Mark Bristow - CEO & Executive Director John Steele - Group Technical & Capital Projects Executive Simon Bottoms - Group Mineral Resource Manager Graham Shuttleworth - Financial Director & CFO

Howard Flinker - Flinker and Co. David Haughton - CIBC Capital Markets Luke Nelson - JPMorgan Chase & Co. Tanya Jakusconek - Scotiabank

Ladies and gentlemen, welcome to Randgold's Q2 Conference Call. I will now hand over to Mr. Mark Bristow, CEO. Sir, please go ahead.

Thank you very much. Good afternoon and good morning ladies and gentlemen as you we presented our quarterly release to the market at the London Stock Exchange today and again the plan is softening to work just quickly take you through the presentation touch on the highlights and give you an opportunity to ask questions. So as you all appreciate and we would have looked at our results already, after a soft start to the year there has been a steady improvement and the results of this quarter really put us in a position where we can point to our ability again to come in within that guidance and right now we are guiding within the mid-range of our guidance for 2018. As is usual I will start with the health and safety and on top of the strong production quarter we were able to record a good safety and health performance with both the last time injury frequency rate coming down materially and we had only one loss time injury this quarter and more importantly the total injury frequency rate decreased further, in fact this quarter it's the lowest levels that it's ever been. The malaria again continued and again brought that infection rate down and as you're probably aware we have been managing the recurrences of [indiscernible] and the DRC medical team has become quite an expert in assisting in partnership and to dealing with those outbreaks. On the environmental points we again improved our power and water usage efficiency and moved all our mines on to the new ISO standards and then turning to the group summary these are our KPIs for the quarter which you know point to a robust all-around performance with the on or ahead of plan apart from Tongon which we will touch on later and some good results from our Brownfields and Greenfields exploration and I point out that we have paid our $189 million dividend as promised. Just digress a bit on that subject of dividend and the whole underlying ethos of the Randgold way and here is an interesting comparison as you can see our dividend yield of 2.7% is significantly higher than that of our peers and we have always prided ourselves in capital allocation and I think we can safely say we are one of the industry leaders in that as well as and as you can see how we have got the adjusted EBITDA margin and I would point out that again we lead the industry in this and any another or most other financial ratios and I would just remind everyone we have never impaired either. So that's what's separates us from the industry and I came back to this at the end of the presentation. These are the financials for the quarter and really support the KPIs, I touched on earlier and we expect this performance improvement to be maintained in the second half of the year. I would also just point out that we have always been about profitability. Earnings per share was impacted by operational foreign exchange losses in the current quarter that compares to the gains we had in the previous quarter and that's the way it is and also quarter two is always a hard cash tax period for us and again that's a good thing to pay tax as we have always said. [Indiscernible] operations started usual with Loulo, Gounkoto complex, strong performance from the underground mines more than offset decrease and the contribution from Gounkoto which is busy with the super pushback and as a consequence of this production was up, costs were down and profits for mining increased. Also as recently announced we have an agreed 50% tax exemption for four years for Gounkoto to support the super pit development and negotiations with the government really were conducted on a win-win basis and both sides gets something out of this because the extra five years on the operations keeps Loulo, Gounkoto complex about 600,000 ounces over more than 10 years and it also more than doubles the revenue that the government gets out of the compared to the original feasibility study for Gounkoto. These are the numbers and as you can see the 4% increase was back of improved throughout and this was largely driven by the fact that we mine some of the satellite pit oxide [ph] and that help us get this throughput and the whole objective of that was to offset the lower feed from Gounkoto. If we look at Loulo standalone just reinforces the performance of our underground miners and Loulo generating well and it was formented by the feed from the [indiscernible]. These are the results and just reinforced what I've just said. I would just point out that the grade of 4.1 grams, if you look at a lot more tonnage fit into the plant, the underground grade was just around 5 grams and the open pit [indiscernible] all came in at 2.6, and it was that extra throughput which we moved from Gounkoto to Loulo that helped push the numbers for Loulo this quarter. That's the Loulo underground update which really is on track, just a point of the development Yalea as I pointed out a couple of quarters we had significant developed resources in Yalea, we wanted to cut back a little bit and redirect the development to the more the southern decline. We've done that and we were able to do that, that's why the lower development in meters for this quarter. But the production in the form of tonnes mined was exactly as planned. We continue as we do across our portfolio in maintaining a strong focus on mineral resource management. And again, you can see how our drilling continues in the south of the Purple Patch at Yalea depth. And we have a new extension to our plus 10-gram orebody highlighted there in the yellow outline. All the holes drilled they average plus 10 grams and a width of 9.5 meters. And there are many other opportunities on the Loulo permits, and exploration continues to test multiple targets. The most advanced one is Loulo 3 where drilling has been completed and the block model has been updated. And as part of an initial review of the trade-off between the open pit and underground's mining plans for this orebody. And at Gara, the main system remains open to both the north and the south and we're working on that. And again, we've been working on some very interesting results in Gara West, which is a subparallel mineralized structure west of the main Gara orebody. On Gounkoto standalone, again gold production, as I already pointed to, was down but in line with plan as we deal with the pushbacks. And at the same time, mined through a lower grade portion of the orebody. The big news is the tax rebates, which we got under the investment convention, and this will add more than $100 million to the value of the super pit project. Numbers show the impact of the pushback constraints on the grade. But as this progresses, we should start seeing an improvement through to the end of the year, and the reserve grade at Gounkoto is 4.6 grams a tonne. The Gounkoto permit is also - has a good share of prospects and near mine targets that we've successfully converted into our plans as the MZ board done plans. And at Faraba North, we have a large scale mineralized ten - trend which is turning into a very interesting target, and the next slide is really just a long section of that. You see quite a strike length - very similar in ways to Yalea and that the shallow drilling didn't produce the grades. But as we drilled deeper, we hit consistently high grades along the entire struck. And another question is, does it continue or is it going to be more like a Sofia, narrow high grade at depth and that's the big challenge as we manage this - and so the stars are the next round of boreholes and we'll continue to test big step - depth extensions. And again, the Loulo, looking back, zooming out at the Loulo district, we've got quite a lot of work on that whole struck. It's been a prolific producer of substantial gold deposits, and more recently, we confirmed the re-award of a new Bambadji joint venture permits as sweet which you can see in bold on the Senegal side of the river but adjacent to the Loulo Gounkoto permits. And we've already got about 22 targets that we need to start working through and prioritizing. Morila continued on a - and delivered on a steady quarter, a little bit delayed in the Ntiola permitting but we're now fully in the swing of mining at Senegal, continued to contribute to the Morila business plan up towards the end of the year. And higher cost - these are the numbers, attributable mainly to the late startup of Ntiola operations, and again, we continue with our closure work and the move of the - of ultimate final rehabilitated Morila footprint into an agripole agri-village style concept. And again, we're working to get this project endorsed by government as an official government-sponsored project. And we hope that we'll be able to prove that this proof-of-concept would give options and formulate our after mine closure strategy as a group. Moving south to Côte d'Ivoire, Tongon came back strongly after dealing with work stoppages that straddled the first and second quarters. And again, the third quarter started with another work stoppage and I - we eventually decided to lock the workforce out. It was led by a small group of people and we just felt that we had tried unsuccessfully to find a solution which - where we were constantly forgiving and almost rewarding bad behavior. And with the support of the traditional leaders in the community, the community leadership, local authorities and the government, we engaged in dialogue and agreed that we were going to force the rule of law and stop this constant intimidation and holding to regimen and effectively where a handful of people are stopping people from coming to work. We've effected that. We are now in the process with government to restart the operation and we are selectively employing and we will pull back up to full production around the middle of this month and we've revised our guidance to 250,000 ounces for the year. These are the numbers, speak for themselves. One thing about the Tongon team, they're very efficient. They get that very quickly. I mean they've done a jolly good job in managing the situation there again, and we've got no doubt that once we get up and running, we'll be full bore then and delivering sort of efficiencies that you see here. We did also - this last quarter have to deal with a mill motor shaft failure. And - but the team, again, did well in managing it, and we are running on a slightly lower card motor at a moment while we wait for the new 8-megawatt motor to arrive, but they managed it pretty well. The cost, sorry, just to go back. I don't - if you can go back. The costs that you see are a little up, and that's because of the next slide. The power balance between grid and diesel, thermal-generated was quite high this last period, as it is at this time of the year and it will continue to be like that until we get the ring line installed, until the tar utility installs the ring line to ensure more stable tar. And when is that due, John, it's this year, is it?

Yes.

End of this year. And then we move on to our portfolio, particularly starting with the Tongon - Nielle mining permits. We've now got some really advanced targets which are starting to demonstrate the potential to add to the life of the mine in, particular the Djinni and Mercator projects. And again, that whole strike length we've now traced the structure that hosts the South and North Tongon pits down to the south and we've just recently completed the eco-evaluation of that entire struck and we are waiting for the results. On the greater portfolio, in the Ivory Coast, our greenfields teams continue to progress our extensive portfolio. And Boundiali, specifically, we recently completed a VTEM survey, detailed VTEM survey along the entire length of the Fonondara structure and adjacent structures. And again, that - the results coming out of that work is really highlighting the definition of additional targets along the entire Fonondara structure and the extensions to that. We also did a VTEM survey on Mankono, and again, that has really helped us link on - on what we know in the Mankono permits with the Endeavour part of that joint venture now, and again, the work highlighted a number of intrusives that have the potential to be mineralized like the Mankono main orebody, which is a low-grade intrusive host of ore deposits. Leaving West Africa and going across to Central Africa. Kibali was clearly the star of this quarter, this past quarter, posted a record quarter again, second in a row. On every front, production, profit and costs and it's achieved while ramping up the underground production and transitioning to owner mining and it also started the commissioning of the Azambi hydropower station. So busy quarter for the team and a job well done. And it's on the back of this we are guiding that we'll be there near abouts in the next two quarters largely in line with quarter two as far as production and performance goes. and with that, we are guiding plus 3% - I mean plus 5% overshoot on the beginning-of-the-year guidance, maybe a little bit more. And really, on a net basis, we have reinforced our guidance. It's still within the range that we guided in January for 2018 because what we've taken out of our guidance for Tongon, we put back in to Kibali. These are the numbers. That speak for themselves. All good with a grade improving material. One point I would point you to is the recovery is now at feasibility and we always said that would happen as we get closer to processing the dominant feed from underground, and that's what's happening now. Underground production continued to ramp up, and the most important is the ratio here between hoisting and trucking - tramming, and you see we are slightly above nameplates mining because of the way that we're tramming some of the ore, not fully - not just using the shaft, and that was good and that drove the improvement in grade in the previous slide, and that's where we will sit now. We're pretty much at steady-state going forward. And you would - I would remind you that our guidance of 7 50, 7 50, 7 50 [ph] our the next three years that already we're showing that we can manage that comfortably with our current run rate. Just as the - with the Azambi commissioning, we're now getting to the point where we will be able to deliver on our low-cost or more affordable power profile towards closer to $0.10 a kilowatt hour on an annualized basis, and that's really the very foundation of the profitability at Kibali and the basis on which we developed the project. The capital projects, there's a picture of the KC - the Azambi official opening and that brings really most of our big capital projects to an end. We've still got the tailings dam and wall that's being - lift that we're going to do and also that we're busy with. That is ahead of schedule and trending under budget. And we've also got a project - slag storage project and an upgrade on the [indiscernible] plant which we're busy with. But other than that, we already start winding down our capital as we've indicated in the market before. Then move on to our exploration drive. And you'll remember we cut back on the reserves and we transitioned from geostatistical to sort of hardwired orebody models. We've said at the time, we'll be bringing a lot of those reserves back as we drill them out. We got a big focus on those three areas that are indicated by the red arrows. And the potential within those zones at this stage is more than 2 million ounces. So exciting stuff and there's still more in the pipeline as we really get ahead around these models of these - sort of the cluster of orebodies that makes up KCD. On the surface side we've made a lot of progress in understanding the KZ structure, which is the main structure. It's - most of the mineralization is associated with in some form. And we've progressed our most advanced new discovery, the Kalimva-Ikamva project where we are now at a stage where we are moving towards drilling it after the reserve, and that's quite an exciting project, and then we continue to look at below the main KCD cluster of orebodies and searching for some deeper orebodies and we've certainly now delineated one new opportunity which I'll show you and I believe is in the next slide. I'll get to that. This is the Kalimva orebody, just to give you a sense of the work that the team has done. And again, this is now in the process of being drilled out to first resources and eventually we'll end up with a feasibility study. This is the work we're doing below the main KCD deposits. You can see at the end that purple line in the MZ cross-section. And then the red bottom MZ is the cross-section where the orebody comes to surface. And this - we're getting some pretty good intersections, but this part - this orebody from what we see in it could well have a refractory component to it because it's got quite a lot of arsenic in it, and we're waiting for enough drill sampling to be able to do some metallurgical cases. On the greenfield side, in the DRC, this is the Kilo joint venture work that we're doing. And again now, we've established camps in the northwest of the concession area and we're already into the physical sampling and follow-up. And as I've said before, this project needs - we need to find big deposits here because the access is pretty remote. Going back to West Africa and just to update as I promised. We were going to complete the initial or preliminary updated financial evaluation, economic assessment of the project to give you a scale on where we are. It's going very well, Massawa in the last quarter. And just to give you some color, we've summarized the base case which is basically the reserve base models using a brand-new plant, new design, fit-for-purpose processing plant and as well thermal-generated power, heavy fuel diesel, a bit of solar. The upside case is a case that John and his team have been working on which looks to grid power because the Senegalese are currently rolling out a big grid power infrastructure and we expect it to be up and running in the early part of the next decade. And we are assuming it will be late in 2022. It will be operational. So we've modeled two years of rental car and then moving to grid power. And at the same time, this also includes - upside case includes the utilization of refurbished front end of the Morila plant. And we've done a lot of work on that and we're pretty comfortable that, that plant we can use optimally rather than just making it fit. And with that, we take the IRR up to just short of 20%, but you can see a significant - in the robust business now. And we would - I would also point out, we're really focused on the risk, particularly on mining because with narrow orebodies, we've used 4.5 by 4.5 meters mining blocks to estimate and plan the mining. And we know we can do that because that's the sort of stuff we do in the narrow parts of the Tongon orebody. So there's a chance to improve on that, but we'd rather to do that once it's passed all our tests and we're in the mine and not make the decision on trusting that, that could happen. We are at a tender on the mining contracts and we are far advanced in the permitting. We are working with interministerial commission on managing all the permits that have to be done with - advanced the environmental baseline studies and the social baseline studies. We've started the infield social engagements, community engagements and so on. So everything is on track to be able to make the application for a mining permit at the end of this year. Where next? So we're still shy of the 3 million ounces, but we've made - as you know, things have changed quite a lot in the Massawa Main Zone or Central Zone, and we've now got $1,000 pits which supports the 2.6 million ounces. But there's very real continuity at depth, the challenge is how we sample it and get enough of a sample so we've been trialing some large-diameter RC drill holes. They seem to be working, we've - we're using very high compressors to make sure that we don't smear the samples and we've done some twin holes and the objective is to prove the concept of this continuation of the high grade and then we'll have to go back and then drill it out. But we've done a rough estimate that if you just take the Central Zone and you extend it at the same thickness and grade as it is in the main - in the $1,000 pit, we will extend the $1,000 pit down to add another 500,000 ounces. And on that basis, it takes us over the 3 million and it would also lift the IRR significantly above the 20%. And then we've still got some upside in the sense, KB is a very exciting exploration target. We're now pretty sure that it's - I mean it's a series of [indiscernible] orebodies, not linear. So more hard - more difficult to get to your model to work and so we're busy in a pattern drilling approach, trying to work out the geometry and get ourselves to a point where we can really model the potential of these orebodies. And then we've also - with the success of this now, we've gone back into some of the more refractory orebodies like Delya and its extensions, which is high grade. We're happy - we're going to be putting in a bioleach - a small bioleach processing facility with - into the life of the mine. And so it makes sense to be able to delineate the payable reserves that we'd be able to feed that process. Before I wrap up here, just a quick reminder of our commitment to stakeholder and community development programs. And in particular, on this partial list is the well-attended responsible citizen workshop we held at Kibali and collaboration with civil society. And really it just demonstrates how important our social license is. You've seen us work as we've managed our way through - and kept our relationship with the Malian government. The way we work with dealing with some of these very complex social situations that have been impacted by the Civil War in the North. And more recently, we're dealing with a country and a government there where the authorities sort of are already distracted by political issues in the DRC and it's important that we make sure that we - people are remembered and are sensitized to the civic duties and that we behave there as a member of the civil society. And then Randgold Resources is a long-term player, as I've always said, which aims to deliver sustainable profits and this is our revised 5-year forecast which forms part of our 10-year plan and takes into account the Tongon issue as well as the Massawa's potential contribution and subject to getting Tongon back on track as scheduled. As I pointed out earlier, our forecast for the year remains in the mid-range of our guidance. And I'll end as usual with a peer comparison and there's none more compelling than the dividend yields I showed you in the beginning of the presentation, amplified here by the implied operating cash flow payout ratio and the growing dividend history for Randgold. And the conclusion has to be that Randgold is and has been a leading performer in the global gold mining industry when it comes to delivering value for all its stakeholders. So with that quick summary, happy to take questions. We've got some of the team players around here. I got Simon Bottoms who runs our QRC on the mineral resource management; John Steele in the capital projects. Lois; and our CFO, Graham Shuttleworth; and myself. So we'll hand it back to questions.

[Operator Instructions]. Howie Flinker, Flinker and Co. Please go ahead.

What's your CapEx going to be this year? Is it going to be about $150 million?

So our guidance for the year is $225 million.

Oh I forgot. Okay. And in your cash flow statement, your noncash expenses differ from your D&A. What is the difference?

So it's a small difference there and it's to do with the accruals. So in the financial statements, if you look at the movement in your property, plant and equipment, obviously, we include certain accruals and for capital expenditure, whereas in the cash flow section, we show you just what we actually paid. So there's a small difference between there.

Okay. And finally, what's the status of your court case between you and Mali?

So we are going through what's called a mediation process where we've appointed a mediator jointly, and the object of the exercise is to get them to help us come to some sort of a meeting of the minds in terms of trying to find a solution to that rather than the next step, which would be to go into national arbitration. So we're working through that process and the objective is to have that done before the end of the year.

And is the governmental side willing to accept judgment of a mediator?

Yes. [indiscernible] there are, Howie, but we're fully committed by both sides. We have had very constructive engagements. As Graham says, we jointly appointed the mediator or facilitator. And the key thing is we need an independent person to point you - to point to the interpretation. And the big challenge in Mali has always been that the tax authority, the tax collectors are the judge and jury. And the ministers [indiscernible] the finance minister in a poor country always struggling to make ends meet and now he's got the United Nations on his back as well. It gives more capacity and it gives the minister comfort to at least deal with our issues.

Could this cost you more than $25 million or $50 million, if anything?

We don't think so because if we did, we would have provided for that, and we haven't. So we believe that we've got us a strong case, and yes, we believe that our financial position as it's represented in the financials is consistent with our - the expectations of the outcome.

But you believe now [indiscernible]

We've advanced paid some of the tax, Howie, which we feel is fair and proper and maybe - some points maybe even generous. But we've demonstrated our commitment to putting down what we think is fair.

Our next question comes from David Haughton, CIBC.

Maybe I can kick off, seeing that we're talking taxes just a moment ago, Graham. The Gounkoto tax break, when is that expected to start coming in?

So the period runs from the 1st of January 2018 until the 31st of December 2021. We only got the approval in July. So the impact of that adjustment will be made in quarter three and then going forward.

Okay. So we'll see an exceptionally low tax paid in Q3 and that will be the balance up for the first half of this year.

Yes. I mean when you say exceptionally low, I mean it's - it will be an adjustment of between $5 million and $10 million.

Okay. And the overall corporate tax rate looks like once it kicks in, it would go down towards the 20% level. Is that a reasonable assessment?

Well, so you'd have Loulo at 30%, Gounkoto at 15% Tongon at 25%. So it's going to be a blended version of that. So somewhere between 20% and 25%.

Okay. Now perhaps to, the Tongon strike, it keeps on flaring up. Is the resolution that you've got now a resolution that all parties are happy with and the workers are prepared to go back to work without any more difficulty?

We've been there before, David. The point is not as - is to deal with some of these - I don't know if I touched on it. I think I did last time. 30% of our workforce are ex Force Nouvelle soldiers, which we worked with the government on the demob strategy to reintroduce them back into civil society. And that's worked really well apart from a few that just haven't made it, and those are the people that seem to be effecting these stoppages and holding people to ransom. We have agreed and we worked through and we've got a protocol that everyone signed. And the government is in charge of this and there will be sanctions so - and there's quite a few criminal events now following the latest one. And I mean we have full support from the national union leadership. They've forced the resignation of the sort of working bodies and the union committee at Tongon. We've got the national civil society leadership behind us. The government in the form of ministries, all the relevant ministries, plus the elected member of Parliament who's the facilitator in the process leading up to this last stoppage as well as the governor and all the other local authorities. So this is not a social license issue. It's not a robot. It's an issue of people wanting more and more money and it all goes back to when the soldiers rebelled last year January and we've had this sort of dripping tap. And - I would point that one of the - the thing - probably the thing that we didn't do right was force this earlier because it's not in Randgold's character to initiate confrontation, and we would always believe it's better to have proper dialogue. And we did after the last one, the last stoppage, and again, it was a government-mediated process through the labor inspector and various elected officials plus the government. And we had - there was about 26 grievances. We ticked off 24 and we - to the - an extra production bonus that people wanted. It suddenly became a 7-month guaranteed extra pay, of which none of us agreed to, neither the government. Because remember it's a shareholder of the mine as well, and that created this second round of - or another round of work stoppage. And it's really barrier and intimidation rather than specific work stoppage. So we exercised which I should have done in April, a lockout under the labor code and with the full support of the government and also the community, the traditional leaders. And we are now busy reemploying people, calling them back to the mine and we'll - at the same time there's a process of - legal process going on led by the public prosecutor and his departments along with the whole string of bailiffs. And the first line of work is on site working as we start up the - and we're going to do it in the phased manner, we expect to get back to production in the middle of August. And there's a lot of government engagement at this stage and local authorities as we proceed with this restart. So a strike have - are rights to workers and we don't mind that and we deal with it. I mean we like to - prefer to find solutions before it gets to that point. But these sort of work stoppages have started to become concerning to us and the authorities because just now you're going to get violence and that's not good. So yes, I'm pretty comfortable that - I don't think anyone suspected - I always threatened and I think our threats became weaker because we ended up always trying to find a solution. And now that we've drawn the line in the sand, everyone understands the - that this is not - you've got to be able to - you got to be committed to work in the team and do a day's work to be able to get some pay. And just, by the way, for - out of interest, Tongon is right up at the top as far as pay in the country. We pay - our lowest paid workers are at 5.5x the base - the minimum salary, and that's before bonuses and other production incentives so - and we've already taken one full pay scale out of the bottom of our pay scales as we advance our workforce in the efficiency stakes. The mine, as you've seen in these quarter two results, very efficient, when it operates. We just have to deal with a couple of these out-of-town instigators. And again, we don't do those, the things we expect the government and the officers of the law to affect those sanctions.

Okay. So quite the good thing in Q3 but, off to a fresh start pretty much in a couple of weeks' time and onwards?

Yes. Next week or so, it should be up and running. We're guiding to 50,000 ounces for the year.

Yes. Just looking at Kibali. So it seems to be outperforming expectations. I mean as you had mentioned in your commentary, the mill throughput is higher than what, well frankly, I'd been expecting. And it looks as though the material handling in the underground is working according to plan. What kind of grade are you expecting through the underground? What did you get in Q2? And what's your expectation for the balance of the year?

So you're right. And the throughput right now remember, in our guidance, our 5-year plan, we had 7 30 this year and then 7 50, 7 50, 7 50, 7 50 [ph]. And what you're seeing is a run rate at around that plus or minus 20,000 ounces. And we did - we were cautious about the ramp-up in the first quarter of the mine, and we certainly are performing at above the 7.2 million tonne a year run rate on throughput. Personally, David, I am not comfortable at pushing this clock at the rate that we're doing at the moment for a long period of time because you'll definitely start seeing the maintenance cost going up. So it's a debate we're having in the team, and I would prefer if we're going to push harder in the throughput that we work to increase the underground feed and don't run down our open pit reserves out of schedule so just to give you that comment off the back. On the grade going forward, we expect the grade to be pretty much what we achieved this quarter with a run rate about the same, around 200,000 ounces for the next two quarters. Simon?

So the grade's going to step up but we're forecasting slightly lower throughput.

Throughput. When the grade steps up, we have to drop the throughput because it's harder material, and that's the way I want it to be.

Okay. And what grade was the underground for Q2?

I can tell you - let me get the [indiscernible]

No we haven't got - the split. 4.7.

Underground grade for quarter two, you're talking about, 4.7, and it goes up to 5 for the next two quarters, around there.

So we have to get to around about 5 to achieve that 200,000 ounces a quarter kind of thing? Then the last one, Massawa.

I've been talking about 2.6, 2.7 and underground's around 5.

Okay. And Massawa, the more you drill, the more interesting it looks. So in the Central Zone, you've got the commentary that as you go north, the grade gets lower. But my recollection is, that so as you go north, the ore is more refractory. Is that correct?

Yes. Exactly right. Exactly right. So that's interesting. And as we - and the southern - so the Central Zone proper has got much higher grades as we drill them out and - as we drilled it out and it's better metallurgy. And the reason is you got more and more gravity recoverable gold. Do you want to add anything, Simon?

Yes. I mean on the slide in the presentation, you can see that the impact of the change in recovery, because we modeled with each drill line and you can see the impact on the pit. So you can see the pit shallows and grades to the mill, and that's reflecting that modeling of that recovery into that pit optimization. So as well as just modeling the grade, we're actually modeling that recovery profile on a holding basis, right the way along the pit.

And at that - on that same diagram, I presume that we're looking at the same one, you've got the purple colored areas that are greater than 8 grams at depth. Does the pit get down on that? Or is that something that is an underground potential?

No. That's what we're highlighting as a potential that the pit will go down and the scenarios Mark was referring to earlier that we've run just looking at taking the grade tonnage profile from the areas we've drilled out within the current pit shell and just transplanting that into those areas highlighted with the potential. Then the pit would actually go down deeper and add an extra 0.5 million ounces within the pit.

Our next question comes from Luke Nelson JPMorgan.

Firstly on Kibali recoveries which were a standout for me. With the ore mix now moving towards steady state, it appears, is it now something - is that level of recovery something we should be thinking as indicative into H2 and beyond? And secondly, another one on Kibali. Obviously, starts to be a pretty big seasonality in terms of costs in the dry season. So just in the context, that Azambi is now operational and you're moving towards steady state production, can you give some sensitivity the cash cost to electricity, electricity rate, particularly if it sort of off splits around that targeted $0.10 per kilowatt hour mark? And then third and final question, just on Loulo. My understanding was Q1 there was some paced back flow issues which limited float access which was meant to reverse in Q2, which didn't look like it was the case. So any comment there would be appreciated.

So the last one first is at a depth because Loulo underground performed very well and it was - it made its target at 5 grams. We've still got some catchup to do but we're certainly top of the pace back fill. On the seasonality of Kibali, is there - and if I look at of our cash cost, just let me grab it - it's probably - the issue, it's probably, I want to say, 20% impact when you're at the peak of the rainy season but it's only - it's very small. It's short, the peak as you can see in that graph.

So Mark's right. It varies between sort of 10%, 20% of costs. It's hard to...

You definitely have costs in it, but the problem is we're marketing it by the grade improvement right now. So you won't see that. But Kibali is a low-cost producer and it's going to get better and better at it. But if you look at this last cycle and the - these rainy - it's not actually the rainy season that impacts this cost. It's the river run. It's the amount of water in the river. It depends on the catchment area, which is very broad. But if you look at it like this year, back in over the last 2 years, we had quite a long flat January, February, March, April. It's perfect for building Azambi. But this year it was sharp, as you've seen. But when you're up there with very high $0.25 per kilowatt hour, and that's changing now because Azambi brings slightly some higher background. Even in the low river levels it still contributes hydro more than it did last year and this year. So my - if this was all thermal, you'd be sitting at about 35% of your cost, power costs. And it's not, as you see, even in a high river period. So I would say it's 20% impact on costs when the river level is very low, the peak of these things, and when it's very high, it's almost nothing, because as you get down to less than $0.05 per kilowatt hour. So we've been playing around with this, Like, and seeing how do we schedule maintenance and can we exploit this more intelligently in our operations and we've got a little task team that should been working on this and see how we can maximize the benefit of this cyclicality in the power costs.

Okay. Now that's clear and then just on the part of my question on recoveries going forward in Kibali.

The recovery is just the impact of the underground ore component. It's a nice clean floating ore. We're getting 95%, 96% recovery in flotation and then subject that out to find grind. So what you're seeing is as we've got the underground ore percentage and the feed mix up, we're seeing the improved recoveries now up to this - almost the 90% level, but definitely the 89% level which we forecasted feasibility.

And again, Luke, I would just add to that, like we've done and demonstrated in Tongon and also in Loulo, Gounkoto, we're a geocentric organization. We pride ourselves in being able to be able to classify our ores. We've demonstrated to you time and again that we can manage these complex feeds. It takes a bit of time but we get on top of it. So on top of the fact that it's just driven by the feed components, we've also got better in the grade control and the processing team responding to those different reagent mixes. And so it's going to be somewhere between 86 and 90, 91. And so it all - it washes out at around the upper 80s, and that's what we planned in our feasibility. There will be times when we process some of the high sulfur arsenic ores out of the pits, which we know that we've classified those and we're going to explain them and manage them.

Our next question comes from Tanya Jakusconek, Scotiabank.

Just have two quick questions. Maybe just on - coming back on actually - Loulo Gounkoto, can we talk about how the grade changes for Q3 and Q4? Because I guess the improvement in grade is going to help the overall production improvement in the second half of the year.

Tanya, so the reserve grade for Gounkoto is 4.6 grams a tonne. And so when you look going out, we are getting out - we've been through a lower grade area. And when you look at Gounkoto for quarter two and quarter four, we are going to be sitting at - I mean sorry, quarter three and quarter four, we're going to be sitting at sort of around the reserve grade.

For quarter three.

And quarter four, we're going to be quite active, it's...

Quite a bit higher than that.

Much higher than that. Because we actually go into a very high grade zone of the orebody, which drives that turnaround. Simon, do you want to comment?

So as we're opening up the development towards the southern portion of Yalea, that gives us access to new very high grade areas in Yalea. And then the same environment in Yalea also giving us access, and we're actually retreat mining from that area. So we're locked in, in the sequence that we're mining there. But that sequence is key to those grade fluctuations in the plant.

And on the underground side, we'll be better than 4.1 because we have less feed from the [indiscernible] and more feed from the underground. But also some of the tonnage we'll start to move to Gounkoto, but not a lot. But when you go to quarter four, we'll just have underground again and that will be just over 5 grams. So that's what drives the recovery, Tanya. It's the extra orebodies.

Just trying to get - because 3.8 grams a tonne has been sort of your grade in Q1 and Q2 but for the complex. So I was just trying to understand where this grade is.

On the next feed, for the complex, you're looking at sort of 4.6 to 4.8 and around 5.5.

For Q4.

Maybe a bit higher.

But better recovery because of the better grade.

And better recovery because of the better grade, about 1% better.

And then my second question has just to do with Massawa and lots going on there. And now I understand you're doing some work on the metallurgy. I guess from my perspective is when will we have enough in place? And I know, Mark, you would say it takes as long as it takes. But when we will have enough time in place for us to actually know whether we go ahead or not? Do we need another year with all of these changes? Or are we still trying to target toward the end of the year?

We're targeting, making the application for the mining permit in January 2019.

[Operator Instructions]. We have no further questions. Dear speakers, back to you for the conclusion.

Thanks, everyone. And again, as usual, if you've got any follow-up questions, we're all available to take them. Otherwise, you can always work through Lois or Kathy du Plessis at DPA and set up a time for us to pick up with you or just send me an e-mail and we'll fit you in. Otherwise we'll see most of you in Denver. Thanks for your time.

This concludes today's conference call. Thank you for your participation. You may now disconnect.