GNC Holdings, Inc. (GNC) is one of the market’s most heavily shorted companies with, at last report, just under 27 million shares sold short. The short interest represents nearly a third of the company’s public float despite actually being marginally less than the recent peak over 30 million shares. The short trade has been exceptionally profitable over the last few years as the company has struggled with multiple challenges, but the persistence of the short interest in the face of the significant stock price decline draws into question the rationale of initiating or maintaining a short position at the current price. The historical short narrative with respect to GNC – weak conventional retail and increasing competition – has been largely accurate. However, given the company’s restructuring, refinancing, and ongoing (though depressed) profitability and cash flow generation, does the short opportunity on a going forward basis still have a sufficiently positive expected value?
Indeed, why continue to short the company’s shares given the dwindling opportunity for gain versus the growing risks of holding a short position? The present maximum potential gain for the outstanding short interest is only around $90 million versus approaching a billion just a few years ago when the short case was significantly stronger from both a financial and retail environment perspective. In reality, as we discuss below, the potential gain is actually probably somewhat less – closer to $50 million – as we believe an outright bankruptcy in the near future is a rather remote possibility.
In contrast to the potential losses should the trade run against the short position, this strikes us as a relatively small opportunity in comparison to the risks. The easy money in shorting the company has already been made over the last two years and, in our view, the current large short interest position represents an overly pessimistic view of the company’s prospects.
We assess the potential value of a short position using an expected value model based on the valuation ranges for various outcomes developed in our prior articles. We find that an objective evaluation of the probable outcomes results in a negative expected value for short positions at the current share price, thus creating a material risk/reward imbalance and placing existing short positions at significantly greater risk of loss. We conclude that a short position in the company now entails unnecessary risks and those with a short position would be well served to seek more compelling opportunities.
In this forum, and especially with heavily shorted companies, there tends to be a binary reaction wherein articles that are not outright cheerleading a company are dismissed as drivel from short sellers of a company. In essence, there is limited room for evaluation of both the positive and negative aspects of a company which are typically myriad in these circumstances where a company is faced with unusual challenges. We’ve been healthy skeptics of some of the claims of the company’s cheerleaders as well as the company’s pronouncements when these seemed patently unreasonable. In many instances, we have been proven reasonably accurate in our assessments, notably on free cash flows, valuation of subsidiaries, etc.
However, we aren’t present in this forum to win the Seeking Alpha popularity contest, but to present our views on complex situations as we assess the relative benefits and risks associated with a company and its securities. We leave it to the reader to assess the accuracy and reasonableness of our conclusions based on their own interpretation of the available information.
Nonetheless, we reiterate that we are not – and have never been – short any security issued by the company, although this would have been a highly profitable position over the last couple years. We believe GNC is undervalued at the current price, and therefore hold an effective long position through short put option positions in the company as we have consistently disclosed in our articles. Ironically, in the context of this article, this should open us up to criticism of being biased towards the positive with respect to GNC. The comments, we anticipate, will remain highly entertaining.
We should also note that this presentation is a simplification of the more complex assessment necessary for comprehensive short analysis and expected values for the purposes of relative brevity.
Now, proceeding on…
The Prevailing Factors
In our view, there are only two material risk factors that would potentially justify a short position in GNC at the current market price, which are:
- A failure of the proposed preferred stock transaction with Harbin either due to regulatory issues or some other reason without a suitable alternative; and
- a reacceleration of the downward trajectory in physical same-store sales results that overwhelms the relatively positive recent performance of the company’s online segment.
In this article, we discuss these two factors and the potential outcomes as they relate to the opportunity (or lack thereof) with respect to a short position in the company.
A Basis for Assessing Relative Risk
In considering the risk/reward balance of a short position (or, for that matter, of a long position), it’s necessary to establish valuation bounds for the various potential outcomes. The process is inherently uncertain and subject to various assumptions, but in this article, we use the valuation ranges we’ve developed in much greater detail over our earlier series of articles about the company.
Specifically, we have previously established a valuation range (in the event of the successful closing of the currently proposed Harbin preferred stock transaction) of $3.25 to $6.50 and a valuation range (in the event of a failed transaction without a substitute) of $0.00 to $2.00. In essence, these ranges establish valuation bands for the best- and worst-case outcomes of the company relative to the Harbin transaction. The corresponding market capitalization (on a fully converted basis in the case of the closing of the preferred stock transaction) ranges between $0 and $900 million.
We also made an assessment of the most probable values within these ranges based on various factors which do not necessarily correspond to the midpoint of the ranges. In the event the Harbin transaction is approved in its current form, our bias is slightly above the midpoint of the range, at around $5.00 per share, or a market capitalization of around $687.5 million, a good part of which is represented by the infusion of cash associated with the sale of preferred stock.
In the event the Harbin transaction fails and is not replaced with an alternative transaction, our most probable value point estimate is around $1.00 per share, or a market capitalization of roughly $87.5 million, reflecting the fact that although the company would find itself in a very challenging financial position, even under this scenario, it will likely remain profitable, continue to generate free cash flow and reduce debt, and at least muddle on for some period of time. We therefore don’t foresee the company’s shares losing all their value even in the event of a failure of the Harbin transaction.
Finally, we also define a middle ground, roughly corresponding to the valuations between the two ranges for a transaction success or failure, which accounts for a modified transaction which may not be as beneficial to the company as the current Harbin transaction. A modified transaction, possibly in response to reservations on the part of CFIUS, may contain a different joint venture structure, smaller equity infusion, lower cost debt component, or some other form. This modified approach, while not allowing the company to reduce debt as significantly, would nonetheless provide a measure of relief to the company in terms of lower interest expense and leverage, allowing more financial room to maneuver.
The Harbin Transaction
The first consideration is the outcome of the currently proposed preferred stock transaction with Harbin Pharmaceuticals Group. GNC has repeatedly expressed confidence that the transaction will close by the end of the year, providing an infusion of cash which will immediately reduce debt (and allow the company to more aggressively reduce debt going forward) while providing a buffer against any weakness in the business. The transaction is currently subject to review by CFIUS, and while we have noted some potential concerns about the approval process, we believe it’s more likely than not that the transaction will ultimately receive regulatory approval and any potential concerns will be overcome.
The question then becomes, strictly on the basis of the success or failure of the Harbin transaction, is there positive expected value in a short position based on an assessment of the relative outcome probabilities? Alternately, and perhaps easier to evaluate, what is the necessary outcome probability to achieve a break even expected value for a short position?
The answer is relatively straightforward in our models since the current share price is almost exactly equidistant from our most probable values under either scenario, that is, the necessary probability is quite close to 50/50.
So, is this a reasonable probability of the outcome? In other words, are the odds of the outcome of the Harbin transaction an even split between success and failure, or is one outcome more likely than another?
It’s easy to compare this probability against the simple historical success rate in CFIUS reviews of transactions involving Chinese participants. Interestingly, despite the ongoing trade tensions between the United States and China (and the corresponding rhetoric from both sides), the approval rate of transactions involving Chinese participants has remained at about 57% during the Trump administration. A probability of success above 50% results in a negative expected value associated with a short position, so simply based on historical experience, a short position at the current share price appears an ill-advised proposition. In fact, simply using this historical success rate, even if the valuation on a failed transaction outcome were closer to $0.50 or $0.75 per share versus our $1.00 estimate, the expected value of a short position remains negative, suggesting a decent amount of downside risk.
We can expand on this assessment as well by objectively noting that, for the most part, the majority of transactions rejected (or withdrawn) involve companies with specific manufacturing or technology expertise, creating a more direct connection to national security that is more difficult to alleviate versus those which may be faced by GNC during its review. It’s arguable, therefore, that the 57% success rate, which includes the entire university of reviewed Chinese transaction, is probably still a low estimate of the actual probability, further widening the negative expected value of a short position.
It’s therefore possible to say that, from a strictly objective standpoint based on past experience, the odds of a successful transaction are greater than 50/50 and, therefore, a short position has a negative expected value when specifically considering the outcome of the Harbin transaction. Indeed, even if one believes that there is a more likely than not probability that the proposed preferred stock transaction will ultimately fail, the risk/reward balance from a short standpoint does not strike us as especially compelling.
The more challenging question, though, is how far above a 50/50 (or 57/43) ratio is the probability of a successful transaction as this is inherently a subjective evaluation. We don’t feel it’s especially necessary to make such an assessment since we have already determined that a short position has a negative expected value, but it may be useful for the second step of our assessment.
Our view, which others may debate, is that the probability of a successful transaction review is probably closer to a 2:1 split, i.e., the probability of a successful outcome is closer to 65% and the probability of an adverse outcome is closer to 35%. In this case, the negative expected value of the short position is simply larger than it is under the lower probabilities noted earlier.
We’ve previously commented that GNC is effectively playing chicken with CFIUS (though not necessarily intentionally and primarily with respect to the timing of the regulatory review process), but the same can be said for those short.
In summary, in terms of the outcome of the Harbin transaction, we find that there is almost certainly a negative expected value associated with a short position in the company’s shares.
The second critical factor upon which a short argument could be based is the future performance of the company with respect to same-store sales, particularly accelerating declines in the physical store segment. A sudden reversal in same-store sales performance would not be unprecedented – indeed, the sudden declines which impacted the company only a few years ago were part of the impetus for the company’s significant changes to its pricing strategy and rewards programs.
The changes were painful but probably necessary under the circumstances and certainly established a foundation for stabilizing same-store sales results. Physical same-store sales results remained relatively weak, but the online segment showed improved results which have greatly benefited the company from a revenue perspective over the last one and a half years.
So, from a short interest standpoint, what is the probability that same-store sales will indeed resume a steep and sustained downward trajectory? The company has clearly been incorrect in its historical assessment of the growth potential in its market, suggesting variously the possibility of positive same-store sales in the physical store segment (which has so far proved elusive) and, in the second-quarter conference call, revising its prior projection of positive overall same-store sales results for the current year in the low single digits to flat results as the best-case scenario. The company’s consistent optimism – often misplaced – has not been a comfort to shareholders as much as it’s been a blessing to those with short positions in the company’s shares.
The company’s proven unreliability in projecting forward same-store sales results, however, doesn’t mean that forward same-store sales results are about to fall off a cliff, or at least sufficiently so to do serious damage to the company’s financial condition and raise the potential of insolvency. We believe the highest probability outcome is that the company’s same-store sales results will remain flat to slightly negative on a going forward basis – not exactly an optimistic projection.
However, for the expected value calculation, we need to ask a slightly different question about the probabilities since the level of the potential decline in same-store sales results has a different impact under different financial scenarios. The evaluation then must turn to how significant a decline in same-store sales would be necessary under the different potential scenarios to put the company at significant financial risk.
In order to assess the potential answers, we utilized our existing financial models for the company which have been tested against actual results over the last several quarters to determine what level of decline in same-store sales results caused a corresponding reduction in cash flow generation and profitability that would place the company in critical financial straits.
The threshold changes depending on the various transaction outcome scenarios. The variation occurs because, depending on the proceeds received from a transaction and thus the amount of debt which could be eliminated and interest expense avoided, the company is able to absorb different magnitudes of decline in same-store sales before impairing its financial condition and ability to repay debt. The repayment of debt is ultimately the critical factor for the company’s long-term viability given the significant high rate debt load and associated interest expense.
GNC is, obviously, most at risk in the event the company does not close a corporate transaction in any form and is forced to address the full magnitude of its current debt load solely from operating results. In the event of a failed transaction, we estimate the maximum ongoing decline in same-store sales the company could absorb is in the range of -2.0%. This rate, or anything in excess of it, would place the company at risk of defaulting on the terms of its existing outstanding debt, in particular with respect to the minimum required principal payments associated with the refinanced term loans. A -2.0% same-store sales figure would, in most cases, be a modest setback, but absent some form of debt relief, even a modest decline would be sufficient to present significant challenges for the company.
However, in the event the Harbin transaction does close in its current form, we estimate GNC could absorb same-store sales declines as high as -6.0% before the company began to experience significant financial distress. A negative same-store sales figure of this amount is not unprecedented – same-store sales results were even worse in the last half of 2016, causing the company to implement aggressive changes to its pricing strategy and rewards programs. However, results in the last year and a half have remained well above this threshold, although same store sales at physical locations took a distinctly negative turn in the most recent quarter.
A modified transaction that provided some level of debt relief although not as much as the current Harbin transaction would permit the company to absorb negative same-store sales results somewhere between the -2.0% and -6.0% thresholds.
The importance of a completed transaction – virtually any transaction – thus becomes apparent, allowing GNC to remain profitable and cash flow generative despite some level of weakness in same-store sales results.
Source: GNC Financial Reports
We also very roughly estimated the impact of same-store sales declines on cash flows and profitability to assign a dollar value to the impact. The financial models suggest that declines in same-store sales results, even in the low single digits, generally have a rather modest impact on cash flows and net income. A decline in same-store sales of 200 basis points, in a typical scenario, reduces net income by around $15 million a few years into the future which represents a cumulative reduction in cash flows available for debt repayment in the range of $30 million, plus or minus $20 million depending on assumptions. This amount is not insignificant for the company, especially with debt-bearing interest rates over 11%, but at the same time is not so significant that it would threaten the survival of the company assuming completion of some form of debt-reducing transaction. It’s certainly not substantial enough to the degree that the risk/reward balance of a short position is suddenly tilted towards reward.
Indeed, in order to be meaningful from a short standpoint, declines in same-store sales would need to be in the mid to high single digits at least just to begin to justify a short position.
It’s then necessary to assign probabilities to each outcome for each specific transaction scenario. A highly simplified example is presented in the following table:
We next assign probabilities to the various outcomes – in combination with the success or failure of a preferred stock or similar transaction – to determine an expected value, a simplified summary of which is presented in the following table.
Source: Winter Harbor Advisors
In this example, we use the historic success rate of CFIUS reviewed transactions involving Chinese participants as the transactional outcome probability. The share price reflects, roughly speaking, the minimum, most probable, and maximum per share values based on the ranges developed from our valuation models for each outcome. We then assign an outcome probability for each share price associated with financial distress, i.e., that same-store sales results will be weak, flat, or improving, reflected by the range in values in each scenario. The revenue (same-store sales) probabilities vary between the transaction success and failure of share price groupings to reflect the additional sensitivity of the company to smaller same-store sales declines in the event of a failed transaction versus a successful transaction. In the case of the $3.50 share price, the bottom of our valuation range for a successful transaction, we assign a 57% probability of transactional success and a 30% probability that same-store sales performance will be on the weak end, thus causing the share price to be at the bottom of the transactional range.
The resulting expected value – $2.80 – represents the probability weighted expected value per share of the company’s common stock based on the previously established valuation ranges and outcome probabilities. In this case, a short position at a share price of around $3.00 would have a positive expected value of about $0.20, a relatively small positive expected value.
It’s worth noting that, in regards to same-store sales performance, we’re actually somewhat pessimistic on the revenue front believing that the highest probable outcome in terms of same-store sales will be flat to marginally negative.
The simplified expected value model can be expanded to include other potential outcomes, for example, the likelihood that a modified transaction would be consummated in the event the current Harbin transaction were ultimately unsuccessful. We also believe that the historical average success rate for CFIUS reviewed transactions involving Chinese participants is too low a figure to use for the probability that the transaction will be approved for the reasons previously discussed, in which case the expected value model can be modified as presented in the following table:
Source: Winter Harbor Advisors
In this case, the expected value of a short position is negative – the current share price is below the expected value, and a short position does not make economic sense.
Clearly, the assignment of probabilities to potential outcomes is both art and science. It’s art insofar as there is little certainty on specific values; it’s science in that there are available data points that can provide guideposts as to the baseline values from which it’s possible to make reasonable assessments as to whether the probability of a specific event is more or less likely than the baseline. Ultimately, this is what the assessment comes down to – is the probability that the GNC transaction will fail actually more than would be implied by historical data? Is the probability that revenues will significantly decline greater than those assigned in this analysis? It’s often easier to set aside the question of assigning specific probabilities in order to focus on the objective evaluation of the relative probabilities.
The bottom line in this analysis is that we have developed a variety of probability matrixes based on various outcome probabilities. It’s impossible to present the full extent of these expected value projections here, but we find that in essentially every case using probabilities which we consider reasonable, the result was a negative expected value for a short position.
Our expected value analysis also considered how sensitive our estimates of expected value – again, based on our valuation models – are to changes in the probabilities.
Consider, for example, a simplified scenario that is both highly pessimistic about the outcome of the Harbin transaction as well as on same-store sales. In this context, we do find a positive expected value in a short position, but even then, it’s a relatively modest positive expected value.
Source: Winter Harbor Advisors
In the event we expand this probability matrix to include a case wherein the probability of a transaction of any sort is even with that of a failed transaction while continuing to assume that same-store sales will significantly decline as the base case, the positive expected value widens, but only marginally.
Source: Winter Harbor Advisors
In other words, while the expected value calculation is clearly a subjective measure, it requires a very pessimistic view on both outcomes (of the transaction and with respect to same-store sales) to achieve a sufficiently large positive expected value to even begin justifying a short position. We believe from a purely objective standpoint, based on known quantifiable data, that these scenarios are extremely unlikely, suggesting a significant amount of risk relative to a modest potential reward from a short perspective.
Finally, it’s necessary to consider the potential impact of intangible considerations, such as shifting market sentiment, market volatility, etc. In particular, there are two significant considerations from a short seller’s standpoint.
First, the market has shown a large degree of market price volatility around retailers and, in particular, the value of Vitamin Shoppe (VSI), one of GNC’s competitors. Vitamin Shoppe has experienced a nearly three-fold increase in share price since bottoming last year and a near doubling in just the last few months. Vitamin Shoppe is certainly in a different financial position than GNC, reducing the potential comparison, but while the question of whether this volatility is justified is open to debate, there is clearly a larger potential exposure to positive surprises and shifts in already dire sentiment given the depressed share price than there is to negative surprises. The outcome is a result of the inherently imbalanced relationship between limited potential gain and unlimited potential risk associated with a short position, especially when a security is priced reasonably.
Second, account must be taken that the success or failure of the Harbin transaction is largely a binary outcome – there are few interim stages outside of a modified transaction to alleviate potential concerns raised in the CFIUS review process. The result is that one must develop a view as to whether approval of the transaction will occur and, if one is reasonably convinced that it will, it’s worth considering whether assigning a probability of 57% or 65% is truly appropriate despite the expected value standpoint. It’s nonsensical to believe that the transaction will be approved and short the stock with a negative expected value; it’s even more nonsensical to short the stock when the share price upon approval of the transaction is virtually certain to be higher than the current market price. Indeed, to maintain a short position at the current valuation, one must be convinced that the transaction will fail, a position which has little empirical support.
Our view is that the proposed preferred stock transaction (or, at a minimum, a beneficial transaction in some form) will likely receive the necessary regulatory approvals and ultimately close, although there remains timing risk which we’ve addressed in earlier articles. In addition, we believe the company is likely to experience ongoing modest declines in physical same-store sales results that are largely or entirely offset by online same-store sales performance such that the probability of revenue declines in physical stores to overwhelm online results is rather low. We also consider an outright bankruptcy of the company unlikely (though not impossible), in which case the price floor per share is probably closer to $1.00, not dissimilar from many other marginal and at-risk retailers that have sufficient wherewithal to muddle through in financial purgatory for a period of time while avoid outright bankruptcy.
We see no reasonable set of probable outcomes that would argue for establishing – or maintaining – a short position at the current share price. A short position under these circumstances would have very limited potential reward with comparatively high potential risk. Indeed, even under relatively pessimistic scenarios, such as a high probability of significant decreases in same-store sales, we conclude that the current valuation is reasonable to slightly undervalued based on our assessment of the outcome probabilities for the most critical factors affecting the company.
In short, we believe the expected value of a short position at the current price is negative and represents a losing proposition.
The probability weightings we have assigned to the critical outcomes can be debated and each should develop one’s own view of which outcomes are more likely than not and, moreover, how much more likely those outcomes are for the company. However, our analysis suggests that the probability weightings are tilted towards the modestly positive.
While there was a highly compelling short argument against GNC when the shares were trading at $15.00 and above, the balance between risk and reward from a short perspective has shifted dramatically over the last two years. The potential gain has dwindled (and, in our view, become markedly negative) with an exceptionally pessimistic view necessary to maintain a short position under the current circumstances.
We are, of course, somewhat biased towards the positive view based on our perspective and valuation models. Clearly, at least some of those short hold different views of either the potential downside (or upside) under the various outcome scenarios and/or the relative probabilities of those outcomes. We therefore don’t entirely discount the possibility of a short argument – not considered here – which we’d be interested in entertaining should one be presented in the comments on another article. In this regard we as always look forward to feedback.
However, we fail to see any such scenario and, even were some of the company’s shares shorted as a hedge against, for example, debt, the potential gain is likely insignificant against the value of the hedged asset.
Ultimately, when it comes to short positions, we are very conservative in our risk assessment given the inherent risk imbalance associated with limited gain versus theoretically unlimited risk. It’s unlikely, as we’ve noted before, that the company’s shares will change significantly until the outcome of the Harbin transaction is known, but even under this condition, a 50/50 coin toss as the best possible outcome for a short position is an exceptionally poor scenario.
The ongoing high level of short interest in the company represents, in our view, an overly pessimistic viewpoint that, even were it to come to full fruition, represents an exceptionally modest gain in comparison with the very real (and in our view probable) potential for loss.
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.