Andrew Bonfield on Caterpillar’s success through value management (Create More Value podcast (Ep 30)

Thank you, Greg, for having me.

Jim Umpleby, the new CEO who came in at the beginning of 2017, introduced Caterpillar's new strategy. He very simply wanted a strategy that everybody could understand, so he put it on a single page. At the heart of that is profitable growth. Profitable growth is an interesting concept for many companies because it's neither grow margins or grow top line—actually he wanted to grow both. So both organic growth and margin expansion was at the heart of that strategy. What surrounds the strategy are things like the company's values, the code of conduct, and the way the company has operated for many years. Effectively there are four pillars—there were three originally and one has been added. Those four pillars are expanded offerings, operational excellence, services, and the new one, which is sustainability. Those all form part of and are covered by the operational execution model. The high-level result has been really, really strong—not only stock performance but also really strong financial operating performance. Since I've been here in 2018, profitable share has nearly doubled. Free cash flow is very strong; last year we achieved the best-ever result from a free cash flow perspective, $10 billion. What we've done is, using the O& model, effectively focus on what we call OPAC, which is operating profit after capital charge. It is an EVA-type approach, an economic value added approach, and that really strongly correlates, we believe, to cash flow, which obviously most strongly relates to total shareholder return. That's been the driver behind being successful.

The operational execution model is the heart of what we do. It is the focus for that long-term profitable growth. Effectively, it's a disciplined investment framework. It goes down to by market, by application, by product—really understanding what is happening and how value is being created in the individual business units and in the individual product lines, by product, by market. Then we can actually work out where we need to invest and where we need to fix as part of that strategy, to make sure we're profitable and driving OPAC across the whole organization.

Effectively, we're trying to use facts and data as a basis for making those determinations. So yes, there is obviously some element of allocation in that, which sometimes creates a little bit of a challenge. But the real focus is to make sure we're looking at everything across the level playing field, across the whole organization, and that then enables you to make better resource allocation decisions. The heart of the O& is about where do we allocate our resources behind the greatest opportunities for long-term profitable growth, and where do we take resources away from parts of the business. Either through thinking about restructuring them, or if we can't find a way to restructure to profitability, to actually exit those businesses. This is done on an annual basis, on an ongoing basis. We do our O& reviews across the year for each of our divisions, and each of those divisions will come and present with a level of detail that we can look through, also making sure we understand the market itself. Part of the understanding is where our competitive position is relative to others in that market and how that impacts profitability as well.

Effectively, like any manufacturing company, we get the benefit of operating leverage—so you spread your fixed costs over a greater base. Where you get volume benefit, obviously that creates an opportunity to grow OPAC faster. That's really as simple as it is, as long as you don't have to continue to build extra capacity. That's always a reminder. One of the things Caterpillar would have done historically would have been to chase the business cycle. As demand continued to grow, the business units would find it quite difficult to meet production need, and therefore they would say, 'We need more capacity,' and actually overbuild capacity because they forgot that capacity has a memory. That's the benefit of having a capital charge, which effectively says every dollar you invest, you have to get a return on that dollar—otherwise it's excess, and then that's a problem. Finally, one of the things we've found is we've actually reduced the amount of CapEx needed in the business as we've gone up. We've closed a lot of old capacity down, and at the same time avoided restructuring charges because you don't have to shutter that capacity in a period of time.

I would say to you, there are times when we have arguments about allocation, so I wouldn't say we do a perfect job at it. But I think we do a better job because it's been a more consistent process and something we've looked at over a number of years. One of the other things we've done is we've actually delegated decision rights down into the business units themselves, and that delegation actually enables them to have a much greater control over what that cost base is. So even if it's an allocated cost, like network costs, they actually understand how that works within their individual person. They don't have full decision rights, but they will have a greater degree of autonomy—for example, of deciding what applications are going to be put on their systems. That creates that linkage between decisions they're making and costs that they bear. Where you get into more of a difficult discussion is where there doesn't seem to be a real rationale for that. That's the bit where everybody has the challenge. When you first worked back in the SmithKline days, you remember that conversation we had around R&D allocations, which were much more challenging, because in a pharmaceutical business the margins are much higher, it didn't really make a difference anyway—so an allocation was almost irrelevant. Manufacturing is a different environment. It lends itself because it's much more capital-intensive and then also lower margin, so people want to understand how they make their business profit. Some of the allocations—we don't have debate, but they understand those allocations are needed to make them understand how they fit in the whole of the enterprise and have that enterprise do in that strategy.

Actually, one of the things that's interesting with the R&D is that the percentage of revenues does fluctuate a little bit, obviously, because as revenues grow you have a little bit more volatility in revenues versus a consumer products or pharma company. But actually, as far as absolute dollars are concerned, it's pretty much a consistent, growing number within the P&L. One of the things we've done is we've also made sure that we continue to look at how to make those R&D dollars more effective. For example, we had a lot of our engineering based in very high-cost jurisdictions, and actually where we added a little bit more incremental engineering—particularly given the issue about STEM and the number of engineers qualifying in, for example, the US and Europe—we're looking at places like India and China for some of our engineering expertise, because not only is it cheaper, but there's also greater volume and you get more bang for your buck. So we have not taken investment decisions off the table. When we do an investment proposal, we obviously look at the NPV and things everybody knows, with an IRR and so forth. But there are certain times where it doesn't lend itself to that sort of rigid process. Let me give you an example: e-commerce. It's very obvious to all of us that people would like to be able to transact on their cell phone. If I went through my normal process—what we call our greenbooks—for an investment decision for an e-commerce investment, I would probably be negotiating with probably 120 entities across the organization and getting them to agree what they think the incremental sales volume would be. That was going to take forever and would have slowed us down. So what we said is: okay, what do we need to believe to achieve a return on invested capital above the cost of capital? What would we need to achieve in that case? And it became pretty easy to then back-solve the problem and say, 'Okay, I'm going to go through this negotiation phase, I'm going to go direct,' and we made that decision. The benefit of the O& model is it gives you a framework that is really important for most of your investment decisions, but at the same time enables you to make flexible decisions around certain things we've decided to do.

You referred to GPS tracking, connectivity—it's actually mostly done through cell phones, cellular data, but that's really important for us. First of all, it enables us to track machines, see what the machine is doing, enables us to look at what hours are on the machine, and enables us to work with our customers and tell them, 'Hey, hold on a second, you've done X number of hours, you probably need to think about some service associated with that machine,' whether it be undercarriage, or filters, or oil changes, repair, and so forth. So that gives us the data to identify the issue. At the same time, it also lets us know where the machine is and what conditions the machine's working in, so that we can better help and understand and actually better serve that customer. Ultimately, at the end of the day, what is the most important thing for most of our customers? It's uptime, and actually making sure their machine is operating when they need it to. Construction often is a phased process—there's a lot of time where, for example, you'll grade. You'll have the bulldozer going in and pushing the dirt, then you go in with the motor grader to make sure it's flat, then you go in with the soil compactor. At the same time, you need a wheel loader to come in and move some of the excess dirt off the site. You can't—that is a process, it works. So the thing that often happens is the motor grader is standing by waiting to do its job. Suddenly the motor grader is not available because of downtime. That's bad news for the customer because it means a delay in the job. So it's making sure you have that uptime, and that is the sort of thing again—connectivity. We charge a very minimal subscription to give certain data services back to customers, but ultimately we believe the benefit to us of connectivity is being able to help the customer be more successful, which enables them to want to buy more Caterpillar. That's the benefit, and also CAT services, which are obviously where we generate more profitable growth. So that's an organic growth opportunity for us, but it's one where we've taken the decision—we don't want to return on the subscription model, because if we did, again, that would be a lot more challenging. To get every customer to want to do that—the landscape gardener who's using his skid steer to do landscaping in your back garden isn't necessarily going to want all of those services, or not necessarily be willing to pay for all those services. But he gets the benefit of it by us enabling them to say, 'Hey, hold on a second, you need to do your oil change, you need to do your filter change.' All of that is part of that strategy, and again just part of how we use the O& model and then also allow the O& model to have the discipline to enable process to operate, but then enable us to focus on those things which are most important from the strategic perspective.

There's two parts to that. First of all, OPAC, as we say, operating profit after capital charge—we do allocate and have an asset base to each individual business unit, and we allocate a 13 and a third percent pre-tax cost of capital to that asset base, and that would include their working capital requirements as well. So fixed assets plus working capital, and we use that to calculate that capital charge for the entity they're operating in. Again, it's not a full economic value—we don't do things like capitalizing certain things, and we do have depreciation rates that are slightly different sometimes in different business units, some accelerated depreciation and so forth. But it gives a direction of travel and it's a consistent basis across all the entities, which does enable us to at least look and gives a proxy for that cash flow, free cash flow, which is obviously the key driver of the value that we create. We also have another measure we use, which is OPAC PV—OPAC PV is also the present value of future services revenues, because obviously like every other manufacturing business, services are an important part of the value you create, and that obviously is a revenue stream associated with that. Obviously the focus there is about capturing as much of that future revenue stream as possible, because that then creates more value for you. It's not the sort of razor blade model of selling the razor at a discount in order to sell the blades, but obviously services create value as part of that long-term stream that you've got. So part of that has to be part of the thought process, and we use that measure as well.

As regards where we don't use it, let me take another example. We made an investment a couple of years ago—started investing in electrification. We looked at people's sustainability goals, what people were talking about, and we thought we need to make sure we have battery machines available for customers—not to force them to take battery machines, but to have them available if they want them as part of their sustainability goals. Interestingly, take-up to date has been very, very low, but we didn't expect a high level of take-up. So that was an example where we said there's no point for us to try and estimate the future cash flows associated with the investment—we'll be here for the next 12 years trying to do the calculations. And mind you, we are an engineering company—that's the heart of Caterpillar. Engineers are terrific, but one of the problems with that is engineers and accountants together—you try to get people driving to perfection with an art plus a science. It's not quite the right way to go sometimes, Greg. What that does mean is we sort of circumvent that in those circumstances and say, look, there's no point trying to do an NPV calculation. We know that the market—we're going to have to move that way, we know customers are going to want to move that way. Obviously we'll give them the option, we'll create the option for them, and that creates the sort of discussion point around that, rather than trying to get into unit sell price per unit, what's the—when you're actually doing something from scratch effectively and have no idea what the real cost is going to be or what the price point is going to be when you're selling it.

So again, that was just part of that sort of approach that we sometimes take. It's a stage-by-stage approach, but obviously just looking at it in a way that says: okay, ultimately at the end of the day, what is our objective? Our objective is profitable growth. Also, at the end of the day, how do you grow your profit? Effectively, that is by meeting customer needs. So you need to execute your strategy to doing that. Is this part of my strategy? If it's not aligned to the strategy, then it's just a vanity project—you should be killing it. It's got to be aligned to those customer needs ultimately at the end of the day as part of that. That's why again just having the simple strategy on a page and the gear that we use is really, really important, because again it just enables everybody—you can always bring people back to that, even though it may be something which is outside the norm. If we could all build businesses on spreadsheets, we'd all be billionaires and we'd all be incredibly successful. The reality is the world creates—you need to have some judgment factors. It's the same as what's going to happen with AI. Judgment is still going to be an important element of it. AI can give you the base data, but ultimately at the end of the day you still need human judgment in that, and make good decisions around that as best as you can.

I've seen that used particularly in pharma, particularly where you're using licensing products. That's a great way of doing it—what's the probability of it getting to market, what's the income stream. Your point though about back-solving—everything then becomes a back-solving exercise, and in order to achieve a 15% IRR, every single licensing deal I have, because everything at a stage has a thing, ended up with peak sales of $1.6 billion. Why was that? Because $1.6 billion sold the 15% IRR. So you have to be a little bit careful. Again, that's where the judgment comes in. If you give people a spreadsheet, they'll solve the problem. That's a little bit where the judgment comes in again about how you execute against those models. If you've got a spreadsheet which says here's my OPAC, here's my OPAC PV—this is a business where generating OPAC on original equipment is quite difficult, but you can get a very good services flow. You may actually still retain that business; again, that would be part of your decision-making process. However, if you've got a business which is requiring only OE to drive OPAC PV, that's going to come short at some stage, because ultimately demand requires field population, and if you're not providing services that customers need, ultimately at the end of the day—or your dealers—how does that flow into your long-term ability to grow that business? Because ultimately customer satisfaction is really, really key to how you grow businesses for the long term.

So again, all of that has to be—we always balance out. The optional model is one way we look at it—what's the optionality? The O& model gives me the 85, 90, 95% of decisions which can be dealt with in that way. I can look at it, I can make decisions, I can test the assumptions, and it's very easy. The 10% of the stuff is the more difficult things. That's where we have to use our intellect to try and actually work through the decision-making process. And that may be exiting the business, restructuring the business—what can be done? How does this look long term? What do we have to do to make this get into profitability? Do we ever see that being feasible? What is the history? So again, just asking those questions and probing. That's part of the power of the model. The success has been significant margin improvement versus our history, at the same time growing services revenues. We were $14 billion in 2016 in services revenues; we're now $23 billion last year in 2023. So that's a huge organic growth option for us, and much less volatile than OE, so it actually drives more stability and profitability, and that drives free cash flow. We've increased our free cash flow target—it used to say 4 to 8 billion, it's now 5 to 10. That again—the thing about CAT, and this is a bit that I think investors often misunderstand, and one of the things I've been talking to people about—is we throw up cash throughout a business cycle. So even in periods of time where demand's strong, we throw up cash; weak, we'll throw up cash, because we work well in working capital and stuff like that. It's a very strong cash-generating business, and cash flow is a lot less volatile.

What that's enabled us to do is actually go into the market, buy back more stock, increase the dividend. We've got a growing dividend at the same time, been a dividend aristocrat. At the same time, we've grown profitability, we're growing free cash flow, buying back more stock, and we've reduced the share count by 18%. So that's all things where the O& model has really helped, which ultimately has driven stock price. Since I joined in 2018, the stock price has tripled. Since Jim took over as CEO, which is when the O& model took its effect, it's quadrupled. That is part of—and that focus on shareholder value has been really, really important. Making sure—we're not perfect, no we're not, no company is. Anybody who tells you that—as you know—is not telling the truth. But again, it just gives us that framework to continue. There are occasional things where you sometimes come back and say the O& model may have driven certain behaviors which probably weren't right. So we have to always make sure we're challenging people both ways, so that they're not doing things just to cut at some small element of cost which actually cuts into the bone rather than actually cutting away some of the fleshy tissue on top. So it actually gets into the real art of business, and we continue to focus on that.

First of all, OPAC is a measure for compensation, so that's definitely in the proxy and it's talked about, and obviously people are aware of it. We don't discuss it much when we're talking to investors per se, but we talk about the O& model and how that has driven changes. The O& model's driven a lot of change in the way that we manage the business as well, not purely just by products and things like that. For example, inventory—we invest a lot in inventory. If you think about what happens when demand starts to grow, we are a dealer model, so dealers operate and we provide inventory to our dealers, dealers sell to the end customer, and they service the end customer. What happens is, as demand tends to surge, availability is still—we're still ramping up—and availability is here. So one of the things that used to happen is, factories start ramping up, dealers see availability still low, so they basically put in more orders because effectively what they're trying to do is catch up and make sure they get slots. Then we start catching up, and then they give us the cancellation rates immediately, particularly construction, because they don't need it and they've got availability now, so they need to go smaller. So what happens then is we've got CAPEX here and we then have to take down. One of the things through our sales and operating planning process—and again this comes back to effectively using OPAC and the O& model—we're now trying to manage that much tighter to avoid some of those ups and downs.

One of the challenges we have quite often with investors is: the level of inventory is increasing, so they say that's going to be a problem. No, it's increasing, but we're now keeping it to a much tighter range than where it was historically. And in the context of a group that's got $60 billion-plus of revenues, it's actually quite a small number. The amount of focus sometimes we get, particularly from those who are on the shorter end of the investor spectrum, is disproportionate. Again, just trying to explain some of that discipline, because ultimately that helps your cash flows, which helps drive shareholder value. Because what you don't want to do is overbuild, then you get stuck with a lot of—or your dealers get stuck with a lot of inventory, or you get stuck with a lot of finished goods inventory, and then that's the worst of all worlds, because your production costs become much, much greater. So all of that about managing is much more part of that thought process. And so it's again just trying to make sure we use all of those things in the way we actually make decisions, and then obviously explain to investors about it. Because ultimately at the end of the day what we say is, OPAC correlates to shareholder value, we believe. That correlation works really, really well. So again, just trying to keep all of those things, but obviously investors think about it in terms of the output rather than the input. OPAC, in a way, is an input measure which is probably more for management; the output measure is free cash flow, in a way, which is obviously the most important thing.

No, just—it's been really, really successful, and one of the things that gives me a lot of joy is actually seeing this operating. We tried this 20-odd years ago, back in my SmithKline days, and couldn't make it work. But to see an operation and to see it being so successful, and actually to see the results as a result of that, has been phenomenal. Being able to not only execute strategy but actually deliver the performance that Caterpillar has done—the team deserves a lot of credit. A lot of the building blocks, as I say, were in place before I joined, so it was a joy to come in and actually be part of it and see some of the success of some of the outputs. That's what you've been a part of.

Thanks, Greg.