Description

Note

I pitched Whole Foods for my VIC application in August last year. With Jana’s 13F filing, my thesis is just beginning to play out, so I thought it might be useful to post my updated thesis .

Thesis

Whole Foods is selling at a discount to the value of its existing stores despite an easily identifiable and fixable problem. In particular, it’s level of SG&A, whether normalized against its square footage or sales is inflated to an extreme degree. This stems from a management composed of great builders but poor operators who are utilizing IT systems for inventory and workforce management that are more appropriate for a mom & pop shop than a professional company. I believe this opportunity exists for two primary reasons: (1) the market underestimates the level of mismanagement and inefficiency at Whole Foods (2) the market fears permanent Sales/SF and gross margin contraction due to greater competition from traditional supermarkets. It turns out that because the level of mismanagement is so great the potential savings from SG&A cuts more than offset most bearish projections of sales/SF and gross margin contraction.

WFM Is the Least Efficiently Run Grocery Store by a Mile

The stats tell it all:

·         Three Year Average SG&A/SF (SG&A including stock comp, store expense, other costs, and pre-opening, excluding D&A)

§  WFM: $240 (2x TFM)

§  TFM: $118

§  SFM: $129

§  Publix: $103

§  KR: $106

§  COST: $109

§  WMT: $72

§  SBUX: $173

§  PNRA: $155

§  BBY: $132

·         Three Year Average Employees Per 1,000 SF

§  WFM: 5.37 (1.5x TFM)

§  TFM: 3.45

§  SFM: 3.33

§  Publix: 3.48

§  KR: 2.48

§  COST: 2.08

§  WMT: 1.94

§  SBUX: 5.76

§  PNRA: 5.41

§  BBY: 2.69

·         Three Year SG&A Per Employee

§  WFM: $44,900 (1.3x TFM)

§  TFM: $34,323

§  SFM: $38,888

§  Publix: $29,449

§  KR: $42,611

§  COST: $55,693

§  WMT: $38,153

§  SBUX: $31,069

§  PNRA: $25,960

§  BBY: $49,069

·         Three Year SG&A Percentage of Sales

§  WFM: 26.3%

§  TFM: 23.4%

§  SFM: 22.3%

§  Publix: 16.5%

§  KR: 16.3%

§  COST: 9.7%

§  WMT:17.6%

§  SBUX: 37.2%

§  PNRA: 46.5%

§  BBY: 15.4%

·         Cashier Salary in Jacksonville (Glassdoor)

§  WFM: $11.62 (1.31x TFM)

§  TFM: $8.87

§  SFM: $9.26 (Atlanta)

§  Publix: $9.68

§  KR: $8.03

§  COST: $18.05

§  WMT: $9.17

§  SBUX: $8.68

§  PNRA: $8.52

§  BBY $10.83

TFM is the best comparison to WFM given that both sell quality organic and prepared foods and both get about 2/3s of sales from perishables (although it’s important to note that TFM was also poorly managed, leading to a buyout). Despite operating 4x the SF as TFM, Whole Foods’ SG&A/SF is more than 2x as high. I believe this is due to simple overstaffing/bloated management, as evidenced by Whole Foods having 60%+ more employees per square foot than TFM and 2x more than comparable companies. Whole Foods also likely overpays their employees, with cashier wages more than 30% higher than at TFM. While I don’t mind WFM paying above market wages to get the best people and reduce turnover (turnover is only 14%, on par with COST), this is only worthwhile for shareholders if the company employs fewer employees than competitors. WFM’s highly antiquated technology (they had 12 separate ERP systems and 3 POS systems in 2015) also contributes to their bloated cost structure.

Interestingly, WFM’s SG&A per SF is about 60% higher, and their number of employees per 1,000 SF at similar levels, when compared to Starbucks and Panera. You can argue that the butcher, bakery, and sandwich type sections in Whole Foods are similar to fast casual restaurants and should have similar costs structures, but these sections only account for about 25% of WFM’s SF (and SBUX and PNRA are still spending nearly $100 less than WFM on SG&A/SF). Applying TFM’s $118 of SG&A/SF to the 75% of Whole Foods’ square footage that is more like a traditional grocer, and the SBUX and PNRA average SG&A/SF of $160 to the 25% of the store that is like a restaurant, the blended SG&A/SF of WFM should be $129, 45% lower than it was in 2016. Using an alternative approach, if WFM continued to pay above market wages leading to elevated SG&A per employee, but reduced the number of employees per square foot to match TFM, Whole Foods’ SG&A per SF would be about $155, $80 less than it was in 2016.

Two Counterarguments

Bears make two excuses for Whole Foods’ higher level of SG&A compared to peers:

1.       SG&A/SF should be higher because their sales/SF are higher

If this were true we would expect Whole Foods’ ratio of SG&A/SF to Sales/SF to be similar to competitors. This ratio reduces to SG&A/sales. Although the discrepancy between Whole Foods and competitors is not quite as large on this metric as others, at 25.8% WFM’s  SG&A/sales is still 240 bps higher than TFM, 350 bps higher than SFM (despite WFM being 3x to 4x bigger) and 1,000 bps higher than Publix. Furthermore, I don’t believe that there is as strong a relationship between SG&A and sales as bears seem to think. KR and SFM both have lower SG&A/sales  than TFM yet have sales/ SF $100 to $171 higher. BBY is an interesting comp. Despite having gross margins (≈24%) and sales/SF (≈$869) closer to WFM than any of the grocers, BBY operates with SG&A/SF of $132 and SG&A/sales at 15.6%, both in line with grocer comps. This shows that it’s possible for WFM to simultaneously have high sales/SF and SG&A levels comparable to peers.

2.       SG&A should be higher because WFM stores are in states with more expensive labor

If this were true we would expect that if we weighted the PCE or minimum wage for each state by the number of stores in each state, Whole Foods’ weighted store base would have a higher PCE and minimum wage than competitors. As displayed below, the reality is that based on both PCE and minimum wage, Whole Foods’ costs should only be about 4.5% higher than comps. Even against TFM, which is concentrated in lower PCE southern states, WFM costs should only be about 6% higher. To reflect that Whole Foods’ stores are in slightly more expensive states than comps, I increase my previous estimates of appropriate SG&A/SF of $129 or $155 by 4.5% which takes brings them to $133 or $162.

$300 Million of Cuts Is Not Enough

Management has announced $300 million of SG&A cuts but this is a laughably low number. Even if SG&A/SF were reduced to $200/SF, nearly twice the level of any competitor, SG&A would decrease by $612 million from ending 2016 levels. Assuming SG&A/SF got to $150, a number in the middle of my estimates, WFM would save more than $1.4 billion from 2016 levels. Importantly, I believe these cuts can be done without reducing customer service. Publix has the best customer service of any retailer I’ve been to (and is ranked #2 in overall satisfaction among grocery stores by Consumer Reports), but spends less than half as much as Whole Foods in SG&A/SF.

Furthermore, on the Q117 conference call Mackey (CEO) gave an example of the kind of wasteful spending that seems to permeate the culture, while actually reducing sales:

“a trivial example is, traditionally, we’ve done pizza in our prepared foods has been a service department. But we found when we go to self-service, that sales go up meaningfully. I mean, in many cases the sales go up 100%. And that is true in bakery, it’s partially true in meat and seafood, so we’re just sort of thinking through our business model and doing lots of experiments in terms of where we can reduce labor costs that do not affect or negatively impact the customer.”

Cost Cuts Trump Operating Decline

I tend to believe that, although they will decline over the next few years, Whole Foods will ultimately be able to maintain sales/SF and GM margins above typical supermarket levels because (1) margins are structurally higher in perishables, where Whole Foods gets 2/3s of sales vs about 50% for Kroger (2) 30% of customers are tree huggers who will never go to a Walmart or Costco, so margins should be higher to account for inelastic demand (3) the product is of higher quality/taste than traditional supermarket offerings and a great prepared foods section drives people to the stores.

But even if I’m wrong and Whole Foods’ sales per square foot  falls to $850 (from $913 in 2016) and gross margins fall to 30% (from 34.4% in 2016), both below what the most bearish sell side analyst projects, WFM would generate $2.3 billion of EBITDA assuming SG&A/SF falls to $150 and only 100 new stores are opened. At a 7x multiple (a little lower than the acquisition multiples for The Fresh Market, Roundy’s, and Harris Teeter) for the existing stores and a 4x multiple for the 100 new stores I get $45 of value, 35% above the current share price.

Catalyst

 Jana leads management change and/or buyout