Good morning, ladies and gentlemen, and welcome to the Caterpillar 3Q 2019 Analyst Conference. [Operator Instructions]. It is now my pleasure to turn the floor over to your host, Jennifer Driscoll. Ma'am, the floor is yours..

Our U.S. GAAP-based guidance for profit per share continues to include the benefit of the $0.31 discrete tax item we recognized in the first quarter. In a moment, you'll hear from Andrew with a summary of this quarter's financial results, but first let me turn the call over to Jim for our third quarter highlights, which appear on Slide 3. Thank you..

services, operational excellence and expanded offerings. At our Investor Day in May, we announced our goal to double Machinery, Energy & Transportation services sales between 2016 and 2026. This target is challenging yet achievable.

By growing our services, we will help our customers improve asset utilization and availability while reducing their owning and operating costs. We're continuing to invest to drive services growth, including expanding our digital capabilities.

We continue to connect assets and invest in our digital architecture to provide actionable insights to our customers. For example, our Cat Inspect app helps customers identify the maintenance needs of a machine and plan accordingly. We're seeing close to 100,000 inspections on the app per month.

In the area of operational excellence, we use lean principles at our large engine facility in Lafayette, Indiana to improve safety, enhance engine quality and achieve greater manufacturing efficiencies.

Lean improvements drove a 40% reduction in assembly time for our large 3600 engines, allowing us to produce more engines from the same facility while improving safety and quality. Our drive for operational excellence is a never-ending journey. Lean is helping us respond more quickly to changes in industry demand.

And the final pillar of our strategy is expanded offerings, enabling us to grow our business by addressing the diverse needs of our customers around the world. We introduced a new Dynamic Gas Blending or DGB engine for our well-servicing customers.

The new Tier 4 engine, which is unique in our industry, allows customers to replace up to 85% of diesel fuel with natural gas. Our DGB product helps our oil and gas customers to be more successful by improving operating economics by offering fuel flexibility.

We're continuing to offer -- we're continuing our focus on autonomy, semiautonomy and remote operation as we expand our offerings. We believe Caterpillar leads our industry in all 3 areas. Some of our early customers have cited productivity benefits of up to 30% using CAT autonomous mining solutions.

In addition, our customers are seeing real improvements in safety, in some cases up to a 90% reduction in safety incidents. One of our competitive advantages is that we can retrofit our competitors' equipment, making our autonomous solutions an option for mixed fleets. We're encouraged by the recent wins we've had in autonomy this year.

With that, I'll turn the call over to Andrew for a closer look at our financials..

We've said we intend to increase the dividend by at least high single-digit percentage in each of the 4 years. We repurchased $1.2 billion of our common stock in the third quarter. Our quarterly share repurchases -- repurchase plans consider our projected cash flows and takes into account the intrinsic value of our shares.

We will continue to be flexible, and in periods of time our cash flow is more variable as a result of dynamics in the external environment, we'll be able to use our balance sheet to maintain flexibility and returning substantially all of our free cash flow to shareholders over time.

We continue to project share purchases for the second half to be similar to the first half, and we continue to expect to reduce our total quarterly average diluted shares outstanding by about 9% from the first quarter of 2018. Meanwhile, we continue to invest in services, operational improvements and expanded offerings.

And we ended the quarter with $7.9 billion of cash on hand. You may have seen our announcement last month that we had issued $1.5 billion of 10-year and 30-year notes in the third quarter, enabling a $1.5 billion voluntary contribution to the U.S. pension plans.

This action increases the plans' funded status, allowing Caterpillar to further execute our strategy of reducing volatility in our pension liability. A secondary benefit of the contribution is we -- that we now don't expect to make any further contributions to the U.S.

pension plans for a substantial period of time, therefore freeing up cash for discretionary deployment. This contribution has no impact on our credit rating metrics. Finally, it was funded at attractive interest rates; and was actually part of our cost-saving initiatives, as it lowers our long-term pension insurance-related costs.

So finally, let's turn to Slide 14 and recap today's key points. Third quarter's sales and revenue declined by 6%, and profit per share by 8%, due to volume declines driven by changes in dealer buying patterns.

We reduced our 2019 profit per share outlook range to a range of $10.90 to $11.40 based on expectations that dealers would further reduce their inventory levels and that end market demand would flatten in the fourth quarter. We're proactively managing production to address expected changes in demand.

We're working on the competitiveness of our cost structure, and a relentless execution of the operating and execution model remains at the center of everything we do.

Our overall financial position remains strong, and we remain very much committed to our strategy of profitable growth and deployment of capital back to shareholders through a growing dividend and consistent share repurchases. With that, I'll hand it over to the operator for the start of the Q&A session..

[Operator Instructions]. Your first question is coming from Jamie Cook from Crédit Suisse..

I guess, a couple questions. First, on the resource side, I think that sales and margin surprised people a little, while the overall quarter was fairly good. You talked about warranty. You talked about production cuts.

Is there any way you can sort of quantify what the impact of that was in the quarter on the margin front? And is there anything that you're seeing sort of from the order intake side to suggest that there is more downside risk on the sales side for 2020 and what that implies for margins? And then my second question, bigger picture sort of on 2020.

I guess the assumption is you go into 2020 with producing in line with retail demand. Is there any way you can help us with other puts and takes? It sounds like there is a cost-cutting program that could be additive, sort of incentive comp, share count lower.

I'm just trying to think about the puts and takes that we should consider positive or negative. We can make our own assumption on volumes..

Well, Jamie, just a couple of comments. I think it's important to remember for Resource Industries that it's a mix of mining products and heavy construction in quarry and aggregates. So you've seen weaker sales unexpected on the heavy construction and aggregate side of the business.

Mining sales on a year-to-date basis continue to be positive, and rebuilds and part sales remained strong across the board. In the third quarter, we did see dealers reduce inventory related to heavy construction and some isolated pockets in mining for coal-related inventory.

Again, I just want to emphasize that, again, RI does include both that heavy construction and mining, so it's not just a mining story..

And moving on to the margins, Jamie. It really is all around volume and mix. So the biggest driver both on a year-on-year and quarter-on-quarter basis all relates to that. All the other items are puts and takes virtually as we move through.

Remember, versus last year, we have lower short-term incentive compensation, but that is offset partly by a higher warranty, but versus quarter-on-quarter, all the other items are -- really washes all down to volume.

And then moving into 2020, I mean, obviously at this stage, as we've said, the situation global economic outlook is very uncertain, so we are not going to be providing sort of sales guidance or top line or outlook guidance at this stage. We're still in the middle of our budgeting process, and things are very fluid.

However, on some of the things, yes, we are continuing to look at our cost structure. As I've mentioned, we are looking at things like G&A and back-office costs, procurement costs and so forth. All of those are initiatives that are ongoing and will continue to go. They are part of maintaining a flexible and competitive cost structure.

And then as far as share count, as we've said, by the end of this year, we'll have reduced the share count by about 9%. That will have an impact of about -- obviously they're split between 2018 and 2019, so there will be a bit of a tailwind from that. Next year, there will be a little bit negative on short-term incentive compensation.

We expect this year to be about $150 million lower than our base plan. That obviously will be reset for next year..

Okay, but I mean in -- the goal for 2020 to produce in line with retail demand, so sort of in a flat market it's not unreasonable to assume you could probably grow earnings..

Obviously it depends what your assumptions are on top line, yes. So if your assumption is we have a flat retail market, that obviously would flow through, yes..

Your next question is coming from Rob Wertheimer from Melius Research..

Thanks for the commentary on dealer inventory and otherwise. It seems like you made a positive step. And dealer inventories are going down, up last quarter, going down now; and yet that was only maybe a -- I don't know, maybe 1/4 of the total cut to revenues. So I'm trying to square the circle here.

The dealer sales in retail seemed pretty good in your release this morning, up mid-single digit. You're reducing dealer inventory and those sales are up.

And so I guess dealers must have really gotten more conservative on orders, but could you just talk about dealer inventory cut being 1/4 of that overall revenue cut? And then did you see any cancellations in Solar or any direct sales or larger projects?.

Yes. So Rob, if you think about -- I think you're talking about over the full year rather than actually in the quarter because obviously, in the quarter, we saw a quite significant year-on-year impact of dealer inventory because it was up $0.8 billion last year and down $0.4 billion.

So when we started the year, if you remember, our assumption was that actually we would have flat dealer inventories and a modest growth in sales. And obviously now what we're saying is with a $500 million build of sales we are seeing slightly lower sales for the full year.

Yes, we are -- our view is that probably we've lost about 2% on retail versus where our base case guidance started the year. So effectively that's been the big driver. That mostly relates to expectations out there and particularly obviously in the fourth quarter where it's dampened quiet -- dampened down to flat.

We've been running a sort of 4% to 6% for the full year. Our full year expectation was probably at the top end of that range, and we're not going to meet that..

You asked a question about larger order for Solar. No, Solar's business continues to remain strong..

Your next question is coming from David Raso from Evercore ISI..

I think people are just trying to figure out the -- regardless where The Street is for 2020, you've now kind of put out a $2.40 midpoint adjusted EPS for the fourth quarter. And just trying to get a sense; sort of annualize that, say $9.60. The margins for the fourth quarter seem to be implied around 13% to 13.5%. Just trying to get a feel from you.

And I know 2020 is a lot of planning is still going on, but the approach you took to the fourth quarter, to get a sense of that $9.60 run rate -- or do we feel like we're trying to bottom the earnings a bit here if the retail can just be flattish, even down a bit next year? How do you view your margins in the fourth quarter? That 13% to 13.5% implied, is that -- I mean, how much of a hit? Because obviously $900 million in inventory reductions is a bit of a drag that maybe you want to see again in 2020, that big a drag in 1 quarter.

Can you just take us through your thought process on how you view the fourth quarter and those margins?.

Yes, Dave. So the fourth quarter, as you always -- as you know, is always our lowest quarter from a margin perspective. As we think through the year and our production cycles and the way obviously through accounting and the way we benefit from operating leverage by volume, effectively you tend to see Q1 and Q2 stronger margins, Q3 slightly lower.

And particularly in CI you normally see a historic -- at least a 1.5 percentage points drop in margin in Q4. So that is why normally Q4 margins are lower than for the balance of the year. As we look for this year, obviously the dealer inventory would normally be a further reduction in margin because you're having an element of deleverage.

However, there are some things running the other way, particularly things like lower STIP, short-term incentive compensation. We also have lapped a lot of the material and freight cost increases from last year, so that does help from an overall margin perspective.

And then we are seeing, obviously, last year, we did have some negative in Cat Financial in particular as well. And then obviously, as you get to PPS, you're also going to see the benefit of lower share count. So all of those factors are weighing in as we think about the fourth quarter.

I would not read through fourth quarter margins as being our likely margin structure as we move into 2020. We would expect normal seasonable patterns to happen in 2020. Obviously it then just depends on what the volume is and how that volume throughput flows through into variable margins..

Well, that was sort of the spirit of the question. I mean the fourth quarter is usually low, as someone is taking a little bigger hit, as we said, of the inventory reduction.

Because the math I'm running, even if sales are down 5% next year, even if the margins stay that low, you're still run-rating $9.60, all right? That's assuming share repo and everything else.

So just trying to get a sense of we could all make our view of retail demand, but it just seems like, if that's the fourth quarter with that margin -- and you just answered my question. You don't think the margins should go lower than that. It is sort of trying to at least baseline this run rate earnings power.

I mean a lot could change, but I appreciate the answer. I just wanted to get a sense of how you view that fourth quarter margin, so okay..

Your next question is coming from Joel Tiss from BMO Capital Markets..

I just wondered around pricing.

Is sort of the bulk of the pricing -- or any color you can give us on is that coming from new products and features? Or is that just coming more from raw material pass-throughs? And any sort of setup into 2020, how you're looking at pricing potential for 2020?.

Yes. So obviously, as I've mentioned, we did see the rate of price moderate in Q3. That was as effectively we've lapped the price increase that happened in the midyear as well, and obviously we are now working through the end -- the "beginning of the year" price increase.

We have -- given that it gave us an indication of price increases in 2020, we expect price to be much more moderate in 2020. Obviously, it's depending -- or that will also then depend on the competitive environment as well as we move into 2020, and we'll provide a little bit more feel of that when we get to our guidance in January..

And no color around the is it more from new features and new products.

Or is it just raw material related?.

Yes. I mean we tend -- that price that we recognize on the way we give you is -- tends to be around real price increases rather than actually mix increases. So obviously mix will go, which -- with new product would tend to go into the mix bucket when we look at mix and volume rather than price..

Okay. And then just quick, for Jim.

Any pieces of the portfolio that you feel like over the next five years need to be beefed up; or things, as you're getting deeper into your operational excellence, that maybe wouldn't fit? And you don't have to name the pieces but just kind of just more of a structural question of changing the portfolio versus just returning cash..

Yes. We continually evaluate our portfolio. We're always thinking about resource allocation. That's, of course, part of the O&E Model.

We've talked a lot about our intent to continue to invest in services to grow the aftermarket because that represents the best opportunity for future profitable growth for both us and our dealers, but in terms of changing the portfolio, we're always evaluating what potential changes we could make to drive more shareholder value..

Your next question is coming from Courtney Yakavonis from Morgan Stanley..

Just wanted to go back to the dealer inventory destock expected in the fourth quarter. It seems like most of it was coming from resources and then also from APAC construction this quarter.

Can you just comment on how much of that will be coming additionally from those regions versus North America construction and whether you're also expecting a decent amount there? And then just back on resources, where you called out the softer demand in nonresi construction and quarry and aggregate versus thermal coal prices, can you just help us understand how big of a factor each of those was; and just again as we're thinking about the fourth quarter and into 2020, how big of a drag this can still be; or whether you're expecting a replacement cycle to offset that?.

Yes. So let me start on the dealer inventory, Courtney. The -- first, fourth quarter expectations are that most of the dealer inventory reduction will come in North America, which will impact North American revenue -- sales and revenues in Q4.

We actually do expect an inventory build again in -- particularly in China, in Q4, and that's partly in recognition of the fact that there's an early Chinese New Year. And obviously that means the selling season starts earlier in China next year. So that will be a factor as we move into Q4.

With regards to your question on ROI, I think the -- our view is overall mining probably will be -- remain positive for the year. If you look at mining CapEx and expectations of mining CapEx, that remains positive, so our expectation is that will effectively reflect through.

If we look at things like power fleet, it's at the lowest level since we've ever been recording it which is since 2013. So there is latent demand there.

And as we said, we do think obviously miners are being cautious on their capital investments, but there is the demand and replacement cycle that is needed at some stage, particularly on large mining trucks. I will just remind you that is only a portion of our RI business.

I know it's often what people tend to use as the sort of marker but relatively small. I think obviously, as far as nonresi construction is concerned, our expectation probably is relatively that will be a drag, particularly in Q4, as effectively -- particularly that is the area where there's still more inventory to come out..

If I can just add a couple comments, Courtney, on mining. So again, we believe we're in the early stages of our multiyear recovery in mining. Just given the economic turmoil going on, our mining customers are being cautious, and so they are hesitant to pull the trigger on new equipment, although again we're seeing increased sales.

So we're seeing improvement in that business. One thing is -- to also keep in mind is that, when miners sometimes delay, that creates opportunities for us for rebuilds and parts. So that isn't all negative either. So again, it's an opportunity either way..

Would you characterize aftermarket as still being stronger than you would have expected otherwise?.

I say that it continues to be strong. It's how I would characterize it. It continues to be strong, about as we expected..

Your next question is coming from Ross Gilardi from Bank of America Merrill Lynch..

I just want to ask on the dividend, in committing to a high single-digit increase over the next four years. I mean, if you apply a 7% to 9% increase, your dividend is $5.40 to $5.80 in four years. I would assume you plan on covering the dividend with earnings internally even at the trough of the cycle.

And if that's the case, it would seem like you're implying at least $6 of trough earnings, so I just wanted -- I was hoping you could just comment on the thought process.

And in your mind, is there some type of minimum earnings payout ratio that you're assuming at the trough of the cycle in making that commitment to raise the dividend at that level given obviously you have no visibility on what's going to happen in the next 3 or 4 years?.

Ross, so if you remember, when we -- at Investor Day, we actually talked about it in terms of cash coverage rather than actually in earnings coverage.

And actually in cash coverage, even when our expectations are of the low cycle, is that we would expect to actually pay out no more than 50% to 60% of free cash flow in dividends even in the low end of the cash flow cycle. Obviously cash is slightly different.

If you plot cash against earnings per share, there are obviously differences in the way because obviously, if you are in a downward cycle from a revenues perspective, obviously sometimes that actually is positive from a cash flow perspective if you are reducing working capital through that period of time.

So there are puts and takes as to why you can't correlate it exactly to EPS, but it does reflect our confidence. So obviously we do have -- expect both cash flows and operating margins to be positive and to reflect our Investor Day targets through all parts of the cycle as we move forward..

So in saying that, Andrew, just to do the math for everybody, I mean, it sounds like in your view you think you're going to do at least $10 of free cash flow at the bottom of the cycle..

Obviously we said 4 billion to 6 -- I think it's $4 billion to $6 billion was our range that we talked about, on Investor Day, of cash flow. So obviously, yes, you can work that back through in the math..

I mean there was a lot of focus on this 6 to 9 months ago. If you look at the data, your share seems to have stabilized in recent months.

Is that correct? And how has that been the case? Have you had to match the competition with lower pricing? Or is it more new product- and innovation-driven?.

So this is Jim. So market share in any area of the world tends to be -- is always fluid and dynamic. We've talked previously about the fact that we are introducing new products in China, our GC product line. Our dealers continue to build up their capability with better coverage, so it's a whole variety of issues.

And again it's a very dynamic situation, but we're confident in our ability to compete in China long term. And we demonstrate the ability to do that, but there will be fluctuations on a short-term basis up or down. That's just part of the deal..

Your next question is coming from Noah Kaye from Oppenheimer..

Jim, you mentioned progress this quarter with respect to shortening product lead times.

Can you provide some more color around that? I guess, particularly, what's been accomplished internally versus a function of easing pressure from some of these inventory reductions? What have you actually accomplished in terms of making the supply chain and production more nimble?.

Yes. It's a combination of a variety of factors. We had discussed in previous calls that with the sharp increase in volume in 2017 and 2018 many of our suppliers struggled to allow us to help the retailers we provide to given that period of rapidly increasing demand.

So there's been improvement in the supply base because, as I mentioned in my initial remarks, we've also been very focused on becoming more efficient within our factories, reducing lead times, applying lean. So really it's a combination of all those factors..

And then on mining, maybe a question about how your customers are viewing autonomy relative to other CapEx priorities. You mentioned the retrofit offering. You've announced several greenfield projects.

We did see a case recently, I believe, where one of your large mining customers was considering going autonomous but then decided to overhaul its existing fleet and focus on productivity.

So I guess the question is, is that the trend or the exception? Is CapEx discipline generally holding back around a reduction of autonomous haulage? Or does this 30% productivity improvement from autonomy provide enough of a step change in fleet profitability that it would actually drive companies to replace or retrofit fleets earlier than typical?.

We've seen a lot of interest and activity in autonomy. I think, if you look at that 30% productivity increase, it really can be a game changer for many of our customers. Obviously every customer is in a very different situation. They could be a coal customer. They could be as in a variety of commodities.

So customers make decisions based on their particular financial situation, but we're very, very pleased at the adoption rates that we're seeing in autonomy in the last year or so. You mentioned the greenfield projects. Again, we do believe it's a game changer, and we're very bullish about the outlook for that product capability..

Your next question is coming from Ann Duignan for JPMorgan Securities..

Yes. Maybe you could address the comments you made earlier about Asia Pacific sales. I think you said sales outside of China were weaker than expected. If you could expand on that.

And then what specifically are you seeing in China in terms of end market demand and the fundamentals? Any green shoots in that region?.

Ann, starting then with your last question first. So on China, the industry, as you know, for us is mostly hydraulic excavators 10 tonnes above. And the industry continues to be strong. So given the fact that that's the majority of our market in China, we have not seen a decline there. So that's a positive.

As I did mention earlier, outside of China and Japan we have seen some weakness in construction over the last few months..

Where specifically?.

Country-wise, I think it's pretty well dispersed over the Asia region outside of those 2 countries..

And Ann, just remember most of our revenues in those markets are basically China and Japan. So that is the bulk. I mean these other markets tend to be relatively small compared to China and Japan..

Okay. And my follow-up is you've got 130 days of inventories on hand as of the end of Q3.

Where would you expect inventories to end at year-end? And is your assumption at this point that end market demand is flat going into next year? I mean, what are the downside risks that we'll go into next year having to underproduce retails? Are you comfortable that you'll have rightsized your own inventories by year-end?.

Yes. So Ann, as we look out -- I mean, are you talking about CAT inventories or dealer inventories? I think you're taking about CAT [indiscernible] -- yes, yes..

CAT inventories based on then there's 130 days versus 119 a year ago..

Yes. So the rise in that -- obviously there is a lag between, actually as we slow production down, ordering components and so forth before it actually flows all through in today's sales. So obviously you're also reflecting based on days sales, which are also impacted by things like dealer inventory as well.

So that has some impacts unless you've -- you take them into account, but the bigger -- we are looking obviously inventory.

We do normally expect a normal seasonal pattern, which is actually inventories to reduce in Q4, but obviously as we talk at the moment, obviously we're looking at actually reducing material purchases to reflect the production declines we've spoken about. But that should rightsize itself.

As we move into 2020, we would expect to be in a pretty normal position..

Your next question is coming from Stanley Elliott from Stifel..

Quick question. Is there a way to quantify where you will finish 2019 in terms of the service sales versus the 2016 -- 2026 targets? And then also, you've done a nice job of developing a lot of this in house.

Is this something the path will continue forward, or is there something that you'll need to look outside the organization with M&A?.

We utilize outside parties. We are certainly beefing up our internal capabilities as well and particularly in the area of digital, but we're open to M&A. So we've start to think about resource allocation and think about how we intend to grow those areas that are most profitable.

Certainly we're open to that as well and one of those things we continue to evaluate..

Your next question is coming from Timothy Thein from Citi..

So the question is on orders and relative to the guidance that you've provided in terms of what you think end user demand and dealer inventories do in the fourth quarter. I'm curious how you think orders play into this.

And presumably you have maybe a bit less year-end budget flush than prior years, but just curious to get your thoughts as to how that plays out. And help us in terms of think about a range in terms of where year-end backlog may end..

Yes. So as we look out, obviously if you look where we are on backlog at the moment, it's impacted by a number of factors. One of them, obviously primary, is dealer's expectations of inventory reductions. It depends on where we end.

Obviously the big unknown factor is what is dealer's expectations of future growth going to be at the end of the year because that will impact their order pattern in Q4. So it really is a function of that.

As it stands at the moment, you would -- obviously the backlog decline in Q3 reflects a lot of the dealer desire to reduce their inventories, which will -- obviously the $0.9 billion. It depends whether they decide whether they would like to reduce inventories further in 2020, and then at this point in time, we just don't knows about that.

That's too early for us to tell..

Your next question is coming from Jerry Revich from Goldman Sachs..

All right, yes. Jim, I'm wondering if you can talk about what you folks are seeing in terms of the forward-looking parts demand indicators for your resource business as we've seen the useful life assumptions get pushed out by the miners. Presumably you have pretty good visibility on major rebuilds coming up.

Is 2020 a major inflection? And then you've spoke about the moving pieces in the market given the economic uncertainty.

Can you just talk about when -- based on the project and work that you anticipate, when do you expect resources will go back to growing the backlog as we, hopefully, see an acceleration towards more replacement-type levels of demand?.

Yes. Thank you. So for Resource Industries, our rebuild activity and parts activity has been strong and we expect that to continue to be strong, so we're not looking forward to a -- significantly an increase or a decrease. It's been strong and we think that will continue..

Yes. And as far as actually when do we expect the -- [indiscernible] point about what is the timing of any bounce, the -- on sort of particularly around power fleet and replacement, I think it's really, really difficult for us to see when that will happen, Jerry, to put a particular time line to it.

I think obviously, based on all the stats that we're looking at, we do expect it to happen. It's just a matter of time of that, and that really depends on miners' views of their outlook. And obviously, as we said, everything apart from coal is investible, so it's not the investment decision. It's probably their view of the outlook in particular..

Well, just given the history of the last 10 years, I believe that miners will continue to be cautious here.

So again I think it'll be a multiyear increase, so it'll gradually get better, as opposed -- we probably won't see the volatility that we've seen in the past, either up or down, which frankly will be a positive thing for us and the industry, to have it be more steady, more of a steady increase over several years than -- again, than volatility we've seen previously..

Okay. And Andrew, on the free cash flow number at the trough that you spoke about at the Analyst Day, what level of working capital contribution are you folks embedding? I think, in prior cycles, it's generally been $1.5 billion to $2 billion of positive free cash flow, as inventories have come down.

Is that what you're contemplating relative to that trough number?.

I said $4 billion to $6 billion. It's actually $4 billion to $8 billion. So I actually underestimated the top end, but actually what we do -- I mean we use similar working capital assumptions, as we've seen in previous cycles, so it actually isn't working capital which was the main benefit to cash flow. It's the lack of restructuring costs.

It's lower CapEx and also the fact that obviously the structural costs will be taken out. So the improved margins. Those are the three biggest drivers of improved cash flow..

Your next question is coming from Mig Dobre from Robert W. Baird..

Just looking to clarify some earlier comments on the dealer inventory destock. So you're exiting the year with an additional $500 million of inventory at dealer level year-over-year.

So if we're assuming that retail sales are flat in 2020, would that allow you to produce the retail demand? Or is there additional destocking that would be needed?.

So Mig, I need just remind you again it's dealers who make those decisions about their inventory levels. It's not us. Availability and all those things can play a part.

Ultimately, at the end of the day, given the lead time for production and the fact that dealers don't want to miss revenues, they'll make decisions based on that and what their expectations are of the future.

Obviously, when we look at it, we believe that the level of dealer inventory is within the range of probability that we would expect, the comfort level that we've talked about, 3 to 4 months. And it stays in that range. Obviously there can be movements within that range which aren't necessarily we're going to have control. They are dealer decisions..

And again just given the external environment, the uncertainty in the dynamic environment we're in, we believe we're well positioned regardless of what happens positive or negative in 2020. We have -- we shortened our lead times.

We have an appropriate level of -- our dealers have set up an appropriate level of inventory, so we think we're prepared either way..

Okay. Understood. And then my follow-up is really on the levers that you have to manage your costs as we're seeing some volume fluctuation here. I mean, if I'm looking at incentive comp, it came down modestly from last quarter.

As we look going forward, how do you think about any restructuring or any other actions that you might have to undertake if indeed volumes remain weak? Or do you feel like at this point you've got enough flexibility within your cost structure to be able to handle that without any meaningful moves?.

So we continually evaluate our cost structure. There's a number of things that we're working on. Andrew mentioned some things that we're doing looking at our back-office costs, if you will. So we're -- we have some things, projects that we started there. We're looking at material costs.

That's certainly a big lever for us; and that's one of the things that we're working on, both direct and indirect costs. We continue to look at ways to become more efficient. So again, we're continually doing that. We won't make a call as to whether or not we'll have major restructuring or not.

Again, we'll see what the market brings to us over the next few months. Either way, we're ready. We'll be ready to respond..

Okay. And that's our last question.

Jim?.

Well, thank you for your questions. We really appreciate your interest. We'll continue to execute our strategy, with a focus on services, expanding offerings and operational excellence, to deliver long-term profitable growth. And we look forward to chatting with you again next quarter. Thank you..

Thanks, Jim. Thanks, Andrew and everyone who joined us on the call today. Before we close, let me point out Slide 16, where we're providing our preliminary 2020 earnings dates. If you have any questions, please reach out to Rob or me. You can reach Rob at rengel_rob@cat.com. And I'm at driscoll_jennifer@cat.com.

Our general phone number for investor relations is 309-675-4549. And now let me ask Catherine, our operator, to conclude the call..

Thank you. Ladies and gentlemen, this does conclude today's conference call. You may disconnect your phone lines at this time, and have a wonderful day. Thank you for your participation..