How Government Incentives Are Shaping $WM’s Capital Decisions

Today, I’m deep into my analysis of $WM (Waste Management). On any given day, the company’s teams collect waste and recyclables from 21 million homes and businesses, operate fleets along set routes, and move materials to processing or disposal facilities.

As I incorporate insights from the latest Investor Day, I’m reminded how influential—and sometimes questionable—regulations can be from an economic perspective. Back in 2016, nearly a decade ago, I added a note on the CapEx (capital expenditure) section of the WM’s model: “Since 2005, organic growth has been mostly negative—therefore, I will assume no Growth CapEx; this means that recent CapEx is an excellent estimate of necessary Maintenance CapEx.”

That forecast primarily held. For years, most of Waste Management’s CapEx focused on maintaining existing operations. If you look at the first chart below, you’ll see that CapEx was, until 2022, dominated by red (Maintenance CapEx), with only a modest amount in blue (Growth CapEx). But in 2022, there’s a clear uptick in growth investment. What changed?

The answer lies in the Inflation Reduction Act (IRA), which introduced a range of incentives for renewable energy. RNIs (Renewable Identification Numbers) and LCFS (Low Carbon Fuel Standard) credits now make producing RNG (Renewable Natural Gas) from landfills “economically” attractive.

Why the quotation marks around “economically”? If you dig into the actual costs of producing RNG from landfills, estimates range from $ 7.50 to $ 21.50 per MMBtu. For context, take a look at the second figure below, which shows the price of natural gas in the U.S. over the last 20 years. The green line marks the lowest cost to produce RNG—notice how it compares to market prices. In other words, without incentives, RNG production is not profitable.

The key risk for Waste Management is that if government incentives are withdrawn, these new assets could quickly become uneconomical, even if we treat the initial CapEx as a sunk cost. Time will tell whether this was a prudent long-term investment.

$FLS and the New Oil Order: Why Middle East Turmoil Might Have a Smaller Impact Than Before

As part of my ongoing analysis of $FLS (Flowserve Corporation), a global leader in the design, manufacture, and service of flow control systems—including pumps, valves, seals, automation, and related services for the oil and gas, chemical, power, and water industries—it’s essential to understand the broader energy market context in which the company operates.

The first chart compares the price of oil in both nominal terms (blue line) and inflation-adjusted terms (red line). A striking feature of the chart is the dramatic spike in oil prices during mid-2008, when the inflation-adjusted price of oil exceeded $220 per barrel. In contrast, current oil prices are significantly lower, both in nominal and real terms, highlighting how much less expensive oil is today compared to that historic peak.

Turning to the second chart, we see the evolution of land and offshore rig counts in North America. Historically, geopolitical tensions in the Middle East have had a pronounced impact on the US economy, mainly due to the country’s reliance on imported oil. However, the shale revolution has fundamentally altered this dynamic. The 2000s saw a substantial increase in the number of rigs operating in the US, coinciding with significant advancements in hydraulic fracturing (fracking) technology. This surge in domestic production has reduced the US economy’s vulnerability to external oil shocks and has been a key driver of energy independence.

Interestingly, the most recent spike in oil prices did not result in a proportional increase in rig count, as seen in previous cycles. This could suggest several things:

  • Higher Break-Even Prices: Many fracking wells now require higher oil prices to be economically viable, as the most accessible reserves were tapped during the initial fracking boom.
  • Productive Well Inventory: A substantial inventory of productive wells may still exist, reducing the immediate need for new drilling activity.
  • Industry Caution: Operators may be exercising greater capital discipline, focusing on maximizing returns from existing assets rather than aggressively expanding capacity.

For Flowserve, these dynamics are highly relevant. The company’s growth prospects are impacted by capital spending cycles in the oil and gas sector, which are influenced by both oil prices and geopolitical stability. While current Middle East tensions have injected fresh volatility into the market, the structural resilience provided by U.S. shale production and a more cautious approach to new drilling may temper the impact on equipment demand in the near term.

Housing Starts: The COVID-19 Blip and What Comes Next

As I work on valuations for $HD (Home Depot) and $LOW (Lowe’s), I regularly update my broader analysis of the housing sector. While I have a wealth of charts on this topic, including all of them in one post would make it far too lengthy. For now, I’ll share just a couple of key charts—more will follow in future posts.

The first chart below is one of the earliest I created on this subject, dating back nearly 20 years. It tracks U.S. housing starts at an annualized rate, calculated by multiplying monthly figures by 12. Housing starts in the United States exhibit significant seasonal variation, particularly in regions with harsh winters. To smooth this out, the red line represents a 1-year moving average.

Take a moment to review the annotations on the chart—they highlight major economic events over time. Notably, every significant economic crisis in the U.S., with the possible exception of the 2001-2002 recession, has either originated in or been closely tied to the housing sector. The most recent event marked is what I call the “COVID-19 blip,” when zero interest rates and aggressive stimulus measures triggered a surge in new home construction.

The second chart illustrates one of the unintended consequences of this “blip.” The blue line represents housing starts minus completions, while the red line shows a 12-month cumulative total (scale on the left). What stands out is that we’ve just come through a period where more houses were being completed than started.

In 2024 alone, approximately 250,000 more homes were completed than initiated. This mismatch has had ripple effects across numerous companies tied to RIM’s CofC (Circle of Competence), particularly those in the building materials sector, which have reported weak or declining sales as a result.

Here’s where it gets interesting: as discussions about a potential recession heat up, this particular drag on the economy is nearing its end. By April 2025 (yes, tomorrow!), the blue line is expected to turn positive again. This shift will remove one of the most significant headwinds for the economy, given housing’s outsized influence on overall economic activity. To be clear, this doesn’t mean we’re on the cusp of another housing boom—but it does suggest that the lingering effects of the “COVID-19 blip” will finally fade.

That said, history suggests that boom-bust cycles in housing are far from over. So, while this particular drag may be dissipating, don’t get too comfortable assuming stability in this sector—it’s always full of surprises.

Is Sleep Number ($SNBR) Feeling the Weight of Consumer Recession?

When analyzing a company, understanding the industry it operates in is essential. Today, I’m focusing on $SNBR (Sleep Number Corporation), a company whose valuation history over the past 20 years has been remarkably volatile. This volatility stems from the cyclical nature of the mattress manufacturing and retailing industry, compounded by decisions made by two consecutive CEOs to leverage the company ahead of major industry downturns.

While the leverage applied in both cases wasn’t excessive, Sleep Number operates with a naturally leveraged business model due to its reliance on rented stores, which adds fixed costs to its P&L. During the Great Financial Crisis (GFC), the company nearly collapsed despite carrying only a modest debt load (but eventually managing to pay down all its debt). Fast forward to recent years: the outgoing CEO—who earned nearly $80 million during her tenure—chose to leverage the company again by authorizing $1.6 billion in share buybacks. The result? Sleep Number now faces significant financial strain with $550 million in debt.

The chart below provides a visual overview of Sleep Number’s sales (in red) over the past 20 years and projects a base-case scenario for future sales. It also compares these figures to unit sales for the entire mattress industry (in blue). Unsurprisingly, the two are closely correlated—any shifts impacting the broader industry inevitably affect Sleep Number. The dotted light-blue/red lines adjust these figures for population dynamics. For the overall mattress industry, I use the total population. However, for Sleep Number, I focus on individuals aged 30+ years who are more likely to purchase their premium mattresses.

Several reference lines on the chart offer additional context. One highlights industry sales levels during 2009, which we’re approaching but haven’t quite reached yet. If mattress unit sales decline by another million in 2025, per capita mattress sales will align with GFC levels—a concerning benchmark. Another line (light blue) illustrates cumulative under- or over-purchasing trends within the industry. Even if recent declines reflect a correction from prior excesses, seeing per capita sales nearing 2009 levels is striking. This aligns with broader indications that American consumers are facing recessionary pressures—a topic I’ve explored in previous posts.

What Trucking Tonnage Tells Us About Market Crashes and Recessions

I was recently asked if a market crash can trigger a recession. In short: yes, it certainly can—and it did in the early 2000s.

The chart below should look familiar to RIM’s clients. It shows trucking tonnage data from the ATA (American Trucking Association). The thin-dotted blue line represents monthly figures, which tend to be volatile due to weather seasons and holidays. I’ve added a thick blue line showing the Last Twelve Months (LTM) moving average to better identify trends.

Every turn in this LTM line has a story behind it—economies typically grow until something disrupts them. That’s why you’ll see annotations on the chart marking significant events: sharp oil price spikes, declines in housing starts, rising interest rates, banking crises, market crashes, and even a global pandemic (I never imagined I’d be adding that one!).

Take a closer look at the early 2000s. Trucking tonnage began declining around March 2000—exactly 25 years ago—coinciding with the internet bubble bursting. While the tragic events of September 11th certainly disrupted economic activity further, the downturn had already begun well before then. However, the market crash was clearly the initial trigger.

Could we be experiencing something similar today? Nobody knows, but there’s no point obsessing over it. Unless, of course, you invest in companies for which it is challenging—if not impossible—to predict long-term sales and profitability. These types of companies are the core protagonists—in terms of how much they decline—during bubble bursting years.

As equity investors, our focus should remain on thoroughly understanding a select group of companies and accurately modeling their financials based on company- and industry-specific data. Then, when the market inevitably overreacts to short-term news or temporary EPS fluctuations, we’re ready to act decisively. If you’ve been following RIM’s approach, you know that’s exactly what I do every single day.

What Harley-Davidson's Financing Data Reveals About the US Consumer

$HOG (Harley-Davidson) operates as two distinct businesses: a motorcycle manufacturer and a consumer financing provider. In fact, Harley-Davidson Financial Services financed over 70% of the motorcycles sold in 2024. Like any company offering financing, it must set aside provisions for potential credit losses.

The chart below illustrates these credit loss provisions (in basis points) relative to their total outstanding receivables portfolio. Notice the significant spike in 2009 during the Great Financial Crisis (GFC)—a severity that’s tough to surpass. (You might also spot my 2018 comment, noting future cycles would come, though likely not as extreme as the GFC.)

But more importantly, observe how elevated these provisions have been over the past two years (and likely continuing into 2025). Those surprised by recessionary signals today simply haven’t been paying attention to the right indicators. As I’ve mentioned in previous posts, the US consumer has already exhibited recessionary behavior for some time.

US Census Bureau Revisions Shrink America's Youngest Population: Implications for Carter's Future Market

When I’m working on $CRI [Carter’s], I need to update the forecasted population growth for the “zero to 5” segment in the US (done by the US Census Bureau), as they are the company’s core “clients.” The first chart below shows—per various group ages—how many millions of people the US has had and is expected to have over the next few decades.

The yellow arrow highlights the “zero to 5” group. Just below it is the “6 to 10” group. Note that the US Census Bureau drops the number of young children in the US with each dataset revision. The second chart gives you an idea of how much: in 2022, the “zero to 5” group’s estimate was reduced by almost 9%!