Q2 2023 MP Materials Corp Earnings Call
Participants
James Henry Litinsky; Chairman, President & CEO; MP Materials Corp.
Martin Sheehan; Senior VP of IR; MP Materials Corp.
Michael Stuart Rosenthal; COO; MP Materials Corp.
Ryan S. Corbett; CFO; MP Materials Corp.
Carlos De Alba; Equity Analyst; Morgan Stanley, Research Division
David Adam Deckelbaum; MD & Senior Analyst; TD Cowen, Research Division
George Gianarikas; Analyst; Canaccord Genuity Corp., Research Division
Laurence Alexander; VP & Equity Research Analyst; Jefferies LLC, Research Division
Lawson Winder; VP & Research Analyst; BofA Securities, Research Division
Matt J. Summerville; MD & Senior Research Analyst; D.A. Davidson & Co., Research Division
Unidentified Analyst
Presentation
Operator
Good afternoon, ladies and gentlemen. Thank you for joining today's MP Materials Second Quarter 2023 Earnings Call and Webcast. My name is Tia, and I will be your moderator for today's call. (Operator Instructions)
I would now like to pass the conference over to your host, Head of Investor Relations, Martin Sheehan. Please proceed.
Martin Sheehan
Thank you, operator, and good afternoon, everyone. Welcome to MP Materials Second Quarter 2023 Earnings Conference Call. With me today from MP Materials are Jim Litinsky, Founder, Chairman and Chief Executive Officer; Michael Rosenthal, Founder and Chief Operating Officer; and Ryan Corbett, Chief Financial Officer.
As a reminder, today's discussion will contain forward-looking statements relating to future events and expectations that are subject to various assumptions and caveats. Factors that may cause the company's actual results to differ materially from these statements are included in today's presentation, earnings release and in our SEC filings.
In addition, we have included some non-GAAP financial measures in this presentation. Reconciliations to the most directly comparable GAAP financial measures can be found in today's earnings release and the appendix to today's slide presentation. Any reference in our discussion today to EBITDA means adjusted EBITDA.
Finally, the earnings release and slide presentation are available on our website.
With that, I'll turn the call over to Jim. Jim?
James Henry Litinsky
Thanks, Martin, and thank you all for joining us today. Let me give you a brief overview of today's call. To begin, I will discuss the second quarter and more recent highlights. Ryan will then run through the financials and KPIs, followed by Michael, who will review Stage II progress. I will then provide some closing commentary before Q&A.
So let's begin on Slide 4 and jump right to the middle of the slide under Stage II for important and exciting news. We have officially started production of separated rare earths at Mountain Pass. Michael and the team have done an incredible job on Stage II, particularly over the last several months to advance commissioning.
Let me be very clear, we still have a lot of work to do to get each circuit to operate reliably and ramp up production volumes, but the chemistry and separations processes are working as expected. We expect to begin shipping separated NdPr oxide to customers, this quarter.
So with the critical caveat that we must still work relentlessly towards our mission, I want to reiterate that achieving refined rare earth production is a milestone worth accelerating for all of us at MP, our shareholders a's well as everyone in America.
Moving back to the left of the slide. In the second quarter, we achieved record Stage I production for a quarter, which included a planned maintenance shutdown. In fact, this was our third highest quarterly production volume ever, with the only 2 greater quarters being ones where we did not have a 1-week plan shutdown. This translated into strong financial performance despite difficult pricing comparisons and Ryan will provide more detail on this in a moment. Nonetheless, the team's Stage I execution this quarter was very impressive, especially when we consider all the activity on our site related to Stage II commissioning and those achievements.
Lastly, our Stage III downstream magnetics business continues to advance across many functions. We are now installing some of the first metal and alloy making equipment in our Fort Worth facility. Other key equipment is in transit with the goal to begin installation during the remainder of the year. Fort Worth is going to be the epicenter of magnetics in the Western world. Along with the scale of the metal, alloy and magnet manufacturing, we are establishing a state-of-the-art R&D, piloting and product development area at the facility. Some of this equipment is already in the process of being installed, and we continue to make important hires across the magnetic team. In fact, MP, the company overall just surpassed the 600 employee mark, yet as we add depth and expertise across our sites in Mountain Pass, Dallas, Fort Worth and Las Vegas, we remain maniacally focused on preserving our owner-operator culture.
With that, let me turn it over to Ryan. Ryan?
Ryan S. Corbett
Thanks, Jim. Turning to Slide 6. As Jim mentioned, Stage I production continued at impressive rates as we produce 10,863 tons of REO in concentrate in the quarter. Modest improvements in uptime ore grade and feed rate combined to drive a 5% increase in production volumes compared to last year. The improved feed grade and feed rate, combined with strong recoveries to produce the second highest plant productivity, which we measure in tons of REO per hour of uptime in our history, demonstrating the steady progress in improving efficiency in the flotation process.
Moving to sales volumes. The higher production led to a 3% growth in sold volume reaching 10,271 metric tons in the quarter. This is despite an increasing amount of production being sent through the various Stage II circuits as we commission them.
As a reminder, we expect about 2 weeks of production or a bit under 2,000 metric tons of REO to be absorbed permanently into the downstream circuits as Stage II production ramps. And through June, we have seen about 1/3 of that total consumed. The last 2/3 will likely be consumed and therefore not available for sale, over the next 2 quarters.
Realized pricing in the quarter fell 55% year-over-year, driven by the decline in the price of NdPr, which represents the lion's share of the value of the REOs and concentrate. The decline was slightly better than we thought it would be when we last spoke in May, as NdPr pricing had a modest uptick for part of the quarter. Since then, pricing has again fallen into the low to mid-$60 per kilogram, and as such, right now, we are likely to see a low-teens percentage decline in sequential realized prices in Q3.
And lastly, on the far right, production costs climbed 11% over last year, in part from the growth in our head count as we commission Stage II and prepare it for full operations. And to a lesser extent, we continue to see slight year-over-year growth in materials and supplies cost. One area that has been a real positive is our combined heat and power plant, where we have been able to reduce outside vendor and third-party training costs as our own MP staff has gained clinical experience, driven plant efficiencies and operated with fewer bugs.
We expect these kinds of successes will come over time with other parts of our Stage II process and as we attain targeted reliability from the operating equipment. As a reminder, as we move to the back half of the year and begin preparing to migrate to Stage II production, some of these Stage I KPIs will sunset, as we evolve our reporting to focus on our transition to separated products and rare earths metal sales.
Moving to Slide 7. On the far left of the slide, revenue of $64 million declined by 55%, driven by the lower realized pricing we just discussed, as well as a small amount of rare earth fluoride sales in last year's second quarter. And given the nature of our commodity business, the $79.5 million drop in revenue drove a similar decline in adjusted EBITDA to $27 million. The third graphic on the page highlights that tightly managing our cost structure, combined with the low-cost nature of our scaled facility and world-class ore body allowed us to generate a very robust 42% adjusted EBITDA margin particularly when considering recent NdPr market pricing.
I would also point out that despite the very weak pricing in the quarter, Stage I continued to generate a modest amount of positive normalized free cash flow when removing our growth CapEx of $53 million in the quarter, and that was with a significant cash tax payment in April.
Moving to the far right of the slide, our adjusted diluted EPS declined to $0.09 in the quarter, primarily driven by the lower adjusted EBITDA.
But three quick comments on some below-the-line items, which had modest impacts on our adjusted EPS. With the current interest rate environment, we continue to put our large cash balance to work for us primarily in short term treasury securities generating over $10 million a quarter in interest income, recognized on the other income line of our income statement.
Second, with us putting over $300 million of assets into service over the last year, we are starting to see our quarterly depreciation expense tick up as we would expect. As such, we would expect a roughly 50% sequential increase in the depreciation, depletion and amortization line item in Q3 with some additional growth in Q4, as we also start putting some of our Stage III assets into service.
And while we are on the related topic of CapEx we continue to expect about $300 million of spend in 2023, having spent about $130 million through June.
And lastly, on the P&L, whereas total income tax expense in the quarter was down year-over-year due to the lower pretax income, our updated expectation for full year effective tax rate increased slightly, which caused an outsized impact on Q2's rate. With additional information and research on how we expect to implement the IRA bills 45x provision, we have upped our full year tax rate for GAAP purposes to the mid-20s.
To be clear, we have not updated our view on the cash benefit to the company, which we expect to be meaningful. But from a GAAP perspective, those benefits to the income statement may be spread over a longer period.
The updated full year tax forecast created a catch-up for the first half, which looked outsized in Q2 given the much lower sequential pretax income. Looking at the whole first half, our effective tax rate was roughly 23%.
Before turning it to Michael, I would also remind everyone that when we do begin shipping NdPr oxide, the actual sales cycle is going to be longer depending on where the product is delivered. For example, much of our initial production of NdPr oxide will be converted to metal by tollers in Southeast Asia before being delivered to our end customers. Shipping times alone to Southeast Asia will likely take 3 to 4 weeks, not to mention the needed inventory build at solars facility as well as production and delivery times. As such, we expect NdPr oxide revenue will likely be recognized predominantly in Q1 for the shipments that began starting this quarter.
With that, I'll turn it over to Michael. Michael?
Michael Stuart Rosenthal
Thanks, Ryan. Turning to Slide 8. Q2 was a very productive and busy quarter for the Mountain Pass operation. Concentrate production and product grade increased modestly year-over-year despite the semiannual plant shutdown. This is partially attributable to less unplanned downtime and is a solid accomplishment considering the very intense cross-training, personnel reallocation and resource sharing challenges presented by the Stage II commissioning.
The results also led to some extent, the fruits of continued optimization efforts in metallurgy and operational execution. As a reminder, we expect generally stable Stage I production year-over-year in the short term but hope to achieve periodic step changes in output tied to certain discrete projects in our pipeline.
Commissioning activity and breadth accelerated throughout the quarter and into Q3. All Stage II circuits are now in active operation. My previous comments remain valid. Commissioning is a process that demands patience and perseverance with very hard lot steps forward and frustrating setbacks. However, overall, I'm quite encouraged with the progress. Most importantly, we have so far avoided any major health and safety incidents that we cannot become complacent.
I'd like to thank our employees throughout the organization for their hard work, flexibility and resilience during this stressful time.
In Q2 and more so into Q3, we began bulk and individual separations and began product finishing.
In Q3, we expect to begin modest shipments of NdPr oxide along with lanthanum and cerium products and the production of SEG-Plus. We are pleased that the process chemistry and major equipment appear capable of operating in line with the Stage II design basis. We view this as a major success given the complexity of the process.
Our confidence extends across the hydrometallurgy, rare separation and product finishing circuits, but it is still early days and each piece of equipment and process experiences its own challenges along the path to stability.
We had the enormous advantage of inheriting an NdPr separation facility that has produced on spec materials in the past, and it was recommissioned in a manner that allows us to quickly produce on-spec material again. To avoid forfeiting this advantage that shape many months off the commissioning process, we continue to proceed prudently both upstream and in our separation survey.
In regards to the trajectory, Stage II operational performance, as with Stage I hinges on consistent and safe execution and how this manifests in uptime, mineral recovery and throughput. With the prior face of commissioning and commercial operations behind us, we're now working towards achieving incrementally higher uptimes in certain circuit, while maintaining modest throughput, identifying and resolving temporary bottlenecks before ramping throughput to our targeted run rate. As we do so, we are, as we fully anticipated, seeing inconsistencies in the performance of certain areas of the plant, more regularly than we expect in stabilized operations. While we are confident that the chemistry works, inconsistencies can come from a wide variety of sources, including new and legacy instruments and automation controls programming, secondary process equipment and material handling.
As I've said previously, we remain focused on the overall mission and will not sacrifice long-term success, sustainability and profitability of our operation for near-term announcements. But the bottom line is that we are overwhelmed by the commitment of our team and very pleased with our Stage II progress.
And with that, I'll turn it back over to Jim.
James Henry Litinsky
Thanks, Michael. Before we go to Q&A, I wanted to provide some updated perspective on our industry and the general supply-demand environment, both present and future.
NdPR pricing remained weak throughout the quarter, trading in an approximate range in the low to mid-60s per kilogram. This was mainly a continuation of the conditions we discussed in May. We see the electric vehicle and e-mobility categories experiencing strong growth. We expect those to continue to grow at least 20% to 30% per year for quite some time. However, industrial applications, wind turbines, hard disk drives, speakers for PCs and smartphones and other consumer electronics applications have been weak throughout 2023.
Certainly, the economic conditions in China are hitting those categories significantly. As a reminder though, the split between electric transport and then all the other industrial and consumer options in our industry is currently approximately 25% to 75%, respectively. This means that in the near term, a 5% hiccup in those traditional spaces can effectively wipe out the accelerating demand in EVs. Of course, this is a short-term phenomenon because the compounding effect of growth verticals begins to trump the influence of legacy industries within a couple of years. Then, of course, the volatile nature here works both ways.
On the supply side, we think 2023 has been impacted by the reopening of the border between Myanmar and China. We understand that there was a fair amount of pent-up inventory from Myanmar mining production that entered China over the first 6 months of the year. I would encourage anyone interested to Google for stories about toxic rare earth mining and human rights abuses in Myanmar. This supply obviously interferes with normal market forces and more importantly, of course, American National Security and humanity. We expect though that this relative impact to supply is transitory.
So net-net, as I have said many times before, anything can happen in commodities prices in the short term. With that said, we view the recent soft pricing environment, as explainable by identifiable market factors on both the demand and the supply side that fortunately seem to be mostly behind us. Therefore, from our vantage point, the long-term prospects for our industry and the value of our platform remain extraordinary.
On that note, we thought we would be doing shareholders and the clients a disservice, if we did not at least mention the word artificial intelligence on today's call.
So let's go ahead and move to Slide 9. All these pictures were created using the Bing Image Creator, which is powered by Dall-E. We had a little competition internally here at MP to see who could do the best prompt to generate images that outline how exciting MP's opportunity is in this new AI future.
The picture on the left is what rare earth mining could look like in the year 2040. This one is somewhat fantastical, but I think now we have some better ideas for what Michael's next uniform might look like.
The second image is what is supposed to be the Chicago skyline a couple of decades from now, where EVs, autonomous vehicles, EV Tal and drones are dominant modes of electric transport or motion.
I think, we would all agree that this is a reasonable depiction of what might be a reality in the not so distant future. This is a world where the demand for rare magnetics is many multiples of what we see today.
The third slide is where things get interesting. Here, we have both quadrupedal and bipedal robots. The more colloquial terms you might have otherwise heard for these are robot dogs and humanoid robots.
At first launch, even as recently as earlier this year before the launch of ChatGPT, these objects might have seemed like totally distant fantasies, but we are already seeing companies apply large language models to robots like these. Just check out YouTube, where you can see a number of videos where various companies are showing off robots, just like these that are able to be directed with spoken language.
Robots use small motors called actuators in their joints to enable movement and provide balance, strength and dexterity. Magnets, primarily rare earth magnets, are what power that. In simple terms, think of an EV as essentially a robot on wheels. It has a large battery and in most cases, a rare magnet in its motor transforming the stored energy into motion. A robot has a small battery and then lots of actuators around this structure. When size, weight and performance matter, rare earth magnets win. For many robotics use cases, a rare earth permanent magnet is likely going to be the only solution for certain actuators.
How this will all play out and over what time frame? It's anyone's guess. For a rough order of magnitude, though, today's quadrupedal and bipedal robots might use anywhere between 0.2 and 3 kilograms of NdPr, a robot doing a major strength task or one needing balance for some kind of security or military case is likely going to need even more NdPr per unit than 3 kilograms. This compares to about 0.5 to 1.5 kilograms in a typical EV.
Roughly 85 million cars are produced per year, and there are about 1.4 billion cars on the road. Earlier this year, Elon Musk predicted that AI robots would eventually outnumber humans. If he turns out to be correct, it would suggest that the size of the opportunity for MP will be many multiples that of EVs. Moreover, in EV applications, rare earths have maintained a 90-plus percent share. Even as there is an understandable conversation around the trade-offs of significantly better performance with rare earth versus the cost and/or supply chain availability and other risks, in robotics, given the space limitations and performance needs, it is very likely that there will be a significant percentage of use cases where substitution is just not feasible at any price.
I would also add that it is also less likely that the Chinese are going to want to provide magnets to American companies in an area with such direct military application.
So to conclude, AI excitement is playing out right now in the public markets via the chip companies, mega cap tech and a handful of other companies but a profound change from AI over time is likely going to come from the next layer of innovation on top of the infrastructure layer.
Robotics is going to be an important disruptive vertical, where we believe AI is pulling forward into the next few years, we might have taken decades. I do not know when, but at some point, investors are going to start to price that into platforms like ours, too.
With that, let's open it up to questions. Operator?
Question and Answer Session
Operator
(Operator Instructions) The first question comes from the line of Matt Summerville with D.A. Davidson.
Matt J. Summerville
Two questions, I think. First, Jim, the comment you made around Myanmar, can you talk about whether you think that supply spigot, if you will, inbound into China has normalized at this point? Or do you still feel like there's still some excess inventory to be brought over the border, so to speak? Well, I guess that's literal.
James Henry Litinsky
Sure. Matt, thanks for the question. So Myanmar is tough to really know. But what I would say is what we -- it was -- that border was shut off for the most part, and there was a lot of inventory that built. And from everything that we have heard that was sort of a huge rush of supply that came through kind of throughout this year, so we think it's sort of a onetime impact of excess supply. And our -- we believe there's been a recent crack down on that.
And so, I guess that's a long-winded way of saying our belief is that, that's pretty transitory. But as an impact, obviously, that supply exists in some form, but that's what we see from this standpoint.
Matt J. Summerville
Got it. And then, just as a follow-up. I was wondering if you could comment on now that you've produced on-spec material, what kind of first pass yield you're seeing out of the separation and finishing operations and how that compares to where you thought those first pass yields would be?
Michael Stuart Rosenthal
Thanks for the question. This is Michael. I guess, there are so many parts of the process to hard to give you a single answer. We don't expect our initial production yields to -- from concentrate to finished product to be at our desired long-term yields we expect there will be significant improvement. But I guess, what I would say is that we started up our separation process with equipment that had produced on spec material in the past and was restarted in a manner to quickly produce on spec, and we're quite pleased with how the separation circuits are working, and similarly optimistic about how the finishing processes will work. I'll probably leave it at that lets you follow up.
Matt J. Summerville
No, we can leave it there. Thanks, Michael.
James Henry Litinsky
Thanks, Matt.
Operator
The next question comes from the line of Carlos De Alba with Morgan Stanley.
Carlos De Alba
So Michael, I just wanted to follow up on the discussion on Stage II. And so, is it fair to say then based on the comments that you had earlier that all critical steps, all critical processes have been running and are running, and it's now just a matter of fine-tuning and making sure that you can run all the circuits and all those processes in a consistent manner and then just improving, I guess, of time and your metrics like that, but the critical steps are already all running and you feel comfortable with that?
Michael Stuart Rosenthal
I think that's a pretty fair statement. Certainly, the product finishing is last in line, and we can't put the proxy materials into parts of those circuits to run them, so they've had less time, I would say, slightly less -- yes, certainly, there's a little bit more work to do before we can -- just to say that all of the bugs are worked out or were 100% confident, but I would say that's not chemistry there. That's just kind of physical material processing. But otherwise, I think your statement is correct. The general processes are working. We feel comfortable that what we've designed is able to perform to the design basis of our circuits and that with time, we'll get to greater throughput, reliability and performance yield.
Carlos De Alba
All right. That's quite encouraging. And then would you -- how do you feel that you are measuring against the timetable that you had plan or defined for the ramp-up of Stage II? Are you on time, you think you are a little bit ahead, or are you a little bit behind schedule? How would you characterize that?
Michael Stuart Rosenthal
Yes, I think we've said all along that the process will be nonlinear, so it's sort of hard to kind of benchmark against the unexpected. We knew there'd be a lot of unexpected challenges, and we continue to see that play out. Certainly, there are many challenges we anticipated and then challenges in places that we thought would be a little bit easier, overall, really proud of the team and how they're working through those challenges and feel comfortable that the targets, we set out remain our targets.
Carlos De Alba
Right. No, fair enough. And maybe last question for Ryan. What could comment on the cash flow generation in the quarter? I think, I calculate -- we calculate something around $10 million in 2Q. That was below our expectations, both consensus and our own despite the fact that you guys bid on EBITDA, right, both also the consensus and our number. So some color there would be great.
Ryan S. Corbett
Sure. Carlos, I think probably the missing piece on consensus versus what we actually achieved was likely a fairly significant mid-$20 million cash tax payment relating to prior year taxes. And so obviously, as you know, often a mismatch between P&L timing and cash flow timing on that cash tax payment. So that is the vast majority of the delta. Obviously, if you normalize for that, we would have probably significantly beat consensus on those numbers.
Operator
The next question comes from the line of George Gianarikas with Canaccord Genuity.
George Gianarikas
I just like maybe a little bit of an update, as you now start to ramp and the NdPr production to what you think your production costs per kilogram will be going forward?
Ryan S. Corbett
Sure. George, we'll stick with George G. Well, I can take that. It's Ryan. I think what I'd say on production cost is obviously it's early. I think we've been very clear that the early production is going to come at a different cost structure than the run rate production, of course, as you'd expect.
We are obviously very proud of our position on the cost curve that's well demonstrated in our Stage I business. We continue to have significant confidence in maintaining that lead in that position on the cost curve, as we transition to Stage II separated product production. We gave a little color to investors and analysts when we originally went public on where we think roughly cost per kilogram would be for NdPr. There are a lot of things that since then have changed, namely commodity reagents, but there's also a lot that has stayed the same. And so, this is something that I think other than major commodities and inflation, as Michael laid out, with the chemistry working the way that we expect, that is the fundamental driver of the economic framework that will be made out to all of you. And so, we don't expect that to change, which is super encouraging.
George Gianarikas
Great. And maybe as a follow-up, last quarter, Jim, you made a statement that you think that at $60 a kilogram that China Inc. is likely unprofitable. Is that still your view? And do you suspect that in and around this price range, we should see some stabilization?
Martin Sheehan
George, thanks. Yes, prices in May were roughly around where they are today. So that statement is still true. I mean we -- it's always very difficult to read the tea leaves in China. And if you recall, and I said something like this last quarter, but it's difficult because there's always the ability to sort of shift various things across various entities, as they do in the supply chain or in the stream, if you will.
So -- but from everything that we see, that remains the case. And so it is encouraging. And as I said in the prepared remarks, we think that those -- these items are behind somewhat behind us. And so pricing feels more stable here. And I would argue that for all the reasons I've stated that, that downside volatility that we saw, there's always not if or when likelihood that there's that upside volatility the other way, right? It gets reflective. But we shall see.
Operator
The next question comes from the line of David Deckelbaum with TD Cowen.
David Adam Deckelbaum
Jim, I'm tempted to ask you about the time line for getting [Deathpunk] to work at Mountain Pass. But I guess, I'll hold off on that one. But the -- I am curious as you think about, I thought the idea was in the second quarter that you all would be withholding some inventory or perhaps a greater amount of inventory to be used in the Stage II testing, I just wanted to confirm whether that had happened or not. And if, we should still be anticipating that you're effectively still selling a significant amount of concentrate in the back half of this year.
Ryan S. Corbett
Yes, David, it's Ryan. I can take that. What I've laid out in the prepared remarks was absolutely a reiteration of that expectation that there's a pretty significant amount of inventory that gets consumed into the (inaudible) as we commissioned. We had talked about roughly 2 weeks worth or, let's call it, just shy of 2,000 tons of REO.
We, as of June 30, had consumed about 1/3 of that. And so, it's safe to say that the remainder of that will get consumed as we ramp and hopefully as quickly as possible. And so, I would absolutely expect as you model out Q3 and even Q4 to think about not just the impact of the rest of that inventory getting consumed, as effectively line fill into the downstream circuits. But in addition, I talked a little bit, in remarks about the sales cycle as we transition to separated products.
And so with that, I just want to reiterate, obviously, the timing there as we start to consume that REO in concentrate into the downstream circuits, converted into separated product and then either sell it on or bring it to tollers to be -- to turn into metal and sell to our ultimate customers. There's also the timing impact of that longer sales cycle.
And certainly, as you'd expect, as we ramp our separated product production, and as we sort of continue to fill that channel, particularly the metal production channel, this will get lapped and will be a onetime impact, but it is important to think about over the next few quarters the impact that, that will have on the financial model.
David Adam Deckelbaum
Ryan, that's a nice segue into my follow-up, which is just for my own edification. The sales cycle is expected to be 6 months for getting the NdPr metal because I guess when you sell concentrate, [Chancas] taking delivery of that cargo at the port, so you can recognize revenue as soon as that we support versus when you're using a tolling arrangement? You don't recognize revenue until the product that's still, I guess, technically your resource until it ends up in the hands of the customer?
Ryan S. Corbett
Sure. It's a good question. I'd say 6 months is probably not the right answer there. Certainly, it's significantly longer than our current sales cycle for concentrate.
To clarify the prepared remarks that I think you're keying in on, what we had talked about is separated NdPr production NdPr oxide production and shipments that are happening later in this quarter. So later in Q3, likely much of that revenue could be Q1 revenue. So end of Q3, beginning of Q1 is not quite 6 months. Shorter than that for sure. However, longer than exactly what you laid out, that our sales cycle is right now, that we deliver product to the port at Long Beach and title transfers there. That's when our revenue recognition happens for concentrate today.
I flag though that not all separated products will be subject to the longer sales cycle. Part of our products will remain our products as they get told into metal and then the revenue recognition is likely at the port in Southeast Asia when it's delivered to customers, to be distributed to broader Asia into Japan primarily. There will be instances where we sell oxide at the Port of Long Beach as well. And so, there really will be a mix.
What I mentioned, and I think it's important to keep in mind, just for the next quarter or 2 is that our early NdPr oxide shipments likely will be told into metal. And so, these early shipments sort of to help you guys think about your models, will likely be subject to that longer sales cycle. But as we carry forward and sort of start to make more product and broaden the customer base and things like that, there will be a mix.
Operator
The next question comes from the line of Lawson Winder with Bank of America.
Lawson Winder
Nice quarter and great update. Okay. So I wanted to try and put a bit of a finer point on the ramp up. Should we still be modeling a run rate of 6,000 tons of NdPr in for full year 2024?
Ryan S. Corbett
I'm happy to start. It's Ryan, and then Michael you anything else to talk about on the commissioning side, feel free to jump in. But, what I would say is, frankly, a reiteration of Michael's commentary that based on what we've seen run rate production at the end of the year remains our goal. It's nonlinear. It's difficult to perfectly forecast, but that very much remains our goal there are certain areas that I'm sure Mike can better expand on that will continue to tweak to improve throughput and all that versus what we're seeing now.
As you think about '24 though, I'd say it's kind of early to predict exactly what our output and importantly, sales volumes will be for 2024, at this point. We've got a lot of work to do, as I mentioned, in the coming months to finish commissioning and to ramp production. And so I'd say stay tuned. We'll continue to update you on the progress, and we'll let you on calls going forward. But, as we always do, we will formulate our 2024 or our next year production and sales plan based on customer demand and on market conditions, once we get closer to that time period.
Lawson Winder
You actually -- you touched on a really interesting point there, and that is formulating your sales based on market conditions. So, I mean one of your competitors made some comments along those lines recently. Is that, is that how you kind of think about the market? I mean if the market is not willing to take the material does MP just hold it back?
Ryan S. Corbett
It's a great question. We -- I think the great thing about what we've built here is a platform where we have product flexibility, right? We have the ability to produce Stage I product, Stage II products. We're making great progress on Stage III. We have that as a lever to pull as well. And so, we always as a management team are focused on maximizing shareholder value. And so, if that means changing the product mix for a period of time, so be it.
James Henry Litinsky
But to be clear, we have not seen any pushback or any lack of demand. We have lots of conversations and lots of demand. And the goal now is to thoughtfully build out our sales channel and think about the full downstream of our business.
Lawson Winder
Okay. Yes, that's great color. And then maybe just finally, I wanted to kind of understand the heavy rare earth separation circuit, is that going to ramp up right on the back of this? Could we be modeling some costs for that in 2024? Or is this something that's going to happen a little later on, like maybe '25 or '26. And if it is happening next year, how should we kind of think about that in terms of like impact on NdPr?
James Henry Litinsky
Yes. Well, no, by the way, I'm glad you raised that since you mentioned one of our peers, I'm guessing you're wondering about heavies. And there's probably just for understanding the overall project, maybe just a little important background for those who don't know.
But early in '22, both us and Lynas were given awards by the Department of Defense. Actually, our award was announced by the President himself to us. And so, that was exciting. We talked about that. And then there was an announcement of something with our peer a couple of days ago. But what I would tell you on that is I think it's pretty clear that DoD understand the importance of a level playing field and good competition across the spectrum here. And we are fiduciaries, and you can imagine that given how much has changed since early last year in the world that we would certainly expect to proceed with projects that we believe are economically attractive, and that we believe are at a -- we believe we have a good competitive position.
So, I think what I'm trying to get at in being is sort of direct as I can, is that you should expect, if you kind of heard everything I just said that we -- we are in discussions, and we believe that we will have equal support or similar support for our mission with respect to heavy. So I think, that that's a very positive thing that we recently saw.
And so, I think I would leave it at that as far as the heavy project is related, just to say that, again, you can probably deduce from what I'm saying that I think that we will continue to expect some good things on that front.
And then lastly, I think you asked about timing. The heaviest project is to be online for the magnetics business, and that timing remains on track.
Operator
The next question comes from the line of Laurence Alexander with Jefferies.
Laurence Alexander
I guess, first of all, could you just give a quick update on your thoughts around CapEx for this year and next year, just given kind of the milestones you've hit and kind of the inflationary pressures we're hearing about from other companies.
Ryan S. Corbett
Yes. Sure. Laurence, it's Ryan. We mentioned earlier on in the call that our original CapEx guidance for the year of $300 million remains intact. And we tend not to give CapEx guidance further out than that. But, what we have laid out, obviously, a couple of years ago in laying out sort of the broader capital plan across Stage II and Stage III was that $700 million investment. And that also remains on track.
Certainly, as you know, the Stage II spend is essentially behind us, except for some small items from a timing perspective. And so, there are other projects that we're continually investing in and out in past, in addition, of course, to our Stage III business. And so, we've done about $130 million year-to-date in CapEx. We continue to see that $300 million is a fair number to model for the rest of the year.
Laurence Alexander
And then, just to clarify two points you made in the earlier in the Q&A. With respect to the kind of 6,000 production run rate, given the lag you were talking about in terms of flowing through the P&L, were you trying to get at that the comparable run rate for 2024 would be 75%, 80% of that?
And then, I guess the other issue is with respect to the prices the ASPs on Stage II that you will flow through the P&L., can we assume that you are using -- that you'll be able to realize market prices? Or will there be some level of discount either through due to partnership arrangements or because you're kind of building relationships with customers who you expect to translate into magnet customers? Can you just give a sense for what, how closely we should hold you to whatever kind of public market indexes we use for Stage II?
Ryan S. Corbett
Yes, I can take both of those. As it relates to the timing and the sales cycle, I don't think we were trying to make any specific comment as to 2024. I think that -- we're actually trying to avoid that since we don't give guidance. But what I'd say on that is if you have an assumption and our goal has remained to get to run rate production by the end of 2023. We've also tried to consistently flag, of course, that the overlap from a timing perspective of production and sales that we experienced now is not the same correlation in production and sales that we'll experience, once we're producing separated products. And so, I think you're on the right track. But certainly, if you were modeling 100% run rate starting 12/31, you would not have 6,000 tons of NdPr sold necessarily recognized, as revenue in 2024. I think that's a fair statement. Remind me your second point, second question.
Laurence Alexander
Just how closely should your pricing going before to ASP at the time.
Ryan S. Corbett
Yes. On pricing, look, we will have a variety of different contract structures that are based on the indices that we all look at. I think the thing that will be interesting there as well is a difference potentially in timing and then, a variety of different contract structures as it relates to realized price versus market price. So I think, as we get closer to more significant sales volume, we'll continue to keep you updated there, but there's not going to be an easy pinpoint for me to be able to give you at this point to say, hey, take market and that was trapped. But we'll do our best to keep you updated from a modeling perspective as we get closer.
Laurence Alexander
And just, I guess, with respect to kind of one other question earlier, just to help us understand the gains here, if you were to decide that you were to produce the rare NdPr, but then keep it off the market, what's the shelf life of the product? How long could you hold it off the market and still monetize it at the same economics?
James Henry Litinsky
Mike, well, just to be clear, we don't have any expectation of keeping product off the market at this time. So I don't know, is that...
Laurence Alexander
It's (inaudible) yes.
Michael Stuart Rosenthal
Yes. I guess in theoretical terms, I don't believe there should be a shelf life for oxide metal or alloy may have considerations of oxidizing.
Operator
The next question comes from the line of Ravi Sinha with (inaudible) Klein Capital.
Unidentified Analyst
Just quickly on your production facility in Texas. So could you confirm, if your magnet facility is covered under 48C, and if you get ITCs that we can count on?
Ryan S. Corbett
Sure. We had a little to hearing you, but I think the question was about 48C. Is that right?
Unidentified Analyst
Right. If you magnet facility investment is covered under the 48C.
Ryan S. Corbett
Right. On that point, since we spoke about this last, there have been some further updates and clarifications regarding 48C, and how it relates to our Stage III business and our DFW facility. I think, the important update there was from treasury and DOE that magnetic was specifically called out as a priority, that said, it's definitely too early to know if we will directly benefit from this program at this specific facility. And as you can expect for competitive reasons, we're not going to discuss specific support we may or may not be seeking from the federal government in a competitive process. But you could certainly expect that we're always monitoring these sorts of opportunities and will avail ourselves of any opportunities that arise.
Operator
The final question comes from Carlos De Alba with Morgan Stanley.
Carlos De Alba
Yes. Just wanted to see if you have any color on start-up cost. I know those are normalized out EBITDA, but just how do you see the progression? They came down quarter-on-quarter. Is this a trend that we should expect to continue or not really?
James Henry Litinsky
Carlos and Bob, I thought you were coming back on to congratulate us on our big milestone this quarter. But I guess...
Carlos De Alba
That goes without saying. That goes without saying.
James Henry Litinsky
Go ahead, Ryan.
Ryan S. Corbett
No, it's a fair question. I'd say the reason you probably saw it come down a little bit sequentially, is that we started to enter into more normalized operations on various circuits, as we bring them online and process order. So as we've talked about consistently, we've been bringing each step of the process online, testing it, getting it up to sort of what we would consider commercial operations. And so, since we had been running some of the most upstream parts of the circuits for a relatively significant period of time since we got into June, part of those costs kind of fell out of the start-up category.
I think that, that phenomenon will continue, but we've rapidly added significantly more circuits into ongoing operation very recently. And so by their very nature, sort of start-up and nonrecurring costs are very difficult to model and predict. But I would say a fair way to estimate is, I think our Q2 numbers are probably fair to think about for Q3 and Q4 and then, obviously, would tail off very meaningfully.
Operator
There are no additional questions left at this time. I will now hand it back to Jim Litinsky. Please proceed.
James Henry Litinsky
Thanks. I just wanted to say this was a an exciting milestone for us. And so we're -- I'm really proud of the team for the hard work, and we'll get back to work over here, and we look forward to talking to you all next quarter. Thanks, everyone.
Operator
That concludes today's conference call. Thank you. You may now disconnect your lines.
