GNC (GNC) began focusing serious attention on the company’s store base last year with the announcement that the company would close approximately 200 company operated stores over the course of the current year. The company upped the ante in the most recent quarterly conference call by announcing plans to close between 700 and 900 additional stores over the next three years – possibly upwards of 25% of the company’s existing company operated domestic store base. The changes are probably overdue (with consideration given to the terms of the associated leases) based on the company’s statement that 10% of current company operated locations (or possibly less, as discussed below), which comes to as many as 320 stores, are currently EBITDA breakeven or negative. However, this announcement masks an already accelerated trend of closures when considering all store types. Indeed, in the last four quarters, the total number of GNC locations hasn’t declined by 200 but closer to 542 when all location types are included in the calculation.
We discussed our first step in developing updated economic models for the company based on the potential impact of the company’s proposed joint venture. In this article, we discuss the second step by reviewing store closure activity, providing an initial rough sketch of the economics of store closures, and outlining the early framework we’re using to refine our forward models for the company to account for the impact on revenues, operating income, and other financial metrics from closing 700-900 additional store locations.
GNC’s store base peaked at around 9,090 store locations of all types at the end of 2015 just before the company began to run into serious trouble with sales. The company has since seen store closures, particularly in the last two years, erase more stores than were added in 2014 and 2015. Domestic franchise and store-within-a-store locations are still up marginally from year end 2013, but domestic company owned stores and international franchise locations have declined to levels last seen earlier in 2013.
In the third quarter report, the company’s store presentation indicated the following for net store closures over the preceding 12 months in each store category:
- Company Owned Locations Closed: 143
- Domestic Franchise Locations Closed: 78
- International Locations Closed: 107
- Rite Aid Locations Closed: 172
The majority of these 542 closures – a full 414 – have occurred over the last nine months (i.e., since the end of the prior fiscal year). However, it’s necessary to realize that the closure of different store types have occurred for different reasons and have very different financial impacts on the company.
Company Owned Stores
The most notable and direct impact has been from the closure of company owned stores where revenue declines associated with these store closures have amounted to an estimated $24 million over the first nine months of the year and $9 million in the third quarter. Interestingly, the company hasn’t yet publicly ascribed any reductions in selling, general, and administrative expenses to store closures, so it’s difficult to assess the exact economics of the closed locations. However, we’ve developed rough order of magnitude estimates for use in our projections of the impact of future store closings which we consider central to a forward analysis of the company.
Domestic Franchise Stores
The decline in domestic franchise store locations has contributed to a corresponding decrease in revenues from sales to domestic franchisees and franchise fees, although the breakdown between declines associated with lower same store sales at franchise locations and closures of franchise locations is not readily delineated in the financial statements. However, it’s noteworthy that total revenue from domestic franchise locations has not declined proportionately to these two factors, suggesting that weaker locations are closing and (so far) limiting the overall effect on revenues.
International Franchise Stores
The decline in international franchise store locations, however, is rather interesting since it represents a decline of nearly 5% in the company’s international franchise store locations in the last year and more than 8% since international franchise store locations peaked around 2015. However, revenues from sales to international franchisees have not been noticeably impacted to date and have, in fact, increased in the face of these closures even after excluding the impact of increased revenues in China. The exact nature of the international franchise store base is somewhat more difficult to assess because there is not a good corresponding component of the business operated by the company as a reference, but it does suggest that increasing sales in some markets are offsetting losses due to closures of locations. The frequent assertions of the benefits of expansion into Australia and India, which strike us as tending towards the overly optimistic particularly in the near term, may be driving some of these revenue gains while not being reflected in new international locations. Regardless, the gains so far being largely at the margins, while the future international opportunity may ultimately prove significant, the time horizon for such benefits is likely well beyond that necessary for these to have any material impact on the company’s financial condition or refinancing opportunities in the immediate future.
Finally, the closure of Rite Aid (RAD) store-within-a-store locations has been driven primarily by the sale of various Rite Aid locations to Walgreens (WBA). The company provides little information on the financial characteristics of the store-within-a-store model other than to suggest in the financial statements (and in our conversations) that the relationship has a minimal financial impact on the company. However, since the store-within-a-store model is essentially a wholesale partner relationship, the company does not experience the same potential free cash flow benefit associated with the closure of company owned stores which we discuss later (since the inventory is owned by Rite Aid) while absorbing a minimal impact from lower wholesale revenues. However, one interesting question is the rather undefined benefit of revenue transfer from Rite Aid store-within-a-store locations to company and franchise stores. The recapture of revenue due to store closures in the store-within-a-store segment may boost revenues and earnings on the margin, but in either case, we don’t consider the decline in store-within-a-store locations especially material to the company.
It's worth noting that the company's manufacturing and retail agreements with Rite Aid expire in 2018 and 2019. The future agreements are currently been negotiated with Rite Aid.
The past store closures have had a rather modest impact on the company so far indicating how poorly some of these locations were performing for the company. Indeed, the information provided by the company on past store closures acts as the starting point for our assessment of the future potential impact on the company’s finances of the recently announced closure of an additional 700-900 stores.
Future Store Closures
The starting point for our analysis was to develop what is, in essence, a typical company owned store based largely on the financial information provided in the company’s annual and quarterly reports. We used this information to develop an average store by revenues and operating income, as reflected in the following table:
Source: Winter Harbor Capital
In this presentation, it’s worth noting that while the company provides an operating margin for the domestic business as a whole, it does not break down operating margin between the company owned stores and the domestic franchise locations to which it sells on a wholesale basis. The above estimate represents a best case scenario for the average company store location since, while we expect the domestic franchise store revenue base has a higher operating margin than the domestic company owned stores, we have used the consolidated operating margin. In the event we apply the international franchise operating margin to the domestic franchise revenue base and utilize the net amount for our projections, average company owned store operating income is roughly a third of that presented in this table.
We believe, based on our conversations, that while the domestic franchise segment operating margin is higher than the company’s owned stores it is not nearly as high as that for the international franchise business, falling somewhere in between. However, for our initial purposes here, the above estimate can nonetheless represent a workable starting point so long as we realize that the operating income projection is almost certainly too high and represents a best case scenario. In terms of the breakdown of expenses between cost of goods and selling, general, and administrative expenses, the exact distribution is at this time not especially important for the initial framework since we’re focused on the operating income of the average store location.
We next developed an estimate of the average annual revenues generated by the typical company owned store which has been closed over the last year by looking at the revenues the company reports as having been lost due to store closures. We performed this estimate on both the three month (quarterly) and nine month (year to date results) based on the number of store closures which occurred since the prior period and accounting for the store locations which closed during the corresponding period and projecting revenues on a full year basis based on past revenue distributions. The results are presented below:
Source: Winter Harbor Capital
In essence, the company stores which have been closed in the current year are, on average, far underperforming the company’s average company store on a revenue basis. Indeed, average revenues are as much as half or less as the average company store location.
However, it’s worth noting that this estimate can be volatile for a couple reasons. First, we don’t have a good sense of the timing of store closures within a quarter, that is, whether store closures primarily occurred at the beginning of a quarter, the end of a quarter, or some distribution within a given quarter. The timing of closures can have a significant impact on the resulting average location revenues since concentrations at one end or the other of the period can change the revenue lost during the period with early closures resulting in higher averages and later closures resulting in lower averages. The result is that a simple division of lost revenues against the number of stores closed can provide an inaccurate estimate of revenue per unit. On the other hand, given the number of closures and the wide distribution of the company’s stores, we’ve assumed for this initial framework that store closures occur reasonably consistently over the course of a quarter just as they have tended to occur in reasonably consistent numbers from quarter to quarter in the current year. The timing differences may play a role in the difference in average revenue estimates based on the time frame, but in either case, we’re reasonably confident in these average figures at least insofar as their use as a starting point. The company has, in addition, confirmed that the majority of company store locations closed on a year to date basis do materially underperform the average company store location.
The company has in the past stated that many of the costs associated with its company store locations are relatively fixed and, in light of this factor, extended the average company store calculation to establish a selling, general, and administrative cost floor. We combined this projection with the projected average revenues per closed company store location to create two template store locations, one operating at an EBITDA deficit and the other operating at a positive EBITDA level but on a negative operating margin basis, as outlined in the following table:
Source: Winter Harbor Capital
We compared our rough projections with available financial information which, while not directly comparable, at least provides a reference check for our numbers. In particular, we looked at a number of GNC franchises available for sale and the associated financial information although it should be acknowledged that company and franchise operations are different and there is a wide range of franchises available for sale all of which are not representative for various reasons.
Clearly, there is a significant amount of potential variability in these figures. Selling, general, and administrative expenses could, for example, be somewhat lower at marginally performing stores, especially if locations are in lower volume and thus lower cost storefronts. It’s also possible that the average depreciation and amortization applied to a typical store is inappropriate for underperforming stores, many of which we suspect are older stores nearing the expiration of their leases and, as a result, have already taken most of their applicable depreciation where the estimated useful life of the leasehold improvements is less than the associated term of the lease.
Finally, we used these values as a basis for calculating the financial impact on revenues, operating income, etc., three years forward (that is, after the anticipated store closures have been completed) based on a midpoint assumption of around 800 store closures. We divided the total number of stores closed between the negative and positive average stores described above based on the company’s prior comments regarding the number of company operated stores with negative EBITDA results.
The company has been somewhat imprecise in its public statements regarding store locations with negative EBITDA. In the third quarter conference call alone, the company’s chief financial officer Tricia Toliver, stated that less than 10% of company stores were operating at negative EBITDA, followed by a response to an analyst question that the percentage was 10%, subsequently apparently corrected to refer to EBIT rather than EBITDA. In seeking clarification on this comment from the company, EBITDA is the measure used for this percentage.
In the absence of precise percentages, we gave the company the benefit of the doubt and assumed that 8% of the remaining store base (as of the end of the most recent quarter) – or about 262 store locations – operates at a negative EBITDA. We thus assumed that the balance of the proposed store closures operate at a positive EBITDA though perhaps a negative operating profit due to the effect of the depreciation and amortization add back.
The calculation of the impact on net revenues was based on multiplying the number of store locations closing by the respective projected unit economics and adding back to the deduction an estimate of the revenue recapture experienced by the company. In the third quarter conference call, the company stated that the revenue recapture, i.e., revenue which transferred to other store locations and/or migrated to online sales, was above 30%. The number provides a lower threshold for calculating potential revenue capture, but how much above that figure should analysts go to get a more realistic perspective of the company’s revenue recapture potential? In addressing this question, our view was that if the revenue recapture rate has been above 40%, the company probably would have reported this higher number and the same may be said were the revenue recapture much about a third of the lost revenues. We grant that this is highly subjective – the tendency to default to round numbers or consistent increments is by no means a definitive argument, but in light of the company’s specificity in the comments, we chose a revenue recapture rate of around 32% for our calculations.
Utilizing these figures, we estimate that the net revenue loss to the company may be approximately $225 million, or about 15% of domestic company operated store revenues, with revenue recapture of approximately $72 million, for a net reduction in revenues of roughly $153 million. In the context of closing around 25% of the company’s operated store locations, a revenue loss of this magnitude is actually reasonably mild, especially on a net basis. We also find that there is not a great deal of sensitivity in the revenue numbers – in the event our average positive EBITDA store location has around 15% higher revenues than projected in our unit model, this only increases net revenue loss by about 10%, or $15 million, not in the great scheme of things a large difference relative to domestic company store revenues. We’re comfortable in this initial rough estimate with a 1%-2% potential variation in net revenue loss relative to total revenues related to store closures.
We separately estimated the associated decline in SG&A expenses using a similar approach under various conditions. In this case, we estimated that gross SG&A expenses would decline by approximately $79 million before accounting for any incremental SG&A cost associated with the recapture of revenues in other stores or through online sales.
In this respect, a judgement must be made on the cost of recapturing those revenues. The company has stated that this revenue recapture tends to be “highly accretive” which suggests that the SG&A expense associated with the recaptured revenues is low. In our discussions with the company, roughly 50% of revenues are recaptured in physical store locations implying that the balance is recaptured through online sales, though this percentage may be a very rough estimate and not very accurate. Arguably, revenues that are recaptured in other stores may very well have no additional associated SG&A expenses since store fixed costs tend to predominate or, at least, add only marginally to operating expenses. However, it’s also arguable that online sales, due to shipping expenses, etc., may have a higher associated cost that, while still advantageous from the company’s perspective relative to the comparable store operating expenses, increases the SG&A expense associated with the revenue recapture.
In order to account for this potential variability in our initial framework, we simply applied a revenue recapture cost as an offset to the SG&A expense reductions based on a percentage of the revenue recapture. We defined the associated range of SG&A expenses as a low of 0% (no cost to recapture) and a high of 20% (a marginal improvement over store operating expenses), and projected SG&A expenses under ceteris paribus conditions.
It’s worth noting, as we expect this to be a criticism of out approach, that our estimates of gross profits and SG&A expenses as separate items to not in the end significantly impact the outcomes of the analysis. Ultimately, our objective is to assess operating income (from which EBITDA is derived) since this is the metric the company has apparently established as a key component of the closure decision. In this case, although our gross margin and SG&A expense percentages may not fully reflect actual conditions of a specific unit, the operating income metric doesn’t change materially so long as the combined allocation between the two is reasonably accurate. However, this may not be fully reflected in the SG&A expense associated with revenue recapture; in the event the SG&A expense ratio for a store unit is actually higher or lower than projected, changes in the SG&A expense cost associated with revenue recapture could impact the operating income result on the margin.
We find that, after testing for sensitivity around the various assumptions and estimates incorporated in our approach, the closure of company stores will likely be in the long term modestly accretive to earnings. The following table provides a base case as well as the impact associated with various revenue recapture costs:
The revenue figure notwithstanding (which does not represent our estimate of forward revenues but acts as a consistent reference to assess the magnitude of the impact of store closures), the potential long-term earnings benefit to the company from the proposed store closures on a fully converted preferred stock basis is in the range of $0.10 to $0.20 per share. The impact on free cash flows, before considering any contribution from inventory reduction, is reasonably significant given the company’s debt load, ranging from an additional $15 million to $27 million per year. In other words, despite the ongoing nondeductibility of certain interest expenses, the projected store closures if properly implemented (even if our projections are somewhat off) should result in material improvement in the company’s financial and operating results.
However, there are at least a couple additional caveats worth mentioning in this presentation. First, on the positive end, this initial model doesn’t differentiate between the types of revenue being recaptured, that is, whether these revenues are associated with GNC brand products or third party products. It’s likely that the majority (or at least a good portion) of the 30% or more revenue recapture occurring is associated with brand loyalty to GNC brand products which could well boost gross margins and result in additional incremental improvement over the above projections. Second, on the negative side, this initial model also assumes that all of the revenue recapture is occurring to the company rather than franchises which is unlikely since franchises represent roughly a quarter of total domestic locations and may rise to about a third after completion of the store closures. The result is that franchisees may be capturing some of the benefit associated with the revenue recapture, thus limiting the benefits to the company to additional wholesale revenues. The net effect could be beneficial or detrimental from the company’s perspective although our initial impression is that the net benefit remains on the side of the company.
Cash Flow Implications
In addition to increased net earnings, the close of 700-900 company stores should also provide a materially positive cash flow impact over time primarily associated with the reduction in store inventory. In the last year, the company generated significant one time positive free cash flows by squeezing inventory out of the supply chain, but this having largely run its course, future inventory reductions will almost certainly come from reductions in the store base.
We estimate average store inventory based on the company’s finished good inventory position relative to the respective period average company operated store base. This method is admittedly rough – all finished good inventories are not committed to the company operated store base – but nonetheless provides a reasonable (if slightly inflated) perspective of the company’s overall inventory position. The result is average inventory per company operated store of around $128,000.
Source: Winter Harbor Capital
Clearly, the low hanging fruit with respect to inventory reduction was picked in 2017, resulting in the company’s significantly higher free cash flow figure for that year versus our (and the company’s) projections for the current year absent this one time benefit. The benefit from inventory management going forward will be largely based on incremental reductions in store locations rather than further significant improvement in inventory management.
However, it’s worth noting that this calculation almost certainly overestimates the amount of inventory in the company’s stores since we have attributed all finished goods to the company’s owned store base rather than allocating a portion to the wholesale segment. The allocation is not especially meaningful in terms of trend in store inventory since it would essentially represent a lateral shift in the entire chart to a lower inventory per store level. Nonetheless, it does have a potential impact in estimating the inventory reduction benefit associated with store closures.
In this respect, the relatively flat average inventory per company owned store over the course of the current year suggests that there should be material inventory reduction associated with the store locations closed year to date. A simplified calculation that almost certainly exaggerates the potential total suggests the closure of 143 company owned locations in the United States would result in potential inventory liquidation of around $18 million, yet while inventory declined in the first nine months of 2017, inventory actually rose slightly during the first nine months of 2018.
The benefit, we believe, is hidden in the numbers – with an additional investment in inventory of only $6 million for the nine month period, inventory has declined by around $12 million during the course of the current year. We suspect that much of this benefit is associated with the closure of stores.
We therefore estimate the potential inventory reduction (and corresponding contribution to free cash flows) to fall roughly within the range of $40 million to $80 million over the next three years depending on the ultimate number of locations closed and the inventory position of those locations. The variance is related to the variability in store inventory levels – lower performing stores in terms of revenue would likely carry lower than average inventory levels. However, whatever the ultimate figure, the company should clearly benefit from store closures by generating additional free cash flow for debt reduction in addition to higher operating income.
We also wouldn’t be especially surprised, given the company’s aggressive free cash flow projections for the full year, if the company sells down inventory in the final quarter to generate additional free cash flow. The fourth quarter of 2017 saw inventory decline by $31.4 million in a single quarter through much of this gain was reversed in the first quarter of 2018 by $23.8 million in additional investment in inventory.
A Second Approach
In addition to the above approach, we considered an alternate approach to determining the financial impact of store closures on the company’s financial results. In this approach, instead of establishing the company store location gross margin based on the company’s overall gross margin, we estimated the revenue base of the company’s franchise stores (assuming a degree of average revenue parity with the average company owned store) and calculated an implied cost of goods sold relative to revenues based on the company’s product sales to franchisees. The result – a gross margin closer to 52.5% - adjusts the associated SG&A expenses in order to keep the operating margin at approximately 7.5%.
In this case, a summary of the resulting operating metrics for a typical store are indicated below:
Source: Winter Harbor Capital
The resulting financial projections three years forward based on this combination of unit operating factors is as follows:
Source: Winter Harbor Capital
Clearly, this approach results in projections of higher operating income, earnings per share, and cash flow measures than the earlier approach adding about a dime per share in earnings per year and contributing about $15 million in annual free cash flows. The reason is the implicit assumption that while reductions in revenues result in larger gross margin declines, the elimination of higher fixed costs associated with store operations exceeds the loss in gross margin.
Our view is that this approach provides an unlikely optimistic outcome for a few reasons. First, the company explicitly states that the cost of sales includes occupancy costs whereas this approach to estimating gross margin clearly does not include occupancy costs. Second, based on the unit economic which are associated with reasonably known values (i.e., average closed store revenues and operating margin), the store operating expenses and associated losses for an underperforming store are significantly worse than those associated with the lower gross margin model to such a degree that closing the store early rather than waiting for lease expiration makes increasing financial sense.
A Final Note
There is one final observation worth making about the company’s closure of store locations and the revenue recapture that results from those closures. The company’s approach to reporting same store sales results, in part, results in a closed store’s sales only being included up to and including the day of closure. The effect, with respect to revenue recapture, is that revenues which then transfer from a closed store to remaining operating stores (or the online channel) are reflected as an increase in the same store sales results of the remaining stores or the online channel. The impact of this effect is minimal so long as the number of store closures is relatively small, but the closure of upwards of 25% of the company’s stores will almost certainly have an artificially positive impact on reported same store sales figures. A highly simplified example of the impact is presented below:
Source: Winter Harbor Capital
The timing of store closures within a period can impact the reported result, but it’s noteworthy that store closure activity can mask actual results, making weak performance appear less weak and marginal performance appear more robust. In effect, store closures may effectively inflate same store sales results at the margin, though to the extent same store sales figures are a focus point for investors, it’s a worthwhile observation.
GNC’s store base is shrinking faster than the headline numbers associated with the oft discussed closure of company owned and operated store locations. Indeed, erosion is visible across the distribution network in company, domestic, and international franchise, and store-within-a-store locations. The impact of these closures to date beyond the company’s owned store locations has not been particularly negative, though this bears attention going forward.
The company’s announcement of the closure of 700-900 additional company owned store locations may, however, have a material positive impact on the company’s financial and operating results. In the short term, some of these benefits will be offset by related store closing costs and will likely take some time to become apparent. In the long term, all else being equal, the overall impact should be modestly to meaningfully positive for both earnings and free cash flows, supporting additional debt repayment.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Additional disclosure: We are effectively long through short put option positions of various strike prices and expiration dates.