GME: recommending a trade (long)

After taking a first look at Gamestop at $5.50 and passing, the stock is down 28% to $3.98. I recently took another look, refined some of the analysis, and initiated a small trade with room to add more.

Although I continue to believe that GME is facing significant challenges to its business model, I’m slightly more constructive in certain areas and do believe that you’re finally getting properly compensated for these uncertainties at the current price. I also believe that the prospect for near-term capital return should provide somewhat of a floor to the stock near these levels. I’ve refined some of my earlier math and will share my high level analysis below.

It’s also worth noting that Michael Burry (from The Big Short) owns 2.75mm shares and has recently written a letter to the Board advocating for a swift completion of GME’s existing buyback authorization in the open market. The large buyback capacity and high short interest create optionality from a technical perspective. You can read his letter below:

Free cash flow profile

GME continues to face a multitude of near-term earnings issues, including: (1) declining hardware sales ahead of the new Sony/Microsoft console releases in 2020/2021, (2) a tough 2019 AAA software line-up vs. last year, (3) pre-owned revenue tending lower (following new software sales on a lag), and (4) management’s recent inability to reduce SG&A at the same pace of revenue declines (leading to negative operating leverage). All of these factors will continue to pressure results through the rest of 2019.

On the flipside, there are some positives: (1) trough 2019 hardware sales should grow materially with new console launches, (2) accessories revenue continues to grow (led by strength in headset sales), (3) collectibles revenue continues to grow, (4) the new CEO has suggested that SG&A is just now getting a fresh look (although perhaps not at the pace I’d prefer), and (5) the Nintendo Switch’s strong unit sales are continuing, driving a growing install-base that indexes much higher on physical discs (due to a small hard drive). It’s also worth reiterating that both Sony and Microsoft have confirmed that their new consoles will have physical disc drives. I don’t consider Stadia a material threat.

For perspective, the last console launch cycle was in 2013, so 2012 was the prior cycle’s “trough year” and is comparable to 2019. GME’s hardware sales were similarly weak in 2012, however sales grow by ~50% through 2014. If GME’s hardware sales grow at comparable rates off of my projected 2019 trough, then GME should generate an incremental ~$575 of hardware revenue and ~$50mm gross profit at a ~9% gross margin. Collectibles continue to drive at least $20mm of incremental gross profit per year. These two growth categories, along with continued SG&A reductions, have the potential to absorb reasonable further declines in new software and pre-owned sales. After weighing the various positives and negatives, I conservatively forecast GME producing ~$100mm of FCF in 2020 before any capital allocation initiatives. Based on the current shares outstanding, this is $1.10/share.

How should an investor feel about that FCF production? The last console cycle would suggest that gross profit should be relatively flat in the cycle’s initial years. Industry data also suggests that the migration from physical discs to digital downloads has mostly plateaued for AAA games, providing GME some relief from that prior headwind. Fortnite, the biggest free-to-play game that was viewed as an existential threat to GME’s traditional AAA model, has already seen users peak and decline. I also found increased excitement around many unreleased AAA games at this year’s E3 conference, providing a roadmap to better lineup comparables in 2020-2021.

GME’s PowerUp loyalty program, a key competitive advantage, was also recently redesigned to better fit customer buyer habits. I expect these loyal customers to continue prioritizing purchases at GME to maximize reward points and trade-in economics. For perspective, my rough math suggests that PowerUp members generate ~$9 in annual EBITDA vs. only ~$2 for the non-member customer. I think PowerUp customers generate ~80% of GME’s total EBITDA. These customers are consistently more active trading-in games, and GME “shares” the value of the high margin pre-owned product with the customers through rewards. By leveraging GME’s competitive advantage – its leading two-sided trade-in network and inventory – it continues to cost-effectively create degrees of customer captivity to help mitigate secular pressures. The key will be GME’s success in evolving rewards to keep up with customer preferences and perception of value.

Regardless of GME’s ultimate fate, I think the above factors are suggestive that GME has a decent potential to generate a relatively flat FCF profile in the early years of the new console cycle. After that, it’s anybody’s guess. But I do think that investors can expect to see at least ~$100mm of FCF generation for a few years.

Balance sheet

Following my trough 2019E holiday season and assuming no further capital allocation activities, I expect GME to end Q4 2019 with ~$1.3bn of cash. This amount is much higher than Q1’s $543mm balance, however it overstates GME’s excess cash due to the seasonality of working capital – operating working capital actually peaks in Q2 before turning negative in Q4. Game manufacturers generally “fund” the retailer’s holiday inventory build (GME extends payables) and GME further benefits from customer financing in the form of holiday gift card purchases, which generally aren’t redeemed until Q1. From that $1.3bn overstated cash balance, I think ~$700mm should be conservatively earmarked for seasonal peak-to-trough working capital needs. In addition, GME has held at least ~$125mm of cash on its balance sheet in recent years. This implies that GME should end this trough year with ~$450mm of truly excess and distributable cash.

How will the cash be used? The Board has already approved a $300mm buyback authorization, of which $62mm was spent on the recent tender at $5.20/share. Given the further decline in the stock price, I expect management should only feel more compelled to finish this buyback. I questioned the dividend elimination in my original post, but I’ll refocus this post on what I expect to occur. This leaves $238mm of remaining buyback capacity, well below our estimated year-end excess cash of $450mm. This also means that management could comfortably prepay $200mm of bonds in the near-term, or ~40% of the remaining 6.750% notes dues 2021 ($472mm outstanding). This would bring gross leverage to ~1.0x – a reasonable level.

Capital return implications

We can now think about capital return impacts and how this could impact GME’s trading. We know that GME plans to return at least $238mm to shareholders in the form of a buyback. Based on the current shares outstanding, this is ~$2.60/share. This means that GME’s stock shouldn’t decline below $2.60, because at that level GME could theoretically repurchase 100% of the float from its existing approved authorization. Put differently, an investor’s view on terminal value doesn’t matter below $2.60. This should generate buying and short-covering activity well before the stock ever hits $2.60, keeping the stock above $3.00 in the intermediate term.

Burry wants GME to utilize its relatively high trading volumes to quietly facilitate a buyback in the open market. Although this would minimize the price paid (and maximize the pro forma value), it’s potentially unrealistic and the company is more likely do another tender. Given the size of the desired buyback, the company would probably offer a healthy premium to encourage participation.

We’ve already established that the size of the buyback authorization should keep shares comfortably above $3.00, and let’s assume that the shares simply remain at current levels ($3.98). Let’s then assume that the company offers a traditional tender at a ~30% premium vs. today’s price, or $5.20/share (conveniently in-line with the prior tender). GME could buy-back 46mm shares at that price, or 50% of the outstanding float. That level of premium would probably be enough to drive a high participation rate – I’ll assume 80% of shares are tendered. This implies that a subscribing holder would ultimately receive $5.20/share for ~60% of their tendered shares. The remaining ~40% of one’s shares would remain outstanding in an illiquid and potentially interesting stub.

A GME holder could always not tender their shares if they truly found the back-end stub compelling, although this would be unusual at a 30% premium. This is because tender stubs generally trade at a discount to their pre-tender prices. In most instances, a company just bought back a material portion of their shares at a premium to the market’s existing estimate of intrinsic value, so theoretically the back-end stub should simply be worth less. Company’s nonetheless pursue tenders on the view that total “package” to shareholders (the tender consideration received + pro forma stub value) will combine to be worth more than the existing trading price. This is because large tenders generally lever the balance sheet and drive material accretion to the stub’s pro forma earnings per share. If the earnings accretion is large enough, this can provide support to the stub’s pro forma trading value and mitigate the expected back-end trading decline. In GME’s case, the math can get a silly due to the magnitude and potential accretion of the buyback at these levels.

Let’s think about GME’s stub. I think GME can do ~$100mm in 2020 FCF, with a cautiously optimistic view that there’s some reasonable duration to this level given the new console cycle. The street is more optimistic at ~$135mm, but I’ve attempted to start from a conservative baseline level. Retiring $200mm of bonds would provide $10mm of after-tax interest savings, driving our FCF estimate to $110mm. Our $5.20 tender would repurchase 46mm shares, resulting in 45mm pro forma shares outstanding. This means that the stub should produce $2.44 of FCF/share ($110mm / 45mm), for 120% accretion vs. our standalone FCF/share estimate of $1.10! This sounds crazy, but the math of deploying significant excess cash to buy back 50% of the float at a 21% FCF yield ($1.10 / $5.20) is simply compelling.

Let’s assume that an investor buys stock today at $3.98, this illustrative tender offer is subsequently announced, and you conservatively tender your shares to reduce risk. This means that you’d receive $5.20/share for ~60% of your shares, worth $3.12 for every share you own on a blended basis ($5.20 x 60%). Your remaining basis in the back-end stub would therefore be $0.86 ($3.98 less $3.12). Right away, 78% of your initial bet would be de-risked, giving a cheap look at the stub.

At that point, 40% of the investor’s initial shares would remain outstanding in the stub. Relative to an investor’s initial position, they would be looking at $0.98 of FCF/share ($2.44 x 40%). By “creating” the stub at $0.86 and retaining a residual position in the equity, you would therefore get a back-end look at this business at 0.9x FCF ($0.86 / $0.98)! Although GME would be higher risk following this transaction – it just spent $238mm of excess liquidity to repurchase shares – the economics of a buyback at current levels is undoubtedly unique.

Where should the stub trade? I have absolutely no idea, but I clearly think that 0.9x is too low at the beginning of a console cycle with only ~1x of gross leverage. To the extent the new CEO can convince holders that GME can simultaneously ride the new console cycle and find new SG&A reduction opportunities, I think the stub could eventually trade as high as 5x FCF. At that level, the stub would eventually re-rate to $4.88 per original share ($0.98 FCF x 5x), implying $8.00 in total value for the package ($3.12 cash + $4.88 stub).

This probably begs the question – why would holders tender in this example? You could alternatively withhold tendering, receive no cash, but get exposure to 100% of the upside from the stub’s potential re-rating. At 5x FCF, the stub could be worth $12.20 to non-tendering holders ($2.44 x 5x), well above the $8.00 package estimate in the prior paragraph. While this is all theoretically true, I don’t think many are willing to pass up a guaranteed 30% gain on their tendered shares to retain full exposure on the stub re-rate. Most would take risk off the table. There could also be a gap between the timing of a tender and the validation that GME is indeed well positioned to generate ample FCF through the early part of the next console cycle. The stub could eventually work, but not right away – this is another risk that many are unwilling to underwrite. The initial stub trading value could remain highly controversial, although it’s worth remembering that GME’s stock doubled during the last console launch cycle.

Let’s also consider the technicals. Currently 58mm shares are sold short. This is 64% of the current float (an extremely high number), but shorts get some comfort from GME’s high trading volumes at a somewhat digestible 8 days-to-cover. Nevertheless, these 58mm shares will need to be covered at some point. Higher borrow costs during this period could also pressure shorts. Put differently, I think there exists 104mm of natural buyer appetite for shares over time (58mm short-covering + 46mm tendered shares), or 114% of the current float. If GME is aggressive in pursuing the remaining buyback authorization, the technicals are absolutely there for a potential squeeze.

We can now try to frame the risk-reward. I think it’s unlikely for the stock to trade near $3.00 in the near-term given the size of the outstanding buyback authorization. As a result, until management announces a tender (or gradually buys back stock in the open market), I don’t think you’re wearing excessive trading risk. I’d expect management to eventually announce a tender, although realistically this could occur any time from tomorrow until after the holiday season (5 months later). But not until that time will an investor be required to decide if they want to participate. At that point they could tender and receive at least $3.12 in hard cash plus potential upside of $4.88 on the stub “working” for $8.00 in total value. But even if the stub eventually trades to $0, a participating holder would still get their $3.12 of cash, suggesting 22% of conservative downside risk in this scenario. On the flipside, if interim data points are bullish between now and then, an optimistic holder could withhold their shares and get complete exposure to a full stub re-rating, estimated at $12.20 per share.

Taken together – assuming my illustrative tender indeed occurs – a GME holder today is effectively risking $0.86 ($3.98 less $3.12) to have the optionality to subsequently play for $8.22 ($12.20 less $3.98), implying a ~10x risk-reward ratio. This is potentially valuable because you’re only paying $0.86 for two options: (1) the option to tender shares and get a stub re-rating on your remaining position, and (2) the additional option to later decide that it’s fully worth betting on the stub for more upside.

What would I do today? Despite creating my illustrative stub at 0.9x FCF, I’d probably tender my shares and mostly de-risk the bet. There’s a lot of illustrative math in this post, but that idea of cheap optionality on this business post-recap is the main point. And although I’ve laid out a specific tender example to walk through value math, I could also see GME’s stock trading well as investors simply bid up shares after spending more time appreciating the excess cash and float reduction potential at current levels.