Repurchases vs Capex: The Math (Part 1)

In the comment section of my most recent article, a reader made a good suggestion: analyze the returns of Walmart’s (WMT) investment in CapEx relative to alternative uses – primarily share repurchases (which was the topic of discussion). This article will attempt to get us there by answering the following question: what would happened if Walmart stopped building new stores outside of the United States a decade ago and instead committed the capital to share repurchases?

Some background work

Let’s start with an easy question: how much has Walmart spent on international CapEx in the last 10 years? Thankfully, Walmart has broken out U.S. versus International CapEx for some time: in the past decade, Walmart’s international operations have spent roughly $42 billion on CapEx.

Next, we need to find the breakdown in CapEx spend between maintenance and growth (penalizing the actual results for the entirety of international CapEx spend over the decade would overlook the capital spending needed to maintain the legacy business). While the company doesn’t provide this level of detail for the International operations, it has in the U.S. for some time (since fiscal 2007); domestically, the breakdown has been ~50% for “New Store / Expansions / Relocations,” ~30% for “IT / Distribution / Ecommerce,” and ~15% for “Remodels.” Reasonable people can disagree, but I peg growth CapEx at ~30% or more of the total. I assume half of the first category is used for new stores (so ~25% of U.S. CapEx), with another ~5% of CapEx to account for the fact that an expansion (to a Supercenter, for example) is somewhere between maintenance and growth. I think that’s sufficiently conservative.

When we look at the International side of the business, we must consider one notable difference: in the United States, the store count has increased by nearly 1,500 units, or 40%, over the past decade. On the other hand, the number of International units increased from ~1,600 to ~6,300 over same period – an increase of nearly 300%. It’s safe to say that a much larger percentage of CapEx outside of the United States has been spent on growth. Personally, I’m comfortable with assuming three-quarters of the total has been spent on growth CapEx on average (and probably more). I’ll run the numbers at 60%, 70% and 80% to put that estimate into perspective.

So with that, we’re done with the first step: by my math, Walmart has spent $25 billion to $35 billion on growth CapEx in the International business over the past decade. We must also remember to include the capital spent outside the CapEx line (International M&A, like Massmart). Based on the disclosures in the annual reports, the company spent at least $8 billion for International M&A over the past decade. That brings the range to $33 billion to $43 billion.

Here’s the next question: what does Walmart actually have to show for that investment? As I noted earlier, they have a lot more stores outside the U.S. than a decade ago (~1,600 to ~6,300). In addition, they added ~$90 billion in annual revenues (from ~$52 billion to ~$142 billion), good for a ~175% increase from a decade ago. However, the operating margins for the International business have fallen by ~100 basis points over the measurement period, resulting in a much lower increase in segment profits (up ~115%, to ~$6.4 billion annually).

For Walmart as a whole, the incremental profits from the International business (FY15 vs. FY05) were ~$3.4 billion; if we remove this amount from the FY15 reported results, Walmart’s operating income would’ve been ~$23.7 billion – or 12% lower than what the company actually reported in FY15. Clearly growth outside the U.S. has helped fuel earnings growth at Walmart.

Before going further, I should address something here: you can obviously argue that segment operating income in a single year doesn’t capture the value of CapEx spend in a given year (or for a period of time). The money Walmart spent on International CapEx in fiscal 2014 could very well still be generating an operating loss for the region. Of course, we can also say the same about the CapEx spend in FY04, which impacted our starting estimate as well (said differently, the years past FY05 benefit from CapEx spending in the years before that point). While it’s clearly impossible to account for the impact of individual years, I didn’t want to simply skip this. It’s one of many factors that you must consider as part of this exercise (it’s far from perfect).

Assumptions

Now that we’ve done most of the background work, we can start to answer the question I asked above: what would have been the result if Walmart stopped building new stores outside of the U.S. a decade ago and instead committed the capital to share repurchases?

I’ll have to make a few assumptions to run the numbers, so let me lay them out here:

(1) I’ll assume that the repurchase spend ramps over time to account for growing levels of cash flow in the late years. Walmart’s sales in the U.S. increased by roughly 50% from FY05 to FY15 (by far the largest driver of cash generation across Walmart as a whole), so I’ll assume that the repurchase spend in FY15 is 50% higher (in dollars) than in FY05.

(2) I’ll be using repurchase prices based on a multiple of operating income; I’m doing this so I can easily back out the incremental profits from the International operations. Walmart’s tax rate has consistently been in the mid-30s, with net income consistently at ~60% of EBIT. I’ll use EBIT multiples of 7x, 9x, 11x, 13x and 15x; those are equivalent to P/E multiples of ~12x, ~15x, ~18x, ~22x and ~25x.

Just to clarify, here’s how I’m calculating the results: In each scenario, Walmart begins with its year-end FY05 diluted share count (4,266 million shares). I will then show the number of shares repurchased over the course of 10 years at $33 billion (60% growth CapEx), $38 billion (midpoint), and $43 billion (80% growth CapEx), as explained above. I will then show the number of pro-forma shares remaining at the end of fiscal 2015, and what that means for EBIT per share in FY15 (as a reminder, the gross number will be equal to Walmart’s actual EBIT in FY15, less the incremental operating profits generated in the International business).

Here’s the output data (I’m happy to share my work with anyone who wants to see the math):

Repurchases at 7x EBIT (~12x P/E)

Repurchases at 9x EBIT (~15x P/E)

Repurchases at 11x EBIT (~18x P/E)

Repurchases at 13x EBIT (~22x P/E)

Repurchases at 15x EBIT (~25x P/E)

As we can see, the results range from a low of $6.15 per share in EBIT to $7.39 per share in EBIT. How does that compare to what actually happened? Walmart reported $27.15 billion in operating income in FY15, on 3,242 million diluted shares outstanding – equal to $8.37 per share. Clearly that’s much higher than any of the numbers shown above.

However, there’s one important factor we haven’t accounted for yet; part two of this article will pick up here (broken up to account for length).