe it will be a terrific opportunity to increase the valuation visibility of the operation and also provide the market in that period with a lot more information that should have the effect of diminishing the perceived risk. In addition, the capital increase should improve the growth potential of the unit, by endowing MEA with the financial chest to tackle new opportunities. Implicit 6.7x EBITDA14 multiple in our DCF vs 4.2x blended multiple before We broke down MEA Africa into: (1) MEA Angolan construction activity (6.5x implicit EV/EBITDA14 F ); (2) MEA construction activity in other African countries (7.1x implicit EV/EBITDA14 F ); and (3) MEA E&S (6.1x implicit EV/EBITDA14 F ). All together, we are now valuing MEA at 6.7x EV/EBITDA14 F or 1.25bn EV. Given the impact of the change in methodology in our ME valuation, we want to make clear that should we have kept the previous method (3.9x EBITDA14 applied to ME Angola and 4.5x EBITDA for the remaining MEA after updating the peer group multiples), MEA would be worth 872mn and our fair value of ME would be, 3.3/share (including 10% discount) or +1% also including all other changes produced in the reviewed SoP. Mota Engil Africa - Estimates breakdown CAGR Sales ( mn) FY11 FY12 yoy FY13 F yoy FY14 F yoy FY15 F yoy Construction % % % % 19% Angola % % 615 8% 649 6% 10% Africa ex Angola % % % % 37% E&S % 54 80% 66 22% 76 15% 36% Total % % % % 20% Note: BPI estimates for the historical breakdown of Africa construction revenues. EBITDA CAGR mn FY11 FY12 yoy FY13 F yoy FY14 F yoy FY15 F yoy Construction % % 220 5% 234 7% 24% Angola % % 129 3% 130 1% 14% Africa ex Angola % % 91 8% % 42% E&S % 23 84% 28 22% 33 15% 37% Total % % 248 7% 267 7% 25% Note: BPI estimates for the historical breakdown of Africa construction EBITDA. EBITDA mg % FY11 FY12 yoy FY13 F yoy FY14 F yoy FY15 F yoy Construction 19% 18% -1.0pp 22% 4.3pp 21% -1.0pp 20% -1.0pp Angola 19% 18% -1.0pp 22% 4.3pp 21% -1.0pp 20% -1.0pp Africa ex Angola 19% 18% -1.0pp 22% 4.3pp 21% -1.0pp 20% -1.0pp E&S 59% 42% -17.4pp 43% 1.0pp 43% 0.0pp 43% 0.0pp Total 21% 19% -2.2pp 23% 4.4pp 22% -0.8pp 21% -0.9pp 12

13 DCF implicit exit EBITDA multiple of 3.9x - 4.9x For all MEA units, we have assumed a 13.6% WACC computed using a 6.7% country Risk premium assumption for Angola and 6.5% for the remaining African activities. We have determined the country risk premium using the premium differentials published periodically by Damodaran and applying them to the trading yields of USD bonds from Ghana and finally adjusting to euros (differential of 10Y yields in Germany vs. USD). We also apply a 150bps factor to take into account the political/ economic risk in Angola and also other countries in sub-saharan Africa. Our riskfree rate in euros is maintained for the entire BPI European research coverage at 3.25% and the asset beta for construction is the same for our sector coverage of 0.8. We assume a pre-tax cost of debt (Rd) of 9.0%, a perpetuity EBIT mg of 15.2% for the Construction activity (Angola and other countries) vs. 17.2%F for 2013 and a 15.3% average from while for E&S we have assumed a 34% EBIT mg vs. 34%F for 2013 and a 34% average from For the construction business we have also assumed a terminal year cycle adjustment factor of 80%. In essence, that sets a more conservative departing point to calculate the terminal value, recognizing the higher perceived risk in this sector concerning among others the order intake volatility. As a result, the implicit exit EBITDA multiple (in 2020) for the activity in Africa embedded in our valuation is 4.6x for Angola construction and 3.9x for waste, respectively and 4.9x for the rest of MEA, which seems reasonable. For the sake of sensitivity of our valuation, should we have used no adjustment to the normalized FCF (vs 80% factor in our base case), the implicit EBITDA multiple of our valuation of MEA would climb to 7.5x (FY14 F ) and our fair value of ME to 6.2 ( 5.6 post 10% discount), which compares with the adjusted peer multiple range of 6.8x - 9.0x (7.5x average). In that scenario, the exit EBITDA multiple for the construction business of MEA would be 4.9x - 5.2x. We expect more visibility into MEA with the expected IPO of the unit, a situation that would probably decrease risk perceived in estimates and consequently allow for a more aggressive take in our valuation inputs, namely forecasts. DCF MEA Angola (YE14) normalized ( mn) 2014 F 2015 F 2016 F 2017 F 2018 F 2019 F 2020 F figures EBIT Adjusted taxes Depreciation&amortization (-) Capex (-) WCap needs =FCF PV FCF PV explicit period 441 Perpetuity 401 EV 841 ME Stake (51%) 429 Implicit EV/EBITDA 14 F 6.5x Implicit EV/EBIT 14 F 8.4x Exit EV/EBITDA 4.6x MEA - DCF Asumptions Const. Const. Ex. Angola Ex Angola E&S Beta assets Re 15.5% 15.4% 15.5% Rf 3.25% 3.25% 3.25% Beta of Equity Market Premium 6.7% 6.5% 6.7% Rd 9.0% 9.0% 9.0% Tax rate 35.0% 30.0% 35.0% D/EV 20.0% 20.0% 20.0% WACC 13.6% 13.6% 13.6% Perpetuity period EBIT margin 15.2% 15.2% 34.0% g 2.0% 2.0% 2.0% ROIC 13.6% 13.6% 13.6% Rev. cycle adj. 80.0% 80.0% 90.0% 13

14 DCF MEA Const. Ex Angola (YE14) normalized ( mn) 2014 F 2015 F 2016 F 2017 F 2018 F 2019 F 2020 F figures EBIT Adjusted taxes Depreciation&amortization (-) Capex (-) WCap needs =FCF PV FCF PV explicit period 300 Perpetuity 343 EV 643 Implicit EV/EBITDA 14 F 7.1x Implicit EV/EBIT 14 F 9.2x Exit EV/EBITDA 4.9x DCF MEA E&S (YE14) normalized ( mn) 2014 F 2015 F 2016 F 2017 F 2018 F 2019 F 2020 F figures EBIT Adjusted taxes Depreciation&amortization (-) Capex (-) WCap needs =FCF PV FCF PV explicit period 80 Perpetuity 93 EV 173 Implicit EV/EBITDA 14 F 6.1x Implicit EV/EBIT 14 F 7.7x Exit EV/EBITDA 3.9x SHAREHOLDER STRUCTURE OF THE NEW MEA At the time of the IPO announcement ME stated that it wanted to maintain a strong controlling stake, assumed by the CEO to be at least 51% and announced a share distribution to current ME shareholders of approximately 20% of MEA precapital increase. ME also guarantees to current shareholders the sale at the capital increase price of the assigned shares of MEA. 14

15 ME to keep controlling stake We run three scenarios assuming respectively that ME dilutes its stake to 51%, 60% and 70% of MEA (post K increase). We also assume Mota family will not subscribe the capital increase of MEA, which means that it would end up with a direct stake in MEA of 8.6% or 10.2% or 11.9% in the same order, which also assumes the family does not exercise the right to sell the entire direct stake in the operation. The family should be assigned a 13.6% stake in MEA pre capital increase that results from the product of its current 67.8% stake in ME and the 20% of MEA to be assigned pre-capital increase to ME shareholders. 59mn and 555mn MEA K increase proceeds range In order to estimate the potential capital increase proceeds, we have to assume, among others, the capital structure of MEA (we estimate 300mn of net debt), final stake of ME and MEA valuation. We run several scenarios using, apart from our base-case valuation, the same scenarios explored in the valuation chapter concerning MEA. The range of potential proceeds is as wide as the valuations assumed. We indicate a range of 59mn to 555mn varying from a total 70% stake held by ME post capital increase and a valuation of MEA using a 4.5x EBITDA14 multiple (avg. construction peers in SAR) to a final 51% stake in MEA and valuing it using a 9x multiple. In our base case scenario (ME ends up with 60% stake and MEA valued at DCF with an implicit 6.7x EBITDA14), the capital increase proceeds would be 236mn, assuming the 300mn net debt or 311mn assuming 0 debt. ME Fair value ( /sh) EBITDA-multiple 14 F Fair value upside w/ 10% disc. upside African premium based max Multiple of 9x % % African peers Max (7.8x) % % African premium based avg. Multiple of 7.5x % % BPI s Fair Value MEA (6.7x) % % African constr. peers Avg (4.5x) % % "Newco" valuation (ND= 300mn) ME SGPS stake post K incr EBITDA-multiple 14 F 51% 60% 70% African premium based max Multiple of 9x African peers Max (7.8x) African premium based avg. Multiple of 7.5x BPI s Fair Value MEA (6.7x) African constr. peers Avg (4.5x)

16 K increase proceeds ( mn) ME SGPS stake post K incr EBITDA-multiple 14 F 51% 60% 70% African premium based max Multiple of 9x African peers Max (7.8x) African premium based avg. Multiple of 7.5x BPI s Fair Value MEA (6.7x) African constr. peers Avg (4.5x) Deleverage expected at ME holding level The eventual success of this operation should enable the company to alleviate some pressure on its BS increasing its fire power to continue its international growth. On a consolidated level and using our base-case scenarios discussed above, ME ND/EBITDA14 F would fall from 2.2x to 1.6x (range of ). Our base-case scenario assumes MEA will be left with 300mn net debt. In reality, there is no clear information to make a reasonable estimate. The lower the debt left in MEA the higher the potential proceeds, but we also believe ME will take advantage of the operation to reinforce its balance sheet in addition to the proposal to sell ME treasury stock. Consolidated ND 14 F - ME ME SGPS stake post K incr EBITDA-multiple 14 F 51% 60% 70% African premium based max Multiple of 9x African peers Max (7.8x) African premium based avg. Multiple of 7.5x Current fair value MEA (6.7x) African constr. peers Avg (4.5x)

17 SALE OF TREASURY SHARES (5.4%) TO REINFORCE CAPITAL STRUCTURE As part of the above mentioned process and to reinforce the group's capital structure, the EGM will also deliberate on the plan to sell the group's treasury stock which currently represents 5.42% of the share capital. These shares should be sold up to 18 months after EGM approval (June 2015) at a price not below 15% of the average market price in the week immediately preceding the alienation. At the current market price, the cash in from the sale of all treasury shares (5.42%) should reach 48mn or 41mn assuming the 15% max allowed discount. PRIVATIZATION OF EGF The above mentioned sale of treasury shares should help the company to gain financial muscle to compete for the announced privatisation of the Portuguese public waste treatment company EGF. This should materialise over 2014 with press reports suggesting several interested parties (more than 10 companies including the Portuguese ME and Teixeira Duarte but also the Brazilian companies Odebrecht and Solvi and the Chinese company Beijing Enterprises Water Group), pointing to a valuation of c. 170mn. Empresa Geral do Fomento (EGF) is Águas de Portugal Group's sub-holding company responsible for guaranteeing the treatment and recovery of waste. The management of waste treatment and recovery systems is undertaken by 11 concessionary companies created in partnership with EGF and the Municipalities. These companies process around 3.7mn tons of municipal solid waste a year and serve about 60% of the Portuguese population. We understand that this deal could be very interesting for ME with the possible integration of EGF operations with those of SUMA (ME's waste treatment company and market leader with 51% in the privatised Portuguese waste market) creating a new player with a c.80% market share and benefiting from potential economies of scale. CHANGE IN ESTIMATES +4% EBITDA13-15 ON THE BACK OF EXPECTED RESILIENCE IN MEA'S HIGH DOUBLE-DIGIT MARGIN F top line cut by an avg. 2% mainly impacted by a small cut (-2% from ) in the European E&S estimates (3Q revenues stood 4% below our forecast) and a small (-0.4% from ) cut in top line from African E&S activity. Top line from America was also cut by an average of 2% after lower than expected revenues in 3Q (9% below est.). We also highlight the expected higher negative effect of the "Holding" line related to the services provided by the corporate centre. This should rise alongside the company's increasing international efforts with a continuing insignificant impact at EBITDA level (this movement was already notable in 3Q results). EBITDA estimates increased by an avg. of 4% mainly on the back of (1) better construction margins in Europe (which stood 5pp above est. in 3Q) and (2) improved EBITDA margins expected for the African construction market where the strong orderbook and resilience of EBITDA mg in past quarters above 20% (23% in 9M) have led us to revise our 2014/15 margin to 21% and 20% respectively (from prev. 19% / 18.5%). 17

18 Change in Estimates New Old Chg Sales ( mn) 13 F 14 F 15 F 13 F 14 F 15 F 13 F 14 F 15 F Europe % -1% 0% Construction % 0% 1% E&S % -2% -2% Africa % -2% -1% Construction % -2% -1% E&S % 2% -3% America % -6% 2% Holding % 82% 88% Total % -5% -2% EBITDA ( mn) Europe % 0% 1% Construction % -1% 8% E&S % 0% -1% Africa % 7% 6% Construction % 8% 8% E&S % 2% -3% America % -1% 7% Holding % -19% -17% Total % 4% 6% EBITDA mg Europe 9.8% 9.3% 10.1% 9.1% 9.2% 9.9% 0.7pp 0.0pp 0.1pp Construction 5.9% 3.7% 4.3% 4.9% 3.8% 4.0% 1.0pp 0.0pp 0.3pp E&S 18.2% 18.5% 19.5% 17.9% 18.1% 19.3% 0.3pp 0.4pp 0.2pp Africa 23.1% 22.3% 21.4% 24.1% 20.4% 20.0% -1.0pp 1.9pp 1.4pp Construction 22.0% 21.0% 20.0% 23.0% 19.0% 18.5% -1.0pp 2.0pp 1.5pp E&S 43.0% 43.0% 43.0% 43.0% 43.0% 43.0% 0.0pp 0.0pp 0.0pp America 8.8% 10.5% 10.5% 9.0% 10.0% 10.0% -0.2pp 0.5pp 0.5pp Holding 2.2% 2.2% 2.2% 5.0% 5.0% 5.0% -2.8pp -2.8pp -2.8pp Total 15.6% 15.6% 15.4% 15.3% 14.3% 14.4% 0.3pp 1.3pp 1.1pp Regarding the company's target for 2015, we remain slightly below at top line and EBITDA following our revision (5% below and 1% respectively). In Africa we expect the company in 2013 to surpass its topline guidance for 2015 ( 970mnF for FY15 vs. 706mn already recorded in 9M13) although we remain more conservative on the European market by expecting a 4% drop in 2014 for the Portuguese construction market vs. a company guidance of a flat top line. On Latam, our main difference is cantered on expected EBITDA mg (11%F by us vs. 13% forecasted by the company). For this region we remain on the conservative side. We expect the company to enjoy a strong top line increase (helped by the recent 660mn road works contract awarded in Mexico) though EBITDA mg was never above 11% in this region (between 7.2% in 2010 and 10.5% in 2012 and 8.8% in 9M13). If we assume an EBITDA mg of 13% for Latam (vs. our 11% forecast) from 2015 onwards our FV of ME increases by 14% (+ 0.7/sh) to 5.7/sh (including 10% discount). 18

19 VALUATION AND RECOMMENDATION Revisiting our MEA valuation methodology We are increasing our valuation for ME by 50% or + 1.7/sh mainly due to: (1) an estimate revision of MEA (+.5/sh), now expecting a resilient EBITDA mg of 20% from 2015 onwards vs. the previous 18.5%, strong orderbook (more than 25 months of sales), and orderintake prospects for the African market; (2) a change in the valuation method of MEA (+ 1.8/sh) now using a DCF with an implicit 6.7x vs. the previous 4.2x implicit valuation; (3) the MtM of treasury shares (+ 0.09/ sh), Martifer & Glint stakes ( +0.04/sh); (4) lower leasing and factoring after a 6% qoq decrease in 3Q (+0.09/sh); and (5) a 20mn NWK recovery expected for 4Q12 in line with the average of the last 3 years 4Q NWK movement (+ 0.12/sh). Change in Valuation ( )! "##$ %& '( $# '() &## ''*#+ )# '''(,! - ". "#/!# 0 1. %$## 2 3* but Ascendi continues to be a black box The concession arm of ME (Ascendi - 60% stake and equity consolidated) continues to be a "black box" when considering the information reported (equity consolidated P&L and BS contribution) and its significant weight in ME's asset value. On this unit we revised our estimate to incorporate the cut agreed with the government in the 5 Availability payment concessions (we assumed a 20% cut in rents) partly offset by lower maintenance costs and capex. As a result, we assume a negative NPV of 47mn for ME's indirect stake that we deduct from the reported 150mn investment book value of the respective concessions. For the remaining key assets (Lusoponte, Brazialian and Mexican concessions), we used an adj. investment BV reflecting the changes in cost of capital since the investment date eroding 9% of the investment value. A potential deal either at the level of Ascendi (entry of a new partner) or sale of individual assets could increase valuation visibility and endow the group with means to tackle opportunities in the concession business. 19

20 Valuation of ME s 60% stake in Ascendi mn Equity inv.by ME BPI Method 5 Availability payment concessions DDM Lusoponte Adj. BV Brazilian concessions Adj. BV Mexican concessions Adj. BV Total ME - SoP EV/EBITDA14 F Business mn implicit Method %EV MEA x DCF 54% Africa ex-angola (100%) x DCF 28% Angola (51%) x DCF 18% Services & Environment (100%) x DCF 7% Construction (Europa + Latam) x DCF 17% Property (BV) 25 1x BV 1% Services & Environment x DCF 17% Concessions (attributable equity) 253 BPI (Adj Equity Inv.and DDM) 10% Martifer (37.5%) (1) MV 1% Glintt (2.3% stake) MV 0% Holding costs x EBITDA13 F -1% TOTAL EV adjusted x Net Debt YE Provisions -89 Off balance sheet debt (2) -311 Treasury shares (5.4%) at MV 48 Financial Inv. (3) 170 Minorities (4) -120 Equity value # shares (mn) 205 Value per share ( ) % Small cap Discount 10% YE14 Price Target /share. (2) Factoring and leasing. (3) ex-martifer and Ascendi, inc. Idaqua,Tertir's plots of land and assets available for sale. (4) E&S., Bloomberg. 20

21 Keeping Neutral Recommendation expecting more visibility with the IPO details We are keeping our Neutral Recommendation despite the significant YE14 Price Target increase to 5.00 (incl a 10% small cap discount). The upgrade of our valuation of MEA has been the major driver of the target increase. The DCF of the unit has an implicit EBITDA14 F multiple of 6.7x compared to a peer group multiple range of 6.8x-9.0x (7.5x average). Still, we also recognize that the generated shareholding cascade after the IPO may increase the due holding discount. Some investors willing to gain access to Africa through ME will likely move into the newco and the risk raised from the expected 5.4% placement of treasury shares could also pressure the stock until the placement is done. However, in the meantime, the market should be in the process of removing discounts due to the higher visibility of MEA should the IPO go through. As extensively discussed, investors must realize the sensitivity of ME valuation to valuation input assumptions and the risks involved with operating in the construction market in Africa and Latam in particular. In addition, we also remind the high leverage of ME balance sheet, particularly when including the proportional stake of its interests in affiliates even if the associated debt is non-recourse to ME, as mentioned several times by ME publicly. Even after a significant improvement in increasing the average maturity of its debt ME continues to have to refinance 37% ( 357mn) of its debt in the next 2 years. Interest coverage ratio should improve from 2.0x in FY12 to 2.5x in FY13 and 3.3x in FY15. SENSITIVITY ANALYSIS ME Fair value ( /sh) EBITDA-multiple 14 F Fair value upside w/ 10% disc. upside African premium based max Multiple of 9x % % African peers Max (7.8x) % % African premium based avg. Multiple of 7.5x % % Current fair value MEA (6.7x) % % African constr. peers Avg (4.5x) % % Sensitivity analysis of ME Fair Value to a change in CRP (Before 10% Discount) CRP Angola 4.7% 6.70% 8.70% 10% 4.5% CRP Subsharian Africa 6.5% % %