I ended part one (here) with the suggestion that I had not yet fully accounted for the important factors at play. The most notable one is Walmart’s (WMT) repurchases in the decade through FY15 that happened despite the International CapEx investments I discussed in part one. In the decade through FY15, Walmart reduced the number of diluted shares by more than 1 billion shares; the company spent $54.6 billion on repurchases in the 10 years through FY15.
To simply assume $54.6 billion in repurchases in my calculation would overlook the help of the incremental profitability from the International operations (which I’ve fully attributed to the CapEx spend). The incremental International operating profits were a cumulative $22.5 billion; after accounting for taxes (35% tax rate) we can estimate International funded $14.7 billion – or roughly 27% – of the repurchases in the decade to FY15 (that generously assumes 100% of incremental International profits went to repurchases, when in reality Walmart spent less than 40% of total profits on repurchases from FY06 to FY15). As such, that amounts to $39.3 billion in additional repurchases over the decade ($54.6 billion less $14.7 billion from International).
To make this easier, I’ll simply double the number of shares bought at our midpoint ($38 billion) from part one; here’s the updated output at each repurchase level:
| Average cost | Shares repurchased | Shares left, FY15 | EBIT/Share (FY15) |
| 7x EBIT | 1,887 million | 2,379 million | $9.98 |
| 9x EBIT | 1,504 million | 2,762 million | $8.59 |
| 11x EBIT | 1,250 million | 3,016 million | $7.87 |
| 13x EBIT | 1,069 million | 3,197 million | $7.42 |
| 15x EBIT | 934 million | 3,332 million | $7.12 |
After that final adjustment, we can see that Walmart would’ve reported comparable earnings in FY15 in either scenario if its average repurchase cost was ~10x EBIT – roughly 16x earnings. By comparison, if the average repurchase price had been at a low-teens P/E, FY15 operating income per share would’ve been ~10% higher than what Walmart actually reported in FY15.
Repurchases would’ve done quite well as an alternative for Walmart over the past decade.
An important question
You might be thinking to yourself that I’ve overlooked a critical point: in reality (relative to the adjusted scenario I’ve created), Walmart added real assets (stores, distribution centers, etc.) that generate an additional $3.4 billion in operating income each year. Under the assumption that this will continue (as long as we fund maintenance CapEx), isn’t it better to have this additional cash flow – year after year after year – than a lower share count?
It’s instructive to invert the question: could we recreate that cash flow in some way?
To simplify the numbers, let’s assume company A earns $2 billion a year and has 1 billion shares outstanding; the company earns $2.00 per share.
Company B, on the other hand, is like the Walmart I’ve created above: it repurchased a lot of shares instead of investing in CapEx. It only earns $1 billion a year, but with 500 million shares outstanding; just like company A, it earned $2.00 per share last year.
Next year, both companies take their first billion dollars of earnings and reinvest in CapEx at identical rates of return; now company A has a free pass to start pulling ahead, right?
No – because company A only generates half as much in profitability on a per share basis on its first billion dollars of CapEx as company B. If both companies add $200 million in ongoing profits from their $1 billion in CapEx, company A has increased EPS to ~$2.20 per share; company B, on the other hand, has increased EPS to ~$2.40 per share.
If company A doesn’t take the rest of its excess cash and invest it at comparable rates of returns to its first billion of CapEx, they will start falling behind on a per share basis.
Much like the continuing generation of additional funds has implications beyond the current year (due to investments in growth CapEx made years ago), the incremental per share earnings power generated as a result of share repurchases does not disappear either. The implications on a per share basis continue in perpetuity.
Conclusion
Capital allocation decisions like M&A, CapEx or share repurchases are neither good nor bad in a vacuum. The expected returns, which are a factor of the price paid to the value received, is the all-important factor in all three cases. What’s smart at one price is dumb at another.
I look forward to hearing readers’ thoughts on what I’ve presented.
