Everyone here in Oslo, and also good morning to all of you following us on the webcast. We are presenting Hydro's third quarter results today, and as usual, they will be presented by our CEO, Svein Richard Brandtzæg, and CFO Eivind Kallevik. We will also have some time afterwards for questions from the audience and from webcast, and also some one-on-ones in the end. Let's start.
Thank you. The presentation. Prices are falling, but we are also seeing tailwinds from weaker currencies in Norway and Brazil, and also the effects of our commercial and operational improvements. The underlying result of NOK 2.2 billion in the quarter is NOK 500 million lower than the previous quarter, but NOK 700 million better than the third quarter last year. One of the most impressive results in the quarter are related to the alumina cost, alumina production. $217 per ton in alumina cost in Brazil is the best level we have experienced since we took over the Vale assets in 2011. This is due to better production, and especially bauxite production is now lifted to 10.9 million tons in the quarter.
The nameplate capacity is 9.9 million tons , so we are very happy with the development in Brazil, in addition to the fact that the Brazilian currency has been also weaker during the quarter. In Primary Metal, also weaker lower prices, but somewhat compensated also on the currency side with the weaker Norwegian and Brazilian currencies. In the previous quarter, we delivered record results in downstream with a solid Rolled Products result. This quarter we delivered even better Rolled Products result in a quarter which normally, seasonally, is weaker, so we are very happy also with the development in Rolled Products in this quarter. In Energy, the snow that did not melt in the summer, it melted in the third quarter, so we have had high production.
This was negatively affected by lower prices. We announced that we have signed a letter of intent with Vale for their 40% share in the bauxite mine MRN, and we are now performing the due diligence to see if we go through with the transaction. If you now look at the market, we see that there are still oversupply globally. China is still exporting. We see that the demand in China is going down, as well as also some reduction in demand outside China. We are taking down the forecast of demand, total demand this year, from around 5% to around 4%.
If you then go to the supply-demand balance in the global market, including China, we see that there is 1% less demand in this quarter compared to the previous quarter, and about 5% increase since the third quarter last year. This is a reason why we also have taken down the forecast from around 5% to 4%. When we take all the parameters into the equation, we believe that the demand will be a bit lower than what we said in the previous quarter. If you look at the supply side, China still continue to increase production at somewhat lower speed than what we have seen before. We believe that the surplus this year will end up around 1 million tons globally.
If you then look at the situation outside China, also there we saw about 1% lower demand than the previous quarter. In total, 1.5% higher demand than in the third quarter last year with large regional differences. U.S. increasing both at about 2.5% in demand, which is lower than the 5.5% in the previous quarter. We see Europe 1%, very low growth development. Asia 2.5%, and South America - 5%, and with Brazil as the weakest part in South America. If you look at the supply side, there have been some ramp-ups in Canada and some smaller adjustments elsewhere. There are also some curtailments in the U.S., some delayed production ramp-ups in India.
All in all, we expect about 2% growth in supply, and with 2% demand growth as we expect this year. Also, we said it could be a bit higher in the previous quarter. We are now estimating about 2% in the world outside China, and that supply growth at the same level as demand growth. We expect that under supply of about 1 million tons continues for the rest of this year. When we look at the global inventories, it could be somewhat misleading to look at absolute figures on volumes, so we are now presenting the inventory in days. The development in inventory days historically during the financial crisis and as it looks now.
We have seen that inventory days have gone down somewhat since the peak in 2009. During the last months, we also seen that the inventory is somewhat increased. This is not surprising as there is oversupply globally. We should also take into account here that when we are saying around 93 days, 94 days inventory, as it shows in this figure, the numbers are quite uncertain because this is the residual figures after taking into account the reduction in the registered inventories, like LME inventories that have now been reduced with 0.4 million tons in the quarter and now approaching 3 million tons in total in the LME inventories. It is also taking into account the supply and the demand.
If you look at the two biggest aluminum analyst agencies, the discrepancies between these companies are between 2 million tons - 3 million tons of global inventories. The numbers are quite uncertain, but I think the development as we see now is at least a signal that the inventory days is increasing after coming somewhat down, but still at high levels as you see. If you then move over to China again and look at the incentives for Chinese producers to export semi-fabricated products out of China, the blue curve on the slide shows the incentives which has come down significantly after we saw the reduction in the premiums outside China was coming down.
There are still some export, but we see now that after a period of high numbers, we saw a significant reduction this quarter. Since the peak of 0.4 million tons in December 2014, we are now at 0.25 million tons per month. We see that the level now is about 0.2 million tons lower than the peak. If you then take a look at the prices, the realized LME for the company was $1,685 per ton. We have sold 50% of the volume for the fourth quarter at $1,600, and 50% of the premiums at $385 per ton.
The LME traded between around $1,500 and around $1,700 per ton in the quarter, so quite large variation. Premiums in the U.S. are coming down to $160 in the U.S., $135 per ton in European market. Japan traded at or premium price was $95 per ton, but now trading around $90 per ton. If you then take a look at the cost curve, because the question is how long time can this continue with oversupply and the low prices. Looking at the cost curve, we see that around 30% of the global capacity is now losing money, is now cash negative. 60% -70% of this is in China.
This cost curve also has some uncertainties, especially in the flat part of the curve, but there is a significant part of the global production that is now below water. We have seen some curtailments in the U.S. There are also some announcements in China, in the China Nonferrous Metals Industry Association , that there could be 2 million tons -3 million tons of curtailments in China, but that remains to be seen. What is important for us also to see is that the CRU that is setting up this curve are now positioning Hydro in the first quartile, which is a result of the significant cost-cutting efforts we have done now for several years and are continuing with every day, and also the fact that we have tailwind on currencies.
If you then go over to alumina and look at the price development of the alumina, which has been quite resilient on prices, for several quarters, we see now a quite significant reduction. The average price was $292 per ton in the quarter. It is now trading around $259-$260 per ton. In percentage of LME, also reduction from 18.8% to 18.1%, and now trading around 16.7% of LME. Also in percentage, alumina is now coming down. If you then just take a look at the alumina price in Brazilian currency, we see that the weakening currency in Brazil is more than compensating for the reduction of alumina price in dollars.
That is really the tailwind we see on currencies in Brazil. If you then move to China again and look at the export-import balance, we still see increased bauxite import. Now Malaysia is taking over the role of Indonesia as the most important source. We are also sourcing some bauxite from Brazil, but Australia is, of course, also an important source of bauxite to China. Guinea has also now sent a large load of bauxite from Guinea to China, so that can be another source for the Chinese players. Alumina import also increased in the quarter, as you see. If you look at the primary side, it is quite balanced. Scrap import, same as before.
If you look at the same as exports, that has been reduced during the quarter, which is positive as we see it. We also regard the volumes that has been reduced, it's very much volumes that has been re-melted in the West or outside China. If you look at some highlights in the quarter for Hydro, we have invested now even more in our capabilities to supply the fast-growing automotive market. At the Karmøy smelter, we have invested in new ultrasonic equipment to deliver extrusion ingots for automotive purposes. You know probably that we have previously invested in flexible adjustable molds for 6000 alloys for automotive in Årdal and Husnes.
We are on the way to deliver the Automotive Line 3 project in Grevenbroich, where we are increasing our body-in-white automotive capabilities from 50,000 tons to 200,000 tons. We are also doing recycling investments in Weinheim in Germany. In Alouette, Canada, where we have a joint venture smelter, we have entered into a new power contract, 13-year power contract from 2017, which is significantly better than the current LME rate in Canada, which will take down the cost of this smelter. We have signed the letter of intent with Vale, and I will come back to that in a few minutes.
Two days ago, we announced the divestment of the rolling mill SLIM, which is in the Italian market, with capacity of 92,000 tons. This rolling mill has been running around 50,000 tons -70,000 tons during the last years. It is producing more standard strip products and also some standard foil products. In the market, as we see when the premiums are going down and the demand is weakening, we have a flexible production system through the fact that we are able to reduce re-melting of standard ingots in the primary smelters and also taking down re-melting activities in the re-melters that are distributed in the market. This is a flexibility that we are now utilizing.
If you then move to Brazil and the letter of intent with Vale, with the intention to take over their 40% share in MRN. MRN is one of the most cost-efficient and the biggest bauxite mines in the world. It is a first-quartile cost-positioned bauxite mine, and with annual capacity of 18 million tons. We are taking the majority of this volume through an agreement with Vale so far to serve the alumina plant Alunorte, but we also have 2.5 million tons -3.5 million tons bauxite for the external market.
Today, we have 5% ownership, and if we decide to go through with a transaction after the due diligence, we will have 45% ownership and then have also more to say with regard to the strategy and the development. We get access to the income stream, which is the difference between the price of bauxite and the cost of production of bauxite. We also, of course, take part in the sustaining CapEx and the development of this mine going forward. This will be a very important strategic positioning of Hydro to strengthen the raw material capacity in the company going forward.
If you then look at the alumina cost development, as I mentioned, we had the lowest cost in this quarter since we took over the assets in 2011, $217 per ton. As you see, the price of alumina has gone down with $19 per ton, and the cost or the margin has been reduced with only $3 per ton. We are now on the way to deliver the B&A improvement program that will be finished this year, which takes down the cost with NOK 1 billion in Brazil. Also the support from the currency, the tailwind from the currency is helping us in this respect.
As I mentioned, we are improving continuously the bauxite production in Paragominas and now delivered 10.9 million tons annualized rate in the third quarter, and now operating further at same level into the fourth quarter. If you then move over to primary, also here, lower prices, but somewhat compensated by also lower cost, $100 per ton lower cost compared to the previous quarter and $200 lower than the third quarter last year. Again, still $200 lower margin, EBITDA margin compared to the previous quarter. We continue also here with improvements, the $180 program that we have mentioned and talked about before is on track.
Of course, the development is also affected by lower alumina prices and also the advantage with lower currency Norwegian kroner and Brazilian reais compared to dollars. In Rolled Products, as I mentioned, we are quite happy with the development. On volumes, it was 2% improvement in this quarter, which is seasonally normally a weaker quarter than the previous one. Automotive growth was 6% and also general engineering 4% improvement since the previous quarter. If you look year-on-year, 2% improvement in volumes. Also since the third quarter last year, automotive sales is 6% higher this quarter than the previous quarter. We're now reaching more or less the full speed in automotive.
We are now waiting to finish the Automotive Line 3 in Germany to be able to deliver more volumes. General engineering, significant improvement, 14% higher than the previous year. Also here we see improved margins, which is adding to the result in Rolled Products. In Sapa and Extrusion, markets in U.S. i s improving. Good development in the U.S. market, not only due to the automotive growth, but also the fact that building and construction in U.S. is also growing. In contrast to Europe, where we see a fairly weak building and construction market in aluminum. 6% improvement in the U.S. year-on-year, and 2% in Europe as a growth in extrusion.
Moving over to energy, we started earlier this year with a fairly low reservoir level. We had a cold start of the summer, and as I said, the snow that didn't melt in the second quarter have now melted in the third quarter. The reservoir level now is close to the 10-year high average level and is now at a level of 93.7% in the NO2 region. Prices has gone down from NOK 171 per MWh to NOK 102 per MW h as a result of high production. But now in the fourth quarter, we see that it is prices are increased already up to now with about 70%. I'll leave the word to Eivind, please.
Thank you, Svein Richard. Good morning, everyone. Let's turn our attention to the financial results for the quarter. This quarter, we have delivered an underlying result before financial items and tax of NOK 2.2 billion, roughly 50% improvement compared to the same quarter last year, and roughly NOK 0.5 Billion down compared to the second quarter this year. Falling prices were by far the largest negative driver this quarter, reducing results by some NOK 1.6 billion between the quarters. Premiums continued to soften at the same time as the LME price came further down, affecting both the realized all-in aluminum prices as well as the realized alumina price. The realized PAX alumina price followed a downward trend and was on average some $30 below compared to the second quarter of this year.
Overall, these developments gave us a 12% lower realized all-in price on the metal side and some 7% lower realized alumina prices in the third quarter. On the other hand, roughly half the negative price effect was offset by the currency tailwinds, as both the Brazilian reais weakened against the dollars, as did the Norwegian kroner. The net currency effect, the net of both the revenue and the cost side contributed with roughly NOK 800 million between the quarters. Of this, roughly 1/3 was related to the BRL development. We had a combination of higher volumes and margins, which also contributed with positive NOK 200 million. We had higher margins both in metal markets as well as in bauxite and alumina.
On the other side, bauxite, alumina, Rolled Products and Primary Metal also contributed with some higher sales volumes. We have also seen lower costs, net of the currency effects in the third quarter, supporting results with another NOK 200 million. This is a combination of seasonally lower fixed costs within the Rolled Products business areas, lower costs in Paragominas, as well as declining raw material costs linked to the falling LME and the PAX index. On the other hand, net of currency, Alunorte and Energy had somewhat higher costs in this quarter. I'll get back to each of the business areas a little bit later on in the presentations.
If you look at the key financials for the quarter, the revenues for the third quarter decreased by roughly NOK 840 million compared to the second quarter. Well, this is primarily driven as a result of falling prices throughout the value chain, partly offset by the strengthening dollar. The volume developments varied a little bit between the different business areas, but nothing amounting to a negligible and very small total effect. This quarter, we have excluded a loss of NOK 586 million in non-recurring and timing effects, bringing us to the underlying EBIT of NOK 2.2 billion. I will get back to that in more detail on the next slide. Financial items this quarter was a - NOK 3.3 billion compared to the positive NOK 258 million that we had in the second quarter.
This large increase is driven by unrealized net foreign exchange losses of NOK 3.2 billion. This is mainly driven by the unrealized currency effects. Two items: one, is the dollar debt that we carry in Brazil. The Brazilian reais has weakened some 28% compared to the second quarter, giving us a net effect of NOK 1.7 billion in net unrealized effects. The second part is the embedded derivatives that we carry in the power contracts relating to the euro. The euro has strengthened 8% to the Norwegian kroner, giving us a net effect of more than NOK 900 million related to the embedded derivative. As a result of that, we get an income before tax of negative NOK 1.7 billion compared to the positive NOK 3 billion we had in the second quarter.
Income tax in this quarter amounts to an income of some NOK 367 million, driven by the large unrealized currency losses. We get a net income of negative NOK 1.3 billion, down from positive NOK 2.1 billion last quarter. If you look at this from an underlying perspective where we exclude the unrealized currency losses, we have a positive net income of NOK 1.4 billion, down NOK 400 million from the second quarter. That gives us an earnings per share this quarter of 0.61 NOK per share. We turn to items excluded. We had NOK 586 million that we did exclude from the result in Q3. The unrealized effects on power and raw material contracts gave us a positive effect of NOK 73 million this quarter.
That comprises of NOK 120 million in positive effects from the feedstock cost contract. That is driven by the embedded derivative on LME and coal, partly offset by the weakening NOK against the dollar. Moving in the opposite direction on this line is unrealized derivative losses on power contracts, driven, carried in Primary and in Rolled Products. The unrealized derivative effects on LME had a negative effect of NOK 249 million. This is a net effect of unrealized gains and losses on short and long positions throughout the value chain of the company. The metal effect this quarter in Rolled Products, - NOK 344 million, that is driven by the negative development in both LME and in premium, measured in euro.
In Sapa this quarter, we have excluded NOK 65 million that is partly driven by continued restructuring charges in China, in addition to some timing effects on derivatives. With that, let's turn over to the business areas, and we'll start with Bauxite & Alumina. The underlying EBIT for B&A this quarter was NOK 628 million. This is an improvement of NOK 150 million compared to the second quarter. The Q3 results was negatively impacted by 7% lower realized aluminum prices. That is a result of the drop that we see in LME, as well as the PAX alumina index.
This was partly compensated by more volumes sold on index in the third quarter compared to second quarter, both due to the higher aluminum production, in addition to higher volume sourced on legacy LME-related contracts. When it comes to production, we did increase aluminum production Alunorte. As Svein Richard Brandtzæg mentioned, we also have record production in Paragominas of some 10.9 million tons on an annualized basis. On the cost side, in Paragominas, we see a reduction in operating costs compared to the second quarter, which the second quarter was of course affected by the extended maintenance we had. In addition, fixed costs came down this quarter, driven by the high bauxite production.
In Alunorte, operating costs increased slightly, due to the higher energy price that we paid, but again, partly offset by lower bauxite costs as well as caustic prices. These negative effects was more than offset by the weakening of the BRL versus the dollar. In Q3, we realized a historical low implied alumina cost of $217 per ton. This is a reduction of $16 per ton from the previous quarter. Again, very much driven also from the BRL versus the dollar. Let me just spend some words on explaining the currency effect in B&A. Around 15% depreciation of the BRL on average against the dollar in just one quarter has a significant positive effect on the results.
On a short-term basis, roughly 40% of the cost we carry in Brazil is exposed to the BRL development. Since our sales then are in dollar, the weak BRL has become a major tailwind on reducing the cost side of our P&L. In fact, if you look at the Q3 currency movements, that has an impact of more than NOK 300 million for the B&A business area. There is, of course, a flip side to the weakening BRL, and that is the inflation side in Brazil. The estimation this year is that inflation will come in around 10%. That will clearly have an impact on the cost of both our own employees as well as the contractors that we hire to do services in the industry.
We worked hard to offset the cost inflation in Brazil through the B&A improvement program, which, like I mentioned, and that program is still progressing according to plan. If you look into the third quarter, we do expect to continue to lift aluminum production. In addition, also bear in mind that we will have somewhat lower volumes sourced on PAX. As such, our net index exposure should increase coming into Q4. If we turn to Primary Metal, we see that the underlying EBIT in Primary Metal has almost halved from Q2, where we had NOK 1.4 billion to NOK 762 million in this quarter.
The drop in Primary Metal is largely explained by the falling all-in prices, as realized LME prices fell by approximately 7%, and the Hydro realized premium fell by 33% between the quarters. Overall, the declining prices drove the results down by some NOK 1.1 billion. This negative effect was somewhat offset by the currency developments, contributing roughly NOK 400 million net between cost and the revenue side for that entity. Furthermore, we also had some higher production and higher sales in Q3 while fixed costs remained relatively flat. Looking into Q3, we've sold about 50% of the production at approximately $1,600 per ton, excluding Qatalum. On the premium side, we have booked about 50% of the premiums at $385 per ton.
However, the remaining 50 percent is more exposed to the standard ingot premium, and as such, we estimate that the average for Q4, the premium will likely be in the range of $250-$300 per ton. Lastly, on Primary Metal, we do expect somewhat lower sales, which is seasonally common for the fourth quarter. If we quickly turn to the Qatalum financials, our share of net income for that quarter reduced significantly from some NOK 245 million in the second quarter to NOK 26 million in the third quarter.
Now, while Qatalum continues to deliver solid operational performance, producing above nameplate capacity on a continuous basis, its earnings were clearly affected by the falling all-in prices, as well as somewhat lower sales in Q3 compared to the second quarter, only partly offset by lower raw material costs between the quarters. For the next quarter, we do expect higher sales from Qatalum, with the overall results being largely impacted by the development in the all-in metal price. We turn to Metal Markets. We saw significant improvement in underlying EBIT, up to NOK 291 million, following some weak results in the first half of 2015. If we exclude the currency and ingot valuation effects, the result was NOK 189 million, NOK 168 million up from the last quarter.
The improvement comes from significantly better results from our sourcing and trading operations on the moderating and more stable premium development compared to the large trading losses that we realized in the first half of the year. At the same time, as opposed to the first strong half of 2015, results at the remelters weakened as they were impacted by seasonality in addition to lower volumes and margins. Let me just underline as well that we also have intentionally scaled down the remelt production in Q3 reflecting to protect the margins and the business operations that we have. If we look into Q3 into Q4, we expect somewhat weaker margins and seasonally lower volumes at the remelters in our company.
As a cautionary part, remember that the trading results can and will be volatile and are so by nature. However, everything else equal, and excluding the currency and inventory valuation or ingot valuation, we do expect Metal Markets to deliver around NOK 100 million in fourth quarter, which is close to what we guided on, as an ongoing earnings potential. If we look to Rolled, the underlying EBIT increased by NOK 16 million to NOK 331 million this quarter. As Svein Richard Brandtzæg mentioned, it's in fact an all-time high result for this business. Particularly impressive, I think, is the shipments. They were higher in terms of volumes compared to second quarter. This is primarily driven by the strong automotive demand, but also better demand for high for the general engineering products.
It comes from a combination of factors. One, of course, the substitution trends that we talked so much about. There is a positive development, some of the industrial production market that we deliver to, in addition to the fact that there seems to be less imports of products into Europe, giving us a more benign market in terms of the supply situation. Other effects, we have seasonally lower fixed costs in this quarter, and there is a small positive net currency effect also for Rolled Products. On the other hand, when it comes to Rheinwerk, the results in Q3 were clearly impacted by the falling all-in metal prices. Looking into Q4 is a seasonally weaker market, and it's also a period where we do a lot of the yearly maintenance.
We should expect lower sales volumes as well as increased maintenance cost for Q4. As usual, the Rheinwerk smelter, the result there will be impacted by the development in the all-in metal price. Lastly, we have divested or signed an agreement to divest the SLIM Rolling Mill in Italy that was announced on Monday. We expect that transaction to close in the fourth quarter. The impact on our reported figures is estimated to be in the range of EUR 45 million -EUR 55 million loss. This number will then be excluded from our underlying performance or underlying EBIT. If we look to energy, there was a slight improvement in underlying EBIT of NOK 12 million between the quarters, giving us a result of NOK 191 million in Q3.
The results are driven by a high production of 2.8 TWh, an increase of 700 GWh, compared to the second quarter. This is driven by the large inflow of water into the reservoirs, driven by snow melt as well as wet weather in Southern Norway in this quarter. The volume effect from a financial perspective has an impact of more than NOK 100 million in this quarter. Working in the other direction, based on the high power production, that has brought the spot prices down in the NO2 and NO5 area where we have a long position. Came down to close to NOK 100 per MWh or 40% compared to second quarter, partly offsetting the positive effect we have on higher volumes.
Finally, production costs also contributed negatively in this quarter, primarily driven by higher property taxes. Let me just remind you that property taxes for the full year of 2015 is the same as 2014. Due to an accounting rule change, we now have to incur the cost in the books as we receive the invoice, so you have volatility. Q1, Q3, we have high property taxes, Q2, Q4, relatively low, and the swing is close to NOK 50 million between the quarters. Looking at the next quarter, as always, there is a large uncertainty around the power production that's driven by precipitation and weather conditions. We have started this quarter with very high reservoir levels, and average power prices so far this quarter is around NOK 180 per megawatt hour in Southwest Norway and NOK 190 per megawatt hour in NO3.
We turn to Sapa. We have delivered, or they have delivered rather, continued improvement in results in underlying EBIT, more than doubling compared to the same quarter last year. It's primarily driven by strong North American demand, the ongoing improvement in restructuring activities, but also supported by the currency development. The seasonality between Q2 and Q3 is probably less pronounced this period than what normally should be expected. The EBIT between our share of results came down from the NOK 202 million, or is NOK 202 million, down NOK 40 million compared to the second quarter. This is primarily driven by two things. First, the seasonal weaker demand was partly offset by a strong demand in North America.
The second part, remember that second quarter results was strongly impacted by the rapidly falling product premiums in North America, having a negative impact on the Q2 results of some NOK 50 million. The restructuring program initiated by the company in 2013 targeting annual synergies of NOK 1 billion to be delivered by the end of 2016 continues well ahead of plan. Some quick comments on other eliminations. This has declined from a positive NOK 333 million in the second quarter to NOK 12 million this quarter. We've just been through the Sapa results, so let me focus on the eliminations. This was reduced from the NOK 338 million that we saw in the second quarter down to a - NOK 13 million in this quarter, a change of some NOK 350 million.
The driving force behind this is that there is some falling margins in the Primary Metal side and the downstream operations, driven by the fall in all-in metal prices. On the other side, working in the other direction, is that we have improved margins on the bauxite and alumina sales in the company. Now, if we adjust for the internal eliminations and Sapa, the result is NOK 95 million in charges for common services and businesses. This is somewhat down compared to the NOK 150 million in Q2 and somewhat down on the long-term guidance of NOK 150 million, which we got on this line. We turn to the net cash development between the quarters. We started quarter with NOK 700 million in net cash.
We've generated an underlying EBITDA in the period of NOK 3.4 billion. Also in this quarter, we're starting to see net operating capital effects. We've taken out NOK 2.1 billion in net operating capital. This is partly driven by reduction in finished inventories in the company, but there's also clearly a price effect from falling all-in metal prices in this number. Taxes and other adjustments reduced the net cash development by approximately NOK 1 billion. This is partly backing out non-cash elements in EBITDA, as well as the metal effect in Rolled Products. We've invested NOK 1.2 billion in the third quarter, bringing the total investment so far in up to NOK 3.3 billion, and I will address this also on the next page.
Finally, there is a combination of currency effects, but also dividends paid to minorities, amounting to some -NOK 700 million in the quarter. Put all together, we end the third quarter with NOK 3.3 billion net cash for the company. If we look to CapEx, I think we should spend just some time on this. When we end Q3, we now see that or expect that the CapEx will be reduced by roughly NOK 1 billion compared to the previous guidance, from NOK 6.8 billion down to NOK 5.8 billion, roughly a 50% reduction for 2015. There's really two main reasons for this decline. First, there's been some deflationary pressure on CapEx driven by the currency development.
This hits in particular the B&A and Primary Metal operations in Brazil amounting to some NOK 200 million. Secondly, as you should know, we always have a strong focus on cost and value optimization. If we can postpone investments to later periods, we will always do so, as long as we can do that without jeopardizing the underlying operations. This is also what you see so far by moving NOK 800 million into previous or to later periods. Clearly, when we look into Q4, there is still NOK 2.5 billion left to be spent.
We will still work to bring that further down, and there should potentially be more risk on spending less than the NOK 5.8 billion than spending more than the NOK 5.8 billion on aggregate for the year as such. Finally, just a few words on the net adjusted debt. This was further reduced in this quarter by NOK 1.7 billion to NOK 9.3 billion. Main reason for the decline is of course the net cash position that we've just been through. Working in the opposite direction is the increase in net pension liability to NOK 7.1 billion in Q3 from the NOK 6.3 billion that we reported in the second quarter.
This is basically a translation effect from translating the German pension obligations into Norwegian kroner, and then with a stronger euro, this is the effect that you get. With this much, this is Svein Richard Brandtzæg. I'll leave the word back to you.
Thank you, Eivind. As we now undertake the last quarter in 2015, we still have strong focus on improvements along the whole value chain. We have good control of the enablers, as you probably have seen in Brazil with regard to, for example, the bauxite production, which is now running at the speed of around 11 million tons. We have lifted somewhat the alumina production also in Brazil, and we are continuing to stabilize and increase the alumina production at Alunorte towards the nameplate capacity. In addition to the improvements we are doing, we also having a strong focus on working capital. We are taking it gradually down this quarter. It has, of course, been helped by lower prices, but also inventories are being reduced.
We have now sent several people to MRN to perform the due diligence, and when we have all the necessary details in place, we will take the final decisions if we move forward with the transaction with Vale. In all products, we have several steps now, for example, the investment in body-in-white capacity at Grevenbroich increase the body-in-white capacity from 50, 000 tons to 200,000 tons, where several customers are waiting every day for this capacity to come forward. This project is moving according to schedule, budget and scope, and the same is the case with the UBC recycling facility that we are now building up in Rheinwerk. These projects will be moving forward according to plan.
At the same time, we have divested the rolling mill SLIM in Italy, which is again the strategy of Hydro to high-grade the portfolio and strengthen the company in a situation where we can still say that the markets are challenging. Thank you very much for your attention.
Okay. We open for questions from the audience. Yeah, there is a question there.
Sigurd Sanna, stockholder. I have a question about car industry specific Ford making this aluminum car F-150. I think you follow that project. What do you think about that? Is it a future in that or is it Yeah. Is it about the future?
Thank you for the question. Well, I think you are thinking about the Ford F-150, which is a huge transition of a car from a steel car to a 100% aluminum car. This is one of the most, if not the most sold car in the world. It's a SUV, but it's quite a big car and contains a lot of aluminum. The amount of aluminum that goes into this car corresponds to more than our share in our Qatalum's metal, just to give you a size of the demand from only this one car model. This is a big step for a car company, and it seems that this car has been very successful.
What we have seen in Europe is that automotive companies are changing parts of the car gradually from steel to aluminum, like changing doors, bonnets, roof, and so on gradually. The Ford F-150 you can say is a big bang. It's a really big transition of the car. We also hear now from Ford that they are planning to do this with other models. It seems that this has been well received in the market. You all know that when aluminum is moved into cars and replacing steel, fuel consumption goes down, CO2 emission goes down.
You get better driving performance, and it's also in fact a safer car due to the fact that aluminum is better in absorbing energy than steel. There are so many advantages with this that we see, as I also mentioned, it is the fastest growing market segment as we see it now, automotive, and that is also why we are investing in new capacity for body-in-white for the OEMs in the future.
Thank you for a very nice question. Any other questions from the audience? Do we have any questions, Pål, from webcasts? Yeah.
We have one question from James Gurry in Credit Suisse. Will the intention to purchase the 40% of Alumar preclude a significant increase in shareholder returns in early 2016?
I think when it comes to dividends and payback to investors, clearly the most important part for us is to ensure that we can maintain the base dividend that we have of NOK 1 per share, or 40% of net income over the cycle. That's the base fit, I think. We will have to look as to what elements would, you know, present itself also as we get through Q4, and then we will have to get back to potential increased shareholder returns as we pass Q4.
We have another question from Jason Fairclough in Bank of America, Merrill Lynch. How far are you on the cost savings journey, and how much is still to come? Could you deliver another $300 per ton in primary?
Of course, we would be happy if that could be possible. I cannot promise that, but I can promise that we continue every day with the improvement efforts. What we have seen over the years when we increase the momentum of the improvement efforts, we see that there are new opportunities that we didn't see some years ago, and we continue to see new opportunities. The $300 improvement program is difficult to repeat due to the fact that it was, of course, a comprehensive program that was consisting of I would say several hundred different elements. We still see a lot of opportunities in Primary Metal to reduce the operational cost, and we will of course work with that.
First of all, we are going to deliver on the $180 program for the joint ventures, and we will come back on further efforts in the fully owned smelters at a later stage.
We have a final question from Hjalmar Ahlberg in Kepler Cheuvreux. Can you please give some more indication on how realized alumina prices will develop with current market prices? Will you realize prices close to the Platts Alumina Index or below?
Yeah, this really depends on the mix, on what we have on the old legacy contracts, priced on LME and how much we have to sell on the index, on the spot market. This year on average, we expect to deliver 70% on old alumina LME-related contracts and 30% on the index. Of course, as we go further and we go into 2016, we will have more volumes exposed to the Platts Alumina Index, giving us the possibility to lift the average price for our products sold.
Okay. Any other questions from here? No? Then I would like to say thank you very much for coming and have a wonderful day.