4Q: GM | The "T" Word

GM reported a strong year, and put out guidance a little lower than last year’s results, but basically in line. The stock fell 10% regardless, as management said the “T” word. Guidance doesn’t include the impact of potential Trump tariffs.

Market Power In Trucks - Our Thesis

The core of our thesis is that the market often paints all automotive companies with a similar brush, and believes they’re equally bad businesses and that they all face the same industry threats to the same degree.

Most — but not all! — automotive companies are bad businesses. I think there’s a continuum, with the likes of Ferrari on one end, and and a large number of guys like Mitsubishi and the dying Chinese automakers on the other end.

GM is somewhere on the good end because ~2/3rds of profit comes from selling pick-up trucks, large utility vehicles, and some luxury vehicles like Cadillacs. This mostly comes from the unique structure of the North American market, where most of the world’s full-size, high-end pick-up trucks are sold. Generally, these market segments can have higher customer loyalty and lower customer price sensitivity. For example, truck owners have tattoos of their trucks or of the brand far more often than the typical car owner. This bears out in the economics of the business. The margins on a pick-up truck are ~15%, which is about the same as a Porsche, though the price point is a little lower at ~$45K-100K. Selling cars, by contrast, is a 3-10% margin business at <$45K prices. Clearly, pick-up trucks have pricing power. The prices have slowly increased over time, even during times of elevated competition in other vehicle segments, where the prices of those cars weren’t going up. Multiply the margins by the price and you will also see that building one truck on the manufacturing line has the same unit profit potential as building ~4 cars.

This said, be aware GM has underperformed the market since we bought @ $37 in 2019, and is one of our worst-performing names the last 6 years.

Even Okay Businesses Can Go Poorly

That said, Ben Graham always said you are neither right nor wrong based on what the market does. One, GM’s free cash flow generation has been about in line with our base case during the good years, and exceeded our downside case. We felt the stock was worth at worst $30-33 and at most $60-100, so we took the bet at $37, knowing full well it had very high uncertainty and is a cyclical business whose volumes we can’t predict over 5 years. Consider this is a $50 billion company capable of earning >$10 billion annually. I believed free cash flow would grow more than I was modeling base-case because of the truck business’ ability to raise prices whenever it liked. Revenue has grown more than I thought it would, but the EV (electric vehicle) transition has consumed more capex and R&D than I thought it would.

EVs are currently loss-making because of the high battery pack costs. Think about it like this: the big difference between the cost to make a combustion vehicle and an EV is the powertrain. You remove an engine and a transmission, but you put in 1-4 electric motors and a battery pack. The big ones are the combustion engine, which is say $5,000, and the battery pack, which is say $10,000. Consumers aren’t willing to pay more for EVs. So making these cars costs about $5,000 more per car and the prices are the same. These were 5-10% margin cars before, at say $40,000 (between cars, crossovers, and small/mid SUVs, say). So you do the math and you can see the math doesn’t work. The battery pack costs are coming down with time. There are two pieces: the packs and the cells. I’ve seen engineers say the packs are quite overdesigned, and there’s lots of room for costs to come down. The cells are more of a technology thing, where the companies have to figure out more ways to use cheaper materials while still increasing the amount of chemical energy stored per unit of battery weight. This is happening. Finally, as EVs become more popular, there will be more fixed cost leverage as more and more EVs are sold and things like the R&D and the overhead costs stay the same. Not to mention more purchasing volume with the battery makers like LG Chemical.

Battery costs have been falling in recent years, and I expect that to continue. In the most recent quarter, GM’s electric vehicle portfolio is now “variable profit positive” as a whole. That means the average EV sells for more than it costs to make it, including stuff like labor and parts, but excluding stuff like the capital cost of the manufacturing plant, and the R&D to design one. The cost has to come down more for this business to earn an appropriate return on capital. But Mary Barra has generally been hitting her targets, so I’ll let her keep going.

Nonetheless, these are investment dollars GM has been sinking into the ground which have not been paying off for several years so far (but which should eventually), and this has been a drag on cash flow and on the business’ value. I frankly believed the cost to transition would be less impactful than it has been.

Offsetting this, the company significantly cut costs in other parts of the business, like R&D on combustion engine cars, and on overhead expenses. That was a “good guy” vs. our expectations.

There have been other issues. First, I believed the Chinese business, which legally has to operate as a 50:50 JV with a local Chinese automaker, would be doing $1-2 billion annually for GM over time, and that a good chunk of this cash flow was free from excessive industry competition because it came from selling Cadillacs and such. Not just the much-more-competitive Wuling cars. That JV is now doing $1 billion or less, as I underestimated how many new Chinese entrants would flood the market: something like ~100 automakers were formed in the last few years, and the market is now suicidally competitive.

Second, some of GM’s cash flow was ultimately misallocated, though the bet seemed good ex-ante. GM bought Cruise, a robotaxi business in 2016. Annual costs soon rose to ~$1.5-2 billion by 2018-2019, and ever since. A lot of this was funded by selling shares or intellectual property rights in the company so that GM didn’t need to foot the entire bill. For example, technology rights were sold to Honda for ~$1b+, to use in Japan. I liked the way Mary Barra was managing this bet. The company was betting about 10% of free cash flow on the idea, while finding smart ways to raise capital like the Honda deal. The guy who ran Cruise was incentivized to set it up to be spun-off, but GM changed its mind after a few years. Barra then shuttered Cruise this year, even though I think the market would have potentially liked an AI robotaxi company that was just starting to generate revenue. There were even reasons to believe the robotaxis would have better unit economics than human-driven cars. Alas, this cost GM $10 billion and there won’t be anything to show for it. It is what it is.

Why do you still own this crap, Chris?

We still own the stock because our core thesis is still true and because the EV business is the only painful part left. Yet it’s getting more profitable as EVs scale up and replace combustion cars. GM’s place in that market is fine: its EV market share is approaching its combustion car market share as it rolls out a full suite of products.

GM’s also generally been smart with its expense management. E.g., years ago, they designed only one 12-cell battery pack which could fit into any vehicle, where you then just inserted more/fewer cells based on the vehicle’s class. A car gets a pack with only 6 cells installed. An SUV gets the same pack, with 12 cells. A pick-up truck gets two packs stacked on top of each other. So GM only designs one battery system across all its products, saving loads of R&D. They also get the purchasing and manufacturing scale from producing and buying only one kind of battery from suppliers. CEO Mary Barra is a pretty good operator, and her team often come up with smart ideas like these while managing down the company’s costs.

The transition’s still weighing on margins. Revenues have grown 9%/yr each of the last 2 years, while margins in key areas like North America are lagging their potential, at ~9.2% vs. ~10%+ where they should be. Peak free cash flow has grown from my initial $10 billion estimate 5 years ago to ~$12 billion¹. Cash flow is below where it could be. As the EV economics improve further, I think margins and free cash flow have room to increase. We also aren’t at peak auto sales. US auto sales were running at a ~15-16 million annualized rate in most of 2024, whereas we think the US can do ~17-17.5 million at peak.

Should these tailwinds materialize, margins should hit 10-11% with free cash flow well over $10 billion, and this should not be a $50-60 billion company.

I pointed out some wasted uses of cash above. However, the majority of free cash flow is not being wasted. In the year, GM bought back nearly $8 billion in stock, paid $500 million dividends, built ~$2 billion in cash, and paid ~$750 million in debt. With money-losers like Cruise now closed², more cash flow will go toward share buybacks.

The core business is still attractive as well. The company’s done a good job creating more desirable vehicles than in the past and is focused on pricing and customer value, not volume market share. Its dealers are carrying below-average inventory levels, at ~53 days vs. an industry average closer to 75 days. This is also lower than GM’s typical history (if you exclude the tight market just after COVID). That means the cars are selling well and the company is managing production tightly vs. demand.

Furthermore, its sales incentives have been below average for a while now. Like many consumer products companies, automakers use discounts to move product. The more demand for your current products, the less you have to put them on sale to help dealerships move inventory. I no longer have access to this data and try to find anecdotes, but GM was kind enough to provide a chart this quarter:

Source: GM 4Q Earnings Presentation

These mean GM isn’t lying to us when they say they’ve done a good job creating more desirable products. A decade ago, they were still poor quality in the mind of many consumers.

In the context of the above facts, valuation matters. We paid $50 billion for this company. Outside of the COVID recession, it was, and is, generating $10 billion in free cash, and growing. So we had a good margin of safety and I would argue this is a big reason why we are sitting on a gain today and not a loss. It’s also why we still own the stock despite the fact I think it’s a highly imperfect business/industry.

Still, GM is not a “forever” investment because it’s really not a wonderful business. I think it’s quite undervalued and worth more like $80-120 per share, but with high uncertainty, and I have been slowly replacing it with better ideas when I find them, such as Fortrea. There are a lot of other businesses with better economics, and in better-structured industries with more tailwinds. When I find ones I think are more undervalued, we will make the rational choice to deploy capital there instead.

The “T” Word

There’s confusion around how tariffs work. There’s also confusion and uncertainty around how tariffs will impact GM and the industry.

How tariffs work is simple. If something has an import tariff, then when the commodity, the component, or the product comes into the country at the border, the importing company pays a tariff to its own government. It’s usually a percentage of the product’s price. If car parts come into the US and GM is going to use them to make cars in the US, GM pays the US government a tariff. What then happens, is the buyers like GM and its competitors will adjust prices through competitive forces. They may try to squeeze the supplier on cost, meaning they renegotiate their supply contracts such that the supplier bears some of the cost of the tariff. More often, the importers end up having to raise prices on the products they sell, which means that the tariff is absorbed by the end-buyers, whether they are businesses or consumers. For example, if most cars of a certain class are imported by most of the competing automakers in the US, then they will be able to raise their vehicle prices to the end-consumers, and there won’t be a big competitive reason for them not to pass through their cost.

Now, the effects tariffs will have on GM, the auto industry, and other industries is uncertain. And this is the complicated part.

If a BMW is imported from Germany and there’s a tariff on cars from Germany, ultimately BMW, its dealers, and consumers will end up sharing the tariff. It will probably mostly be the consumer, since many German premium cars are imported.

By contrast, If GM makes a Cadillac in the US, its dealers don’t have to work with GM on changing their pricing strategy and such to absorb this cost. This makes Cadillac cars more price-competitive than German-made luxury cars, and GM would get some combination of better profit margins per unit (because it can somewhat follow German car prices upward, but its costs don’t increase) or more market share if it holds down prices.

In other areas, such as vehicles produced in Mexico and Canada then sold into the US, GM probably ends up less profitable.

In the US today, most cars are built in “transplants”, the industry’s word for local manufacturing plants. E.g., back in the 80s & 90s, Toyota made cars in Japan then imported them into the US. As the company gained market share, automation increased, and Japanese wages rose, it eventually made sense to add capacity in the US. This eliminated the cost of freight, which is ~10% the cost of cars imported overseas (they go on rail, then they go on a ship, then they go on rail again, then they go onto a truck, then they end up at the dealership lot). Today, Toyota has plants in Alabama, and other foreign automakers have many plants in the US. There’s a huge global supply chain of imported components, but the cars themselves are mostly assembled in the US. They all source components from the similar suppliers like Continental and Faurecia and Magna and such, so for large automakers they should all end up impacted similarly, and thus be forced/able to raise prices.

Much of the remaining supply is assembled in Mexico and Canada, where the previous Trump administration already renegotiated NAFTA and our countries agreed on the USMCA instead. Very little changed when it came to auto industry requirements. The agreement demanded that a little more of the vehicle’s content be made in these 3 countries than was before, and required that workers make at least USD 16/hr, which reduced wage cost differences to incent new production to move to the US from Mexico/Canada. Except the market isn’t growing and the plants are already located where they are, so there was no point paying to move them unless it’s economically worthwhile. Which doesn’t seem to have been the case. So, not much really happened after a lot of talking and renegotiation. Something similar might happen this time around, because…

Keep in mind, we don’t even know any of this will happen: I have been through one Trump administration already. Trump is a one-trick pony. What he does is make big threats in order to get people to come to the table. Then, a more sensible plan is negotiated and comes out as a result. In many cases, only superficial changes are made so that Trump can claim a win.

The US IRA (Inflation Reduction Act, created under the Biden administration) is also a small problem if the Trump administration removes it. I don’t think this is politically palatable. The IRA created incentives for US automakers and suppliers to invest in electric vehicles (EVs) mainly through battery subsidies. GM and competitors (but not so much Tesla) rely on these subsidies to eke out a profit on EVs. If the Trump administration removes the IRA, the whole EV industry is in a lot more trouble for a few years until they can further reduce battery costs. HOWEVER, the EV incentives require that a lot of the battery pack is produced in the US. Firms like LG Chemical invested in US capacity, such as by using joint ventures with GM. That created American jobs. It would be strange for Trump to can these incentives, which would likely kill American jobs and American manufacturing capacity, which is exactly what he’s trying to bring back with his “America First” policy. I don’t think this part of the IRA will be canned.

Concluding

So, in all, GM still looks cheap to us, but clearly there’s a load of uncertainty at the moment. Without going through the model, I’ll say that at ~5x the company’s earning potential, it looks like a lot of the potential margin impacts are already priced in because of this uncertainty. Meanwhile, there’s a lot of reason to believe the Trump administration won’t kill this industry’s profits with tariffs or changes to the IRA, and at the same time, there’s a lot of potential for GM’s per-average-car economics to improve from here as EV costs fall and as US volumes increase.

Chris

1. GM made $14 billion in 2024, but some of this was from working capital.

2. What’s left of Cruise will be used to continue doing R&D for GM’s advanced cruise control stuff, which will take off 50bps of margin in GM North America as it continues spending $500 million annually, but down from $1-2 billion as the standalone Cruise segment.