Thursday, December 1, 2011

Whole Foods and SuperValu

Investor sentiment varies considerably between Whole Foods (WFM) and SuperValu (SVU), despite the fact that these two companies operate in the same (or similar) industry. This difference in sentiment is exemplified by the opinion of a one Jim Cramer, who notes that even though SuperValu has a substantially lower P/E, Whole Foods is actually the cheaper stock because of its lower PEG ratio. What's wrong with this surface analysis?

First, a PEG ratio is only as good as its underlying assumptions. In this case, Cramer is assuming Whole Foods will grow at an 18% annual rate. But Whole Foods has only grown its revenue at an 8% annual clip over the last three years, and at 6% year-over-year in its most recent quarter. As such, either revenue has to pick up in a big way, or Cramer is assuming Whole Foods will be able to grow its margins rather significantly every year.

This growth rate assumption is just that - an assumption. And history suggests we humans tend to err in our growth assumptions while at the same time remaining very confident in them - a bad combination. Research by David Dreman illustrates this problem, and helps explain why stocks with low P/E's dramatically outperform those with high P/E's, even with the same industry: because we are pretty poor at predicting the future growth rates of individual companies.

So while Whole Foods is clearly the better performing business as of late, having posted far superior margins and same-store sales increases, it's important not to be over-reliant on a large growth assumption. Comparing the price of the businesses as they stand today (i.e. removing growth from the equation), SuperValu has ttm operating earnings of $850 million while Whole Foods has ttm operating earnings of $550 million. And yet SuperValu can be purchased for an enterprise value of about $8 billion (for an EV/EBIT ratio of less than 10) while Whole Foods comes at a cost of $11 billion (for an EV/EBIT of 20).

As such, you're paying more than twice as much for Whole Foods as you are for SuperValu. In so doing, you are making an assumption that there will be a consistently large growth rate differential between the two companies. Considering what we know about how good we are at making assumptions about growth rates, you might not want to make that bet.

The second problem is that even if Whole Foods were to achieve such growth rates, that growth is not free. It takes capital to grow, particularly in a business such as this one. Companies in the grocery business need to buy or lease land, pay for construction, finance inventory growth, add or enlarge distribution centres and add staff if they are going to grow. That capital could come from debt, which would increase Whole Foods' enterprise value (thereby giving shareholders a smaller stake in the company's operating earnings), or it could come from retained earnings, which would otherwise accrue to shareholders.

In the last few years, Whole Foods has grown through a combination of both, as it has used non-cancelable leases (a form of debt) and cash from operations (that could otherwise have been paid out to shareholders) to finance its growth.

One manner in which the contrast between the capital requirements of these two companies can be seen is in their respective dividend yields. Whole Foods pays a dividend of less than 1% while SuperValu pays almost 5%. At the same time, a low-growth SuperValu will have (and has had) much more capacity to deploy the cash flow it doesn't pay out to shareholders towards paying down its debt, reducing its enterprise value in the process (thus making it cheaper).

This isn't to say that Whole Foods should stop growing and should instead pay out its earnings. Such a statement is beyond the scope of this article, and would involve a discussion of a number of issues including business risk and returns on equity and capital. It is to say, however, that there is a cost to growth, though the simplistic PEG ratio ignores this important element.

SuperValu is clearly cheaper than Whole Foods, until you start making assumptions about growth. It's important to recognize the drawbacks of this "assumption" approach, however, and realize that growth must be financed with additional capital.

Disclosure: Author has a long position in shares of SVU

2 comments:

Anonymous said...

Hi Saj,

Good article. The same high growth rate assumption by street caused the dot-come bust.

Anonymous said...

I thought the denominator in the PEG ratio was earnings growth, not revenue growth. I'd gladly pay 2x for Whole Foods business. Remember, SVU still has to reduce prices and ebitda dramatically to be viable. normalized ebitda is much lower. same store sales (ex inflation) have been abysmal for 12 quarters plus. thats not turning soon.