Description

Short CST: Breaking Rose-Colored Glasses

CST is a gas station/convenient store business that was spun off from VLO on 5/1/13. CST has traded well and is at ~$33 now. Given the information vacuum and general excitement around spin-offs, CST is highly misunderstood. There’s nearly 40% downside to fair value of $21 once the market appreciates significant downside to earnings and CST’s poor business prospects. In particular, CST should generate 2013 EPS of ~$1.60 (vs Street’s ~$2.25) and deserves below-market multiple of 13x (vs Street’s target of 15-16x) reflecting its worst-in-class assets with negative same-store sales (SSS). CST’s weak FCF potential (44%-85% of EPS depending on growth assumptions) suggests further downside below $21.

Conversations with VLO and CST suggest both management teams are pleasantly surprised by how well CST has been trading. It’s also interesting that recent CST equity awards were not granted through options but through restricted stock.

Company Overview

• CST operates 1,032 gas stations in the U.S. (mostly in the Southwestern states). CST captures profit both from fuel retailing and c-store businesses.
• CST also has 848 locations in Eastern Canada. CST captures fuel retailing profit from all locations but participates in c-stores at 261 locations (the remaining locations are either operated by dealers/agents or are unattended truck fuel sites without a c-store).
• In 2012, 47% of gross profit came from motor-vehicle fuel, 40% from c-store merchandise, and 13% from other (e.g. car wash, heating oil, ATM, lottery). This business mix is unique as CST has much greater exposure to fuel and less exposure to merchandise vs both publicly listed peers and industry average.

Street’s Perception vs Reality

• There is currently no official consensus as only one sell-side firm has initiated on CST. However, buy-side investors appear to converge around a consensus of ~$420m EBITDA and ~$2.25 EPS in 2013. This represents ~$40m of increase in EBITDA from 2012.
• Examples of Street’s general view can be seen in these articles (http://bit.ly/16xYdax and http://bit.ly/17Z3kRq).
• Specific drivers behind the ~$40m yoy increase differ among investors but generally come from some combination of the factors summarized in the above table and discussed further below. In contrast, a more realistic fundamental view suggests EBITDA should fall by ~$30m yoy to $347m in 2013.

1. US fuel margins

• Street assumes US fuel margins to be flat or rise by 1c/gal based on an apparent trend of rising margin in recent years. Note that every 1c/gal of fuel margin translates to $19m of EBITDA in the US.

• This apparent trend obscures the underlying relationship between fuel margins and changes in wholesale fuel prices. In a period of rapidly falling crude oil and wholesale fuel cost (such as in 2Q12 when Brent fell from ~$125/bbl to ~$90/bbl, the sharpest fall in the history during Q2), CST would reduce retail fuel price more slowly and capture greater margins. Conversely, in a period of rapidly rising crude oil and wholesale fuel cost (such as in 2009), CST would experience some headwind and report lower margins.
• Fortunately, a large historical dataset (quarterly since 2002) is available on VLO’s website (http://bit.ly/13zRx8R) to clearly demonstrate this relationship. Indeed, a regression between realized fuel margins against changes in wholesale gasoline cost (MOIGC87P Index on Bloomberg) from the beginning till the end of each quarter shows tight correlation and conclusively proves this underlying relationship (see regression charts here http://bit.ly/12PWcCF). For any math/stat major out there, the product of “p-values” indicates the odds that this relationship does not hold (and perhaps margins are secularly rising) is 1 in ~3 trillion (~16,000x longer odds than a PowerBall jackpot)!
• Furthermore, CST and VLO have acknowledged the negative relationship between wholesale fuel cost and retail fuel margins in public filings (e.g. see page 45 of VLO 2Q12 10-Q). Other public comps have recognized this relationship as well and indicated that fuel margins in 2012 were abnormally high and should contract in 2013.
• Assuming oil price remains stable for the rest of the year, 2013 margin should be 1.8c/gal lower than 2012, representing $34m in EBITDA headwind in the US. A large yoy decline should occur in 2Q13, when CST will be comping against an impossibly high bar in 2Q12 (VLO’s historical data shows US retail EBIT in 2Q12 was 2.7x the preceding 10-year average). Note that fuel margins in 3Q12 and 4Q12 were below and above average respectively, but together the 2H12 margin was approximately in-line with the normal range and should be flat into 2013 (absent wild swings in crude oil and fuel cost).
• Why is the market missing this fundamental relationship? A constellation of factors align and cause this colossal misunderstanding:

a)     False trend - CST’s limited historical data by chance shows rising margins.

b)    Rose-colored glasses - Investors eager to chase spin-offs are ready to see the upside and management hasn’t proactively managed expectation.

c)     Irrelevant to VLO - Legacy VLO investors don’t focus on retail metrics since it’s more useful to analyze the much bigger and more volatile refining business. Most CST investors aren’t even aware of relevant historical disclosure provided by VLO.

d)    Irrelevant to peers - Perhaps most importantly, even veteran retail sell side analysts and investors aren’t well aware of this negative relationship because (i) the relationship is less important to other public comps with small exposure to fuel margins, (ii) the correlation is less pronounced for other c-stores as they typically cut retail fuel price quickly (and do not benefit as much from the lag) in order to drive more traffic towards the c-store business. (CST can’t afford to do the same since fuel is the main business).

2. US organic and new store growth

• CST’s US business generated $162m out of the $263m in total EBIT in 2012. Most of its US locations are in the Southwestern states (especially Texas) which have strong economic growth. This fact was prominently highlighted in CST presentation and resonates with many investors. A 5% organic growth in the US (similar to Susser which operates in Texas) equates to $8m of EBIT. In addition, CST is investing $115m to open 23 new stores this year. Street assumes 25% EBITDA return on this capital. Assuming new stores open ratably through the year, CST should generate $14m of incremental EBITDA.
• Actual results show that CST’s US fuel volume and merchandise sale (per store per day) have had zero growth over the past two years.

• The same weakness can also be seen in EIA data, which shows flat gasoline demand in the Southwestern region (PADD 3) over the past two and five years. While PADD3 is doing “better” than the US overall (down 3% and 6% over the past 2 and 5 years), it is nothing to write home about.

• CST’s prospect is even worse than it looks because the flat per store metrics (different from SSS metrics which were not disclosed for 2009-12) include benefits from M&A. Since 2009, CST has (a) closed/divested 217 underperforming stores (which boosts per store metrics for remaining stores) and (b) opened/acquired 137 new stores that are larger (which also improves per store metrics). To benefit from significant M&A tailwind and still show no growth means SSS in recent years must be negative. Indeed, CST disclosed SSS metrics for the first time in 1Q13, showing 1.6% decline in fuel volume and 0.9% decline in merchandise sales. These figures benefit from higher growth rates from the ramp-up of new stores which should taper off as the stores mature. SSS figures are slightly worse (-1.1% fuel and -2.0% merchandise) excluding new stores. Because of negative SSS, new store investments are needed just to keep CST from sinking. The “Realistic” case above assumes $5m of net EBITDA increase from new stores, offset by negative SSS.

3. Why CST is different from peers.

• Many investors compare CST to other fuel/c-store retail peers and imply that CST will enjoy similar growth and deserves similar valuation. Figures above clearly shows poor prospects for CST which can be attributed to the following reasons.
• Fuel vs merchandise. The table in the overview section shows that CST is primarily a fuel retailing business while peers are primarily a merchandise retail business. Fuel is in a secular decline while merchandise can generate moderate growth (with the right store format). The distinct fundamentals between fuel and merchandise make CST not comparable to peers.
• Flexibility. CST’s outsized fuel exposure also limits its flexibility to use low fuel price as a “leader” to attract customer traffic to boost merchandise sales. While this strategy may be attractive to peers with a dominant merchandise business (2-3x of fuel gross profit), the tradeoff doesn’t work as well for CST. Effectively, CST is stuck with relatively large fuel exposure (this applies to CST as a whole and separately to US and Canada segments) and cannot easily diversify to merchandise. Over time, as industry fuel volume declines and the fuel-for-merchandise tradeoff becomes increasingly attractive, there will be more pressure on fuel margins which will disproportionately hurt CST. To use the supermarket analogy, CST is selling a lot more milk (a well-known loss leader) than peers.
• Foodservice. The key growth category in merchandise is foodservice, offering not just hot dogs on rollers but actual QSR implants serving made-to-order items such as tacos with a nice sitting area a la Chipotle. CASY’s stores even offer pizza delivery! QSRs require bigger store boxes ranging from 3,200 sq.ft. at the low end to over 5,000 sq.ft. With an average box size of 2,300 sq.ft., CST simply doesn’t have enough space to install a kitchen. Growth from foodservice will have to come from new stores.
• Size. CST already has a large base of 1,880 locations. Adding 23 stores this year (or even more next year) will hardly move the needle, especially considering the likely offset from store closures (an average of 45 per year during the past 3 years). CST’s large size also hinders its ability to increase mix shift towards foodservice. Even peers that started from a much smaller base and have prioritized growth in merchandise show little mix shift from fuel towards merchandise over the past 15 years.
• Timing. CST is also disadvantaged as a late entrant into the large store format. Peers like Susser were picking the best locations for new stores, developing its food and merchandise brands, and winning customer loyalty since 2000. The competition has recently gotten fierce. As Susser noted in its 1Q13 call “The strong economy in our region also is attracting significant numbers of new retail competitors, including expansions by the dollar stores, drug stores, and large box retailers offering fuel. In addition, in certain markets, we're seeing an increase in the number of new convenient stores, being built by traditional operators.” It’s hard to imagine how new CST stores today will be able to match the ramp-up of Susser’s store vintages that opened a decade ago.

4. Merchandise gross margin improvement

• CST has lower merchandise gross margins (29.6% in 2012) than peers (average of 35.2%). Street believes the company is undermanaged and margins can improve. Every 100bps increase equates to $15m of EBITDA.
• Most of the 560bps difference can be explained by CST’s worse merchandise mix. In particular, CST’s lower exposure to foodservice alone explains 330bps of the difference (foodservice carries 30% higher gross margin than average). As discussed earlier, CST’s low foodservice exposure cannot be fixed easily. CST’s higher exposure to cigarettes accounts for another ~50bps of margin downside (cigarettes has ~15% lower margin than average). The remaining 180bps gap can be attributed to CST’s older store base, which is unlikely to command as much pricing power as peers.
• Some have suggested an easy margin fix by replacing cigarette sales with food sales. That’s a good plan if we can put a kitchen on the small shelf behind the cash registers.
• Even assuming some upside in gross margins from greater foodservice mix, this would be offset by increased operating expenses given the service-oriented nature of foodservice. The table below shows greater ratios of SG&A to gross profit for all peers relative to CST.
• It’s more likely that increases in gross margins and lower mix of cigarettes will be driven by declines in cigarette sales (the pace of which have intensified over the past year) rather than by increases in sales of other merchandises (1Q13 result shows same-store merchandise sales ex cigarettes only grew 0.5% yoy). This means gross profit dollar per store is actually falling despite the increase in gross margin (gross profit per store per day in US fell from $939 in 1Q12 to $936 in 1Q13 despite margin rising from 29.5% to 29.9%).

5. Poor FCF prospects

• CST “maintenance” CapEx guidance of $75m (below D&A of ~$120m) ostensibly suggests FCF should be higher than EPS.
• However, negative SSS means meaningful new store CapEx is required to maintain flat EBITDA or generate any growth.
• The table below shows that CapEx needs to be $140-190m to generate minimal 0-3% EBITDA growth, indicating FCF of $0.72-$1.38 or only 44%-85% of EPS.

Other items

• Management: All of CST senior management came from the legacy Retail business or other parts of VLO. CEO was previously VLO’s general counsel. No new talent from the retail industry added.
• Share count: CST has a total of 76m shares, including 61m shares of float and the remaining 20% retained by VLO. Some reports use an incorrect share count of 61m.
• Incremental public company cost: CST has guided to $20m of cost (as noted in company presentation). Some reports (including Stephens’ initiation) fail to capture this cost.
• MLP: Susser’s MLP was anchored around its wholesale fuel distribution business. CST doesn’t have a wholesale business and its fuel retail business is not MLP-able.
• REIT: CST owns real estate in 61% of its total locations (1,152 out of 1,880). REIT conversion (ie opco/propco) is not doable because the distribution of propco to VLO (which owns 20% of CST) would trip up tax rules. Management also wants to retain control of real estate. In any case, on realistic estimates CST is already trading at a premium to CASY (which owns 99% of its stores) and to potential upside from forming a single-tenant propco.
• 1Q13 result: Reported EBITDA of $59m reflects no impact from incremental public company cost ($5m per quarter) and only partial hit from fuel supply agreement ($5m vs full impact of $8m because the agreement was not fully implemented at the beginning of the quarter). PF EBITDA should be $51m. In addition, reported result doesn’t reflect interest expense on new debt ($11m per quarter). PF 1Q13 EPS should be 10c.

Conclusion / Valuation

• Motor fuel demand is in a secular decline thanks to rising fuel economy. Despite being the best house in the neighborhood, the Southwestern states are showing no growth. An eventual decline seems inevitable.
• CST is overexposed to fuel retailing and underexposed to the c-store business.
• CST’s challenge is structural and difficult to be fixed. CST’s businesses are struggling to stay flat despite significant investment in new stores and the benefit from store closures.
• Growth will come largely from new stores, but they are unlikely to be significant relative to CST’s large existing store base.
• Given the above considerations, a 13x P/E seems appropriate for this no-growth, secularly and competitively challenged, and worst-in-class business. This results in fair value of $21 and implies an EBITDA multiple of 6.7x. CST’s poor FCF generation suggests further downside.
• Note that peers CASY, PTRY, and SUSS are trading on average at 7.3x CY13 EBITDA with significantly greater growth than CST (SUSS multiple is adjusted for its ownership in SUSP, which brings the headline multiple down from 8.7x per Bloomberg to 7.4x). Couche-Tard’s multiple of 10x CY13 EBITDA benefits from its significantly lower tax rate of ~15% (vs peers in the mid 30’s) which converts 31% more pretax income to the bottom line (85c vs 65c for every dollar of pretax income). Adjusting for this unique tax advantage, Couche-Tard would be trading in-line with other peers.

P&L Summary

Catalyst