Good morning, everyone, thank you for joining us today for our final chapter in our four-part series. Each chapter in this series was created to provide you with deeper insight into Mosaic's strategic priorities. Today, Clint and I will bring things together as we look at the positive impact of these priorities and their outcomes on the business. Before we begin, however, please review the safe harbor and familiarize yourself with our conference capabilities, especially the Q&A box on your screen. You may enter questions at any time, and we will address them at the end of the presentation. Mosaic has six global strategic priorities that align with our actions, focus our employees, provide a framework for decision-making, and hold us accountable for creating long-term value for our stakeholders.
This approach to managing the business has strengthened our foundation and improved our competitiveness during the recent ag and global uncertainties. At the same time, this approach has also positioned us to accelerate our financial performance as markets improve, which I'm very pleased to say is exactly what's happening today. Our highly successful transformation efforts, which continue to be executed under our strategic framework, produced exceptional results in 2020, driving $318 million of benefit. What's important to remember here is that this performance was driven by the initiatives and actions that we control, as full-year 2020 prices were actually below what we saw in 2019. In addition, as I'm sure you all saw this morning, we have increased our target dividend rate by 50%. This reflects the success of our strategy and the confidence we have in Mosaic's ability to deliver non-market EBITDA growth as we move forward.
Before we turn things over to Clint, I'd like to provide you with a quick update on some of the strategic priorities we covered in previous chapters. In September, Bruce Bodine and Karen Swager provided information about our North American transformation and our drive functional collaboration and efficiency strategic priorities. Since then, we have reached important milestones that reflect the powerful combination of initiative thinking and enabling technology on the future of our business. As you heard in our year-end earnings report, the North American business is delivering production and cost improvements that were almost unthinkable a decade ago. Looking ahead, we expect this progress to continue as we advance our transformational initiatives.
Two significant highlights are the new integrated operations center in phosphates, which is efficiently and safely running our Four Corners field operations today, and our continued accelerated progress at Esterhazy as we near completion by mid-year of this year. This would represent the delivery of this project over two years earlier than originally planned while remaining within our original budget, an extraordinary accomplishment for a project of this size and complexity. In November, Corrine Ricard joined us to discuss our South American growth engine strategic priority, which is very well-named given the significant contribution of Mosaic Fertilizantes. Record-setting financial results continue, with 2020 delivering the best annual earnings and adjusted EBITDA since the acquisition from Vale, despite year-over-year declines in our average MAP selling price.
Mosaic Fertilizantes more than doubled its 2020 transformational performance target of $50 million that we set earlier in the year, and the team is ahead of schedule to achieve their 2023 EBITDA goals. Brazil remains one of the largest, fastest-growing agricultural markets in the world, providing the economic basis for our growth strategy. We expect continued growth on many fronts as we look to the future with the combination of stronger grower economics, efficient assets, and our well-positioned distribution business, as well as the benefit of solid execution. In December, we heard from Walt Precourt and Ben Pratt as we look in-depth at our grow and strengthen our product portfolio and our act responsibility priorities.
Since that time, we have made progress on both these strategic areas, including external recognition for our efforts with the prestigious Campbell Award for EHS excellence and our inclusion in the Barron's 100 Most Sustainable Companies list. Here, I'm very pleased to share that Mosaic is the first-ever company in the agriculture or fertilizer space to be named to that list. We've also built promising new relationships in the soil health space that position Mosaic to benefit in this fast-growing market. As we continue to progress our ESG journey with recent enhanced fertilizers designations for two of our performance products. While we explore new paths for high-value product growth, it is also important to recognize that our suite of performance products continues to build momentum with our customers.
We set new global performance production records of 4 million tons in 2020. We expect to achieve 30% growth in these higher-margin products by the end of year 2023. Now, I'm going to turn it over to Clint to review the financial implications of our prior presentations, including implications for free cash flow and how we think about capital allocation. Clint?
Thank you, Joc. Good morning everyone? In our previous chapters, we identified the adjusted EBITDA growth relative to the 2019 baseline that we expect to realize from our transformational programs, which include the leveraging of technology through our NextGen initiatives and the anticipated completion of our K3 project, all things that we control. As of the end of 2020, we remain on target, in fact, a bit ahead of schedule, in realizing those benefits. While our EBITDA and cash flow profile is improving through our transformation initiatives, we expect our total capital spending to decline from here based on what we see today. We've guided to approximately $1.1 billion in CapEx this year, which includes approximately $190 million for K3 and sustaining CapEx at the top of our $650 million-$750 million range, given the timing of certain work, particularly in phosphates, that needs to happen this year.
With K3 spending reaching completion and sustaining CapEx reverting to more normal levels, we expect total capital expenditures to decline to around $800 million in 2025, based on our current commitments and expectations. Now, this won't be quite a straight line decline. Recall that we've allocated roughly $200 million in 2023 to extend the reserve life of our South Fort Meade mine, which will allow us to extend the life of our current beneficiation assets, which then meaningfully delays much larger investments for new mines in the future. We also expect asset retirement obligations spending of approximately $175 million in 2021, partially driven by the closing of Plant City, to decline to about $60 million over the same time period.
Adding the decline of both capital and ARO spending to the non-market adjusted EBITDA growth driven by our initiative, results in more than a threefold increase in expected free cash flow in the 2020 to 2023 period. Again, all based on things that we control. This is obviously before considering any impact of strengthening global fertilizer markets. Now, beyond the EBITDA growth that we're driving through transformation, we've seen significant increases in phosphate prices and margins, benefits from the Brazilian real depreciation, and recently, more modest increases in potash prices. The EBITDA sensitivities shown on this slide are meant to help you model future EBITDA based on prior performance and your estimates of market drivers. These sensitivities were updated in our fourth quarter earnings materials and show that for every $10 increase in DAP price, holding raw material prices constant, full year EBITDA increases by $105 million.
A $10 increase in MOP implies a $65 million increase in full year EBITDA, net of Canadian resource taxes, and a 0.1 decline in the BRL to USD ratio or a 0.01 decline in the CAD to USD, both add approximately $13 million on a full year unhedged basis. For phosphates, one of the complexities of the analysis is raw material cost. It is important to remember that in our phosphate segment, we use 0.18 tons of ammonia for every ton of phosphate that we produce. And we buy only about one third of what we consume on the spot market. We use about 0.4 tons of sulfur for every ton of finished product in the phosphate segment. The final consideration involves identifying the pricing to use as a base for these sensitivities, particularly if you are applying these to quarterly analyses.
There is a lag or a time between the publishing of various spot prices and when they are actually realized in our P&L. For phosphates, this lag can be up to 60 days. The prices flowing through our income statement could be the prices that you saw in the market two months ago. For potash, this lag can be as high as 50 days. More detail on the timing lags of finished products and raw materials is covered in our modeling education deck on our website. These sensitivities should help you estimate what, if any, adjustments you should make to the non-market EBITDA and free cash flow figures that we just walked through a minute ago. Currency and inflation are also an important consideration when assessing Mosaic's free cash flow generation capability. We have material non-U.S. dollar expenses in both Canada and Brazil.
In both countries, our revenues are pegged to the U.S. dollar. In Canada, most operating expenses are in CAD, which drives our unhedged sensitivity. A 0.01 change in the CAD to U.S. dollar ratio is a $13 million impact to full year adjusted EBITDA. In Canada, our hedges, though, tend to be related to our capital expenditure programs, particularly K3, less so for normal operating costs. In Brazil, we hedge net cash flows. It's an economic hedge, not an accounting hedge, to help smooth out near-term currency volatility. Over time, though, our real base cost will reflect the exchange rate changes. A significant portion of our costs are priced in U.S. dollars, particularly raw material cost and product purchases for our distribution business, we do have approximately BRL 4 billion in fixed real-based cost in our Fertilizantes business.
Over time, a depreciating Brazilian currency lowers real-based expenses in U.S. dollar terms, a benefit, given sales are pegged to the U.S. dollar. There is a close relationship over time between currency and inflation. While a depreciating currency is a benefit to our profitability, real base cost inflation is a negative. The chart on the right of this slide shows the correlation between the currency depreciation and inflation. While currency is significantly more volatile, you can see the strong correlation over time. The inflation index shown, the IPCA, is a broad consumer-based index and is the most closely correlated index to the inflation that we would expect in our real base cost. In 2020, that measure of inflation was around 4.5%, with the average bank consensus forecast for 2021 currently at 3.7%.
Now we'll be doing our best to increase our transformational savings to offset at least a portion of these costs, but we do face risk of near-term inflationary headwinds. What we've covered so far is really meant to help you understand the magnitude and sources of Mosaic's cash generation capacity moving forward. What we're going to cover next is how we think about deploying that capital. Mosaic's capital allocation program is designed to maximize the value of the enterprise by focusing on two principles. The first principle is maintaining and strengthening our core business. We'll do this by making appropriate investments in the existing business to ensure continued safety, reliability, and environmental and regulatory compliance. We also plan to strengthen the financial profile of the company through targeted de-leveraging and active liquidity management. The second principle is driving incremental value through the prudent management of excess capital.
This includes making value-enhancing organic or inorganic investments when risk-adjusted returns exceed cost of capital and alternative capital uses like share repurchases. It also means returning capital to shareholders through a safe and growing dividend, supplemented with additional capital return when appropriate. With that philosophy in place, we've set goals balanced across these three areas, strengthening the balance sheet, investing in the business, and shareholder returns. Our balance sheet goals include paying down $1 billion in debt over time and maintaining investment-grade metrics through all parts of the cycle. Our goals for investing in the business really are simple. We want to focus on high-returning projects which improve our competitive positioning, reduce risk, and drive growth. For shareholder returns, we want to provide a regular dividend that's safe and sustainable, even in challenging markets, and grows over time as our earnings profile improves through transformation and other initiatives.
We may also look to supplement dividends with share repurchases from excess capital over time as appropriate. Each of these uses of capital competes against the other to maximize long-term value creation with an overarching goal of maintaining a certain level of balance over time between the three. Now, as we've spoken about in the past, we remain committed to reducing our total debt by approximately $1 billion over time. With multiple debt maturities over the next several years, we'll have an opportunity to make material progress on that, if not get it done. There are several reasons why we think that this is important. First, as you can see from the slide, our debt metrics are very sensitive to our earnings profile, and given the cyclical nature of our business, the company can have very different leverage profiles over time.
Shown on this slide is what Mosaic's net debt to EBITDA would have been at various times over the last 10 years using our year-end 2020 net debt level of $4.1 billion and holding it constant through that period of time. As you can see, net debt to EBITDA would have ranged from just over one time to over four times, which not only exceeds what we think is appropriate for this business, but that also begins to exert pressure on our equity as investors price in concern over the strength of our balance sheet. We think our targeted level of debt reduction would certainly improve our financial profile, particularly during periods of market weakness, as shown by the dotted line on this chart. In addition to improving financial resiliency through this cycle, though, we also believe that our targeted debt reduction will improve our cost of capital.
We've conducted an analysis covering this 10-year period, looking at different capital structures from single A to a non-investment grade, and taking capital market conditions into consideration at various times, looking at what Mosaic's cost of capital would have been over this timeframe. What this analysis tells us is that 90% of the time over the last decade, Mosaic would have achieved its lowest cost of capital with a capital structure consistent with mid to strong BBB, Baa credit metrics. Not only would a stronger balance sheet make Mosaic more financially resilient, it should also translate into an improved cost of capital.
Finally, as we saw with the acquisition of our Brazilian production business from Vale a few years ago, value-creating strategic opportunities can present themselves during periods of market stress, and we want to be sure that we're well positioned to act without stretching our financial profile, should that happen again. This next slide shows how we've allocated capital in each of our three buckets over the last three years. In addition to repaying approximately $800 million in long-term debt, we've continued to fund our K3 project at Esterhazy and invested an incremental $400 million in transformation and growth initiatives, all of which are delivering very attractive after-tax unlevered returns. Over the first two chapters of this analyst presentation series, we identified $485 million in growth capital investments, including our K3 project, that are expected to drive $700 million in EBIT improvement by 2023.
These are the types of investments with high returns and fast paybacks that we will continue to look for in the future. Today announced a 50% increase in Mosaic's targeted annual dividend. We've also returned $330 million to shareholders over the past three years. What does all this mean for 2021? We expect to deploy $300 million in growth capital, funding the specific initiatives, including K3, that we've detailed in our earlier chapters. All of which deliver attractive returns and position the company for future success. We announced an increase in our targeted dividend to $0.30 per year. Continue to expect to retire $450 million of debt maturing in November of 2021. Beyond these plans, we'll continue to monitor markets, our cash generation, and our cash builds.
If markets remain as constructive as they are today, we would expect our cash balances to build during the year. Should we see that, we'll look to make incremental capital allocation decisions consistent with the framework we've outlined today. With that, I'd like to turn it back to Joc for concluding comments. Joc?
Thanks, Clint. As we wrap up the fourth and final chapter of our analyst day series, we want to reiterate a few key points. Our investments in the business on projects like K3, Transformation, and NextGen Mining have already begun creating value and will meaningfully contribute to our target of incremental EBITDA of $700 million by 2023, apart from any market impacts. Mosaic Fertilizantes has proven to be a pivotal transaction that gives us a platform for growth in South America, one of the world's most important agricultural markets. We are re-engineering how we manage our supply chain, as Karen discussed in chapter one, which is providing us with significant cost benefits.
Our collaborations with companies like Anuvia Plant Nutrients, BioConsortia, and Sound Agriculture, while still very early in the process, show our commitment to expanding our specialty products portfolio over time to meet grower needs while also promoting soil health and sustainability. We believe our efforts to optimize the business will result in meaningful growth in free cash flow, which will be used to strengthen our balance sheet, beginning with this year's $450 million debt maturity, and returning to shareholders through a stable dividend and share repurchases when the opportunities arise. Now, I'd like to shift from our prepared comments and open the floor to answers to your questions. Paul, over to you to facilitate.
Thanks, Joc. As a reminder, you should see a Q&A box on your screen where you can enter questions. We'll do our best to get to as many questions as we can over the next few minutes. Afterwards, feel free to follow up with Laura or myself with any additional questions. Joc, our first question comes from Andrew Wong at RBC. It's a CapEx question. What's the cadence in the expected CapEx decline? Is it gradually dropping from $1.1 billion to $800 million over several years? Is there a step change when Esterhazy K3 is completed at the end of this year?
Okay. Thanks, Andrew. Yeah, there is certainly a step change when Esterhazy completes, although, as Clint mentioned in his prepared remarks, we do have some cost in our phosphate business for developing new reserves. Now, those new reserves will allow us to put off larger investments in new plants. We think those are important investments. I think there's a little blip in the 2022, 2023 timeframe, but after that, it does reduce to much lower levels. Clint, is that about right?
Yeah, no, I think it is. I think there are, I guess, kind of three things that I would note on that. One is, when you look at our sustained CapEx, again, that tends to be within a range of $650 million-$750 million per year. I think I noted in my prepared remarks that we're probably at the high end of that range this year, but I think that should revert back to kind of the mean, if not a little bit lower over time. I think that's kind of more of a near term this year type of phenomenon. I think with K3, we've had $100 million decline, roughly, from last year to this year in that CapEx spend. We would expect something similar next year as that project ramps down. Again, I think you should see a pretty smooth glide path on that.
Really the only blip in the interim is what Joc noted a minute ago, and most of that's in 2023.
Our next question comes from Vincent Andrews at Morgan Stanley. With growth CapEx set to decline significantly between now and 2025, how will the company grow organically? Is the strategy just to manage the cycles and be a return of capital to shareholder story? What organic growth and/or M&A opportunities would you consider if ag remains in a prolonged upcycle and free cash flow remains robust?
Yeah. Thank you, Vincent. Clearly we've said we will be always looking for that balanced approach to capital allocation. When you look at our capital that we have in this plan, that wouldn't include organic or inorganic opportunities that haven't been developed or proven yet. For instance, a good example might be, as we continue to grow our premium product market, our MicroEssentials production capacity may come to be the limiting factor for growth there, at which point we would likely allocate capital for that. Of course, M&A is always on the cards. You've seen our investment in some of the soil health. While those are modest, we'll continue to look at those type of growth opportunities. What I would say is we're certainly not going to ride the cycle per se.
We're going to always be looking for opportunities that can add real sustainable value to our shareholders.
Adam Samuelson from Goldman Sachs asks, what are Mosaic's return hurdles for capital investments, and is there any differentiation between debottlenecking or efficiency versus capacity additions? Also, what are our return hurdles for M&A? How does Mosaic define opportunistic M&A?
Okay, thanks, Adam. That's a lot of stuff loaded into a question here. Without being too specific, obviously the first hurdle for any capital has to be how does it compare against, for instance, either paying down debt or more relevantly, buying your own shares. There's a certain return expected from buying your own shares back, that's the first hurdle. You have to take a look at what's the risk profile of doing anything. There has to be a risk return, including country risk, technical risk, et cetera. There's a bunch of areas. We would probably look, for instance, just as an example, something like cogeneration, which we invested in New Wales a few years ago. We would probably look at a lower hurdle rate than we would, for instance, of an M&A opportunity in Brazil.
Bottom line, we expect those to be higher than our cost of capital. We expect them to have a higher than cost of capital plus a risk premium, depending on the situation. Clint, any thoughts on that one?
No, I think you've really covered it, Joc. I think certainly we look at the risk-adjusted return. We certainly want to beat our cost of capital. We do compare it against the economics of share buyback to be sure that we're allocating capital in the right way. I think some of that analysis does pivot off of some of the risk things that you talked about a minute ago. It also tends to pivot off of, are these direct line of sight, cost reduction, proven technology, what have you. Those kind of things go into the risk analysis as well. I think as you look at some of the opportunity CapEx or some of the growth CapEx that we've been allocating to more recently, what we're finding is very high level returns and very quick paybacks.
One year, one and a half year type of payback, some even under one year payback. Those are the kind of things that we look for, and that we'll certainly look to fund. The other benefit of those is that they're generally not high dollar type of investments. We're also mindful of, as we look at opportunities, we want to be sure that we understand what other options we would either still be able to take advantage of or that we may need to defer and maybe take off the table. I know that's a lot between what I've said and what Joc said, but frankly, a lot goes into that consideration.
Thanks, Clint.
Ben Isaacson from Scotiabank asks, phosphate has received the least transformational benefits in 2020 at $66 million, despite it being your biggest business segment. Why is that? What are the identified phosphate initiatives going forward?
Thanks, Ben. First of all, I have to say that our phosphate business has been working on transformation for a number of years or working on continuous improvement for a number of years. If you look at things like our cost of mining, I think last year at $36, I think that's been consistent pretty much over a 10-year time horizon. I think the folks have done a fantastic job of maintaining a cost profile over time. The other piece to think about on phosphates and why it might be a little smaller number is I think some of the benefits are we've done a bunch of things in 2020, and we're just barely starting to see the benefits come in in 2021.
Two good examples of that, I think the consolidation of our North American business unit, which really only happened in the fourth quarter of last year. We're going to see significant savings there. We're going to see significant savings, that might not directly go to phosphates, I accept. We're also going to see significant savings from this NextGEN Mining. I think we're just barely starting to understand how much benefit that's going to give us as we automate a lot of processes that have been more manual in the past. I think there's a lot more to come. It's more timing than anything. I think as a business of that size, there's lots of opportunity.
John Roberts from UBS asks, after the $450 million bond repayment, what's the priority for the remaining $550 million debt reduction? Will share repurchase be deferred until the remaining debt reduction happens?
Thanks. Clint, why don't you just take this one and talk a little bit, again, about the allocation strategy and the balance of all of our priorities?
Yeah. Thanks, Joc. John, as we think about it, I don't think we need to preclude doing really anything, including share buybacks, until we get to the full $1 billion in debt reduction. What we've tried to do is to build a program that has balance. Certainly, we want debt reduction to be a priority. Once we pay off the maturity this year, we'll still have two maturities over the next two years, in 2022 and 2023. Total's about $1.45 billion. Certainly, we have the opportunity between those two years, to be able to finish out, the debt reduction and hit our target. With that said, again, I don't think we approach it in terms of saying, well, we can't do share repurchases, or we can't do other things until we finish that debt reduction. I think the thought is let's have balance.
Let's plan these things so that we still retain the flexibility to do other things, including investments or capital return while we are moving toward hitting that debt reduction target.
Thanks, Clinton. Let me add, we're fairly certain on paying back the $450 million at the end of this year. I think as we move into next year, our debt metrics will be pretty good for where the market is today. Long-term, we want to take it down by about $1 billion. Whether we do that in 2022 or whether we refinance with a term loan and do it over the next couple of years, I think is yet to be seen, depending on markets and everything else. More to come.
Adam Samuelson from Goldman asks, given the cyclicality of cash flows, has the company considered special dividends or a variable dividend policy as a means for returning cash to shareholders as opposed to share repurchases?
Yeah. Thanks, Adam. The simple answer is yes, of course. We've looked at all forms of how we might return excess cash to shareholders, there's a number of ways to do that. One of them is a special dividend. One of them is a variable dividend depending on earnings in a year. The other one is share repurchases. Of course, the one we've changed today is our regular dividend. We're looking at all of those as possibilities, a lot of things go into it, including what the market looks like at the time, et cetera. Again, we're open to all forms of returns, we just have to figure out what's the best on the day.
Ben Isaacson from Scotiabank asks a few questions on Brazil. First, he asks us to talk about the macro situation in Brazil, things like weaker growth, COVID, inflation, and what that means to operational and financial risks in the Mosaic Fertilizantes business. The second question has to do with what continued growth exists in South America, and are there opportunities for further M&A in South America?
Okay. Thanks, Ben. Well, first, let me talk about the macro situation in Brazil. Obviously, COVID, and you're reading it in everything from The Wall Street Journal on down, that COVID is really quite serious in Brazil today. Having said that, we have been very successful in maintaining our operations, keeping our people safe, and we certainly have not had any transmission that we would say is directly attributed to our own sites. Although there's a lot of, obviously, community transmission in most of the areas where we're operating. So far, we've managed through that quite effectively, and I think at this stage, we believe we can continue to manage through that effectively. As I move to the macro situation in Brazil, there's a lot of uncertainty around the macro situation in Brazil.
I think they just dropped off as one of the top 10 economies of the world. In that, two things that should be mentioned from our perspective is, first of all, agriculture was the one area in a negative growth scenario for Brazil, which was positive growth. The agriculture sector continues to be a strong sector within Brazil. With the real continuing to devalue, we see extremely good farmer economics, which again, bodes well for our products. Fertilizer continues to be very affordable, and we are in a very good position in Brazil to take advantage of that and continue to do well in that market because of our position, both from a production and distribution perspective. In terms of Brazil growth, we still talk about the South American growth engine, and that's mostly about Brazil.
We see opportunities in both production and in distribution where we can increase our footprint. I don't have any that we're going to talk about today, obviously, but we're constantly looking at how we can economically, with a great return, take advantage of what is a very good market in Brazil, a very good agricultural market in Brazil.
Mark Connelly from Stephens asks, as K3 moves towards completion, how do you think about the ultimate return on that investment versus what was expected at the time that it was first announced when some of the assumptions about K1 and K2 were different?
Yeah. Thanks, Mark. Look, a couple of important things on K3. First of all, K3 was ultimately going to be needed because the distances and the difficulty of mining at K1 and K2, we were getting to where the distance to the shaft was 45 minutes by car and stuff. That starts to drive costs up significantly. The other piece of K3 that was so important was, of course, the brine inflow risk, which at the time of decision was over $200 million a year. There's really three pieces to the economics of K3, if you will. First is the sustaining of the business and replacing the K1 and K2 shafts with a lower cost, closer to the shaft type mining, which will drive costs down significantly. The second is the elimination of brine inflow, which K3 will do.
The third is an incremental 900,000-ish tons of new production from that project. In terms of the payback, what we're seeing today is being able to meet our total production needs simply on Esterhazy. Because of K3 ramp up, we're able to idle Colonsay, which is taking high cost product and replacing it with very low cost product, if you will. Knowing that we can meet the market there, that's a pretty good payback and doing what we thought it would do. Obviously, today, I think last year, our brine inflow costs were less than $80 million. They'll probably be almost insignificant in 2021. You can almost look at that as being a $200- and- some million a year return. Although, when you look at it in 2021 terms, you go, well, it's only $110 million or something.
That's not what we were looking at when we made the decision. The reason we've been able to push it down is through technology, but also because our risk profile's changed and the timeframe has changed, so we're not having to make those long-term investments. I think that has turned out exactly as we'd hoped. Being two years ahead of time means it's allowed us to accelerate that. That's exactly what we expected. Then, you'll see it as K3 ramps up, the cost profile of K3 is going to be excellent. Let's call it $50-$55 a ton probably, cash cost, which means that we're going to be almost probably $30 or $40 a ton less than what Esterhazy was on the decision day. I think all of those things come together to make that a pretty attractive investment overall.
Of course, we've been able to keep it under the $3 billion we started with. It's going to be a good, let's call it a real flagship for Mosaic going forward, and I think for good reason.
Joc, maybe if I can just add a couple of things. I think at our Analyst Day in 2019, we talked quite a bit about the economics of the project, and at the time, kind of laid out the net investment in K3, and that we would expect, on an unlevered after-tax basis, a 15%-18% return. I think with the fact that we're still on target, if not inside our budget for K3, and the fact that we have accelerated that project by two years and the benefits associated with it, factoring in some of the acceleration of the CapEx even, we still see us being inside that range, if not kind of toward the high end of that return range.
Yeah.
When you couple that with the reduction of our risk profile on a risk-adjusted basis, I think that project is coming in pretty much on target, if not a little bit better than expected.
Yeah. We can't put dollars on the risk reduction or the idea that we had to, at some point, do something to just sustain the mine over time. I think with all of those things, it's been, I think, better than I would have expected or certainly better than what we presented to the Board 10 years ago now.
Jonas Oxgaard from Bernstein asks, your cost of capital is 5%-6%, which is a very low bar. Do you have a formal return hurdle independent of your cost of capital?
Yeah. Thanks, Jonas. Again, we haven't defined that per se as, we won't do anything under 12% or 10%, as we were talking earlier. Realistically, we're looking for projects that have very superior returns. I would argue that probably, we're looking at stuff that's, call it on a simple payback basis, less than two years-three years for most of the bolt-ons, and then, for a bigger one, it might be a four-year. If you look at, for instance, how we looked at Fertilizantes when we bought it, we were looking at what was the accretive value of it. As you saw there, we thought we could bring significant value to the table, and that's how we looked at it.
I would argue that the return on Fertilizantes, considering last year, $475 million of EBITDA, represents about a four-year payback for the $2 billion we invested there. That's got to be a 20% return. That's kind of the way we would look at it rather than say, hey, here's a hurdle. Hurdles are incredibly difficult to be precise on, and so I think you end up making poor decisions sometimes on stretching to get a hurdle. Clint, anything to add there?
No, I think that's right. I don't really have anything to add on that.
Michael Piken from Cleveland Research asks a few questions on inputs. He asks if we can discuss the impact of the recent winter storms on our internal ammonia production in Faustina, and should we expect a higher mix of spot ammonia over the next several months? The second question as follow-up is, are there further steps we could take to become more self-sufficient in ammonia production, or are we comfortable with the current allocation of how we source ammonia?
Yeah, thanks, Michael. Look, first of all, if we look over the last little while, we did actually sell some natural gas, I believe at about $12, Clint?
I think that's right, yeah.
Every storm has its silver lining, I guess. We were able to sell some natural gas that we weren't able to use because of the storms closing down the Faustina plant for a little while. Overall, I think our expectation is, and this won't change really materially over the next little while, we're gonna be about 1/3 produced ammonia, 1/3 from our CF contract, and 1/3 on the open market. As right now, the CF contract is in the money, we will probably be pushing to the top end of that. We have a variable range of 600,000 short tons to 800,000 short tons per annum, I believe, are the numbers. We'll be pushing that towards the higher end of that, meaning that we'll pick up a few more tons at a little bit below market.
We'll probably try to optimize our production from Louisiana because, of course, it's well in the money. Those ratios will change. Overall, we think that's a pretty responsible way to look at our ammonia purchases, 1/3 being hedged, 1/3 being our own production, and 1/3 being on the open market. Recognize when the prices were low, we were probably a little behind the market because CF was a little above. We were a little under the water there, but now we're in the money on that hedge. Overall, we think it's a pretty good balance. It's all about managing the risks.
Joel Jackson from BMO has more of a strategic question. He asks what the main gap that Mosaic believes it exists in. Is it greater Brazil distribution, perhaps in the north of the country? Is it greater or more diverse rock reserves? Is it a push into greater specialties, or is there something else?
Yeah, thanks, Joel. Look, from a strategic perspective, I think there's great opportunities to expand our distribution in Brazil. Particularly, Joel, you said it, into that northwest. If you go into the northwest part of the country, that's where agriculture is growing, and it is growing quite rapidly. We think there's great opportunities to extend our distribution there. The Government of Brazil has talked about making Brazil more self-sufficient on its own production of fertilizers. I think they're gonna probably look at ways to lower the impact of some of these interstate taxes that have been a bit of a restriction, and I think when they do that, and if they do that, but even if they don't, I think production in Brazil is of value because there's a big advantage to be in-country. I think there's some opportunities there.
Obviously, you talk about new products. I just think that we have to be looking at where these products are going, particularly products that look at better phosphate efficiency, generation of nitrogen at the plant level or at the plant roots. All of those things, they may not be for the next couple of years, but in the next 5-10 years, you wouldn't want to see a revolution in that area and us not be prepared for it or us not be at least in that market. I think we're into those things at the right time. The technology's changing fast. Then the other gap, I think, is just continuous investment in technology to keep our operations as low cost and as efficient as they can be. That was three gaps.
We've got a few more follow-ups on capital expenditures. Ben Isaacson from Scotiabank says, it looks like sustaining CapEx is going down by 2025 versus 2021. How are we able to fight off inflation and reduce maintenance CapEx by what visually appears to be 5%-10%?
Thanks, Ben. I'll put this on Clint as well, but I think the simple answer there is, look, we've gone through a period of time, particularly in our phosphate operations, where we've had to invest in quite a bit in terms of gypstack replacement, new clay settling areas, and those are big chunky investments in sustaining capital. They're business sustaining rather than maintenance sustaining. In terms of maintenance sustaining, really, we're looking at that from a perspective of the better we can do condition monitoring, the better we can do basically mechanical integrity and predictive maintenance, the better we can keep the real sustaining costs down. If you look at it, the last couple of years has been probably three areas where we've spent a little more, and we expect that to go down.
This is particularly in phosphates, but we've had gypstack expansions, we've had tailings expansions. Of course, some of our acid plants have got to 40 or 50 years old, and we've had to replace converters and stuff like that, and those are big chunks. We think that cycle sort of goes through, and then you get into lower sustaining. Brazil and Canada have been pretty low sustaining other than some of the costs of dams in Brazil. Clint, any other areas? I think that's why it's more structural than it is us being able to, let's say, cheap out on sustaining capital.
The only other things that I would note, Joc, is that our sustained CapEx, again, tends to be in a range of $650 million-$750 million. I think this year, in 2021, we expect to be at the high end of that as we've talked about. Again, I think reverting back to more normal levels over time, over that span, certainly by 2025. Part of the reduction is just you're starting at probably the higher end of your range and then kind of reverting back to more normal levels. I think the other thing is as you look at where those CapEx dollars are being spent, the majority of them tend to be in phosphates. I think Brazil and potash tend to be lower sustaining CapEx spends.
I think as we look at the work that the teams have done, Joc talked a little bit earlier about how mining costs have remained pretty flat over an extended period of time, and that's because the team works really hard to manage cost and to basically offset inflation. They do the same things within the CapEx program as well. Certainly, that's something to watch, I think, over time, particularly in Brazil, where you may have an elevated level of inflation. I think generally speaking, the teams work pretty hard to offset that impact. Then it's just a matter of the lumpiness of the work that Joc was just referring to. What year does it fall into? We have a heavier than normal year this year. I think going forward, we see a little bit more normal levels on the back end of that forecast.
Let me use an example as well, I guess, Ben, just to say there are cost saving areas too. You look at, for instance, our Belle Plaine operation up in Canada, where it was probably costing us about $35 million to build a cluster. They've now been able to reduce that down to about $25 million through really thinking through how you build it, how you develop it, how you do the drilling, et cetera. There are areas where we've been able to really kind of rethink the capital needs. The other one, of course, is K3, where now you have a new mine. You go from three mines to one. That's certainly going to reduce sustaining capital. Obviously Brine Inflow, there was a lot of sustaining capital associated with that. There's a couple areas, but like I say, a lot of it's structural.
Jonas Oxgaard from Bernstein. He asks, does the relatively low growth CapEx figure in 2025 imply that there will be no new mines before this decade?
Yeah. I think that's pretty fair to say. We do not expect to develop the next new mine. We've continued to push it off. We're focusing on what we're calling South Fort Meade's extension to the east. We found new reserves there that have extended the life and time of South Fort Meade. Our Wingate mine has, I think, 25 years, lots of reserves there. One of the advantages of buying the South Pasture operation or the CF assets was we ended up with the South Pasture extension, the southern part of the South Pasture mine, which is now being mined. That and the start of the Ona mine is now being mined from Four Corners. One of the big things there is these guys, when you talk about innovation, they're now pumping almost 20 mi from the draglines to the plants. That was unheard of a decade ago.
When I came to Mosaic, we were talking about being able to pump about 10 mi. Now through better technology, this whole NextGEN Mining where we're monitoring it from a central control room, we're using variable frequency drives to manage the pressure across a long haul, and we're able to mine that or move that slurry which is, of course, slurry across a long distance. Bottom line is we're continuously pushing out the need for new mines. Then if you look at the next phase, it'll be actually starting up our South Pasture operation again, which is sitting idle today. We've got lots of reserves, and we've got a way to mine them for quite a few years to come. We're really pushing off some of those investments we were talking about five, 10 years ago by decades.
I think maybe just a couple of more questions given the time. Jonas does have a follow-up. If the CVD complaint is upheld at the final determination, how long are those tariffs expected to be in place? Maybe a follow-up, I think specifically for Jenny, how does she expect this to impact the global phosphate market?
Okay
Our sales in different regions?
Did I hear that was from Jonas? Sorry, Paul, I didn't hear who the question was from.
That's right, from Jonas.
Okay.
At Bernstein.
Yeah. Thanks, Jonas. Well, first of all, let's talk about the mechanism of the CVD, which my best understanding is we should get a ruling today. If that ruling is positive, it is a five-year thing, but it's reviewable, I believe, on a yearly basis, depending, and the Moroccans or the Russians, I believe, can bring up reasons that they may think it's reviewable over time. Expectations though is that this would be in place for about five years probably, or I guess until they change their practices. Noting that the Department of Commerce has ruled that there are unfair subsidies to the tune of 47% for EuroChem, about 9% for PhosAgro, and just under 20% for the Moroccans.
I'll hand this to Jenny to talk a little bit about it, but I do want to emphasize that when we talk about what is the market influence going to be, this will specifically be a market impact on the U.S. market. What we would expect to happen on a macro level is for the global markets, the trade flow will change, but the overall supply and demand doesn't really change. In other words, the Moroccans will probably redirect their product to Brazil and to India. The Russians will redirect their product to whatever markets and move away from the U.S. markets. What we're already seeing is the Australians, the Mexicans, the Saudis, the Jordanians, and the Egyptians are all importing to the U.S.
As a matter of fact, I think we saw imports, and I'll get Jenny to confirm this, in the range of 800,000 tons into the U.S. in the fourth quarter. What we're seeing is a fairly normal lineup of imports, but imports from what I'm going to call fairly traded countries as opposed to from the subject companies of the CVD filing. We expect the U.S. market to probably stabilize at near parity to the U.S., recognizing it's been three years at significantly under parity to the global market. It's driven the price increase lately, we expect that to balance out to being global markets adjusted for freight, basically. Each free market company that imports here will decide based on their best alternative where to send their product.
We don't expect the global market to change, but we do expect the U.S. market to be affected. Jenny, do you want to talk about?
Yeah.
Sorry, the countervailing duty?
Yeah, Joc, I think you got all of the things covered. I just want to reiterate that we've been talking about the global supply and demand does not really get changed with our filing on the countervailing duty case. Since the filing, we have seen the CVD was a catalyst, basically eliminated all the discounted price in the U.S. which have been in place for three years. Now with the trade flow adjustment, today, U.S. market is already traded at a parity to the Brazilian market. That's basically how we see today, as Joc mentioned, the year-to-date input into the U.S. market today is actually back to the normal year, which is even greater than the year of 2019. Lastly, I want to say it is a very tight phosphate market.
It is driven by strong demand around the globe. We are not seeing much more new addition in the supply side of the equation. We feel very confident that the market momentum for phosphate is going to sustain regardless of the outcome of the CVD.
Thanks, Jenny.
Well, we still have a few questions left in the queue, considering we're past the hour, I think we'll try to be respectful of everyone's time. Laura and I will try to get back to everyone who still had questions after the call. With that, I will turn it back to you, Joc.
Thank you. Again, as Paul says, Laura and Paul are available to answer ongoing questions, of course. Let me end by saying, I just want to thank everybody for attending our analyst presentation series. I think they've been helpful and informative. I will leave you by saying we're really excited about our future. We're excited about our ability to deliver on the promise of our mission to help the world grow the food it needs. We're excited about our ability to do so sustainably, and we're excited about our ability to deliver superior results for our shareholders, for our employees, for our communities, and our customers. Thank you for joining us. Have a safe and healthy day. Take care.