TeraWulf: Sound Strategy But High Valuation

Summary TeraWulf has made strategic moves, including selling its Nautilus stake and ending a lease early, to focus on HPC/AI data centers. The company is transitioning from Bitcoin mining to high-performance computing, with plans to announce its first HPC client by year-end. Hedge funds, including Stanley Druckenmiller’s Duquesne Family Office, have increased their stakes in WULF, indicating institutional interest. Despite potential, WULF remains speculative; I maintain a ‘hold’ rating until further progression in the HPC business is evident. My last article covering TeraWulf ( WULF ) for Seeking Alpha was in late-August. For the third time out of five instances previously covering the stock, I concluded WULF was a ‘hold.’ Among other things, part of the focus of that article was Stanley Druckenmiller’s Duquesne Family Office opening a stake in WULF. Since then, WULF shares have ripped higher by about 50%: Data by YCharts Not only has WULF appreciated quite spectacularly since late-August, but it has done so while also outperforming larger mining stocks and the CoinShares Valkyrie Bitcoin Miners ETF ( WGMI ). Quite a bit has happened over the last 3.5 months. In this update, we’ll go over the company’s recent strategic moves, production results, recent trends in institutional buying, and take a look at WULF’s current valuation. Strategic Moves October was busy for TeraWulf with multiple announcements. First, the company sold its 25% stake in the Nautilus joint venture to Talen Energy in a deal valued at about $92 million. Per the announced agreement , TeraWulf received $85 million in cash and $7 million in mining machines. The company said it will use the proceeds from the Nautilus sale to build out its 20 MW CB-1 facility at Lake Mariner. This is part of a core strategy that is centered around expansion of the business into HPC/AI data centers. On the expansion point, TeraWulf also entered a new 35-year lease agreement with Somerset Operating Company, LLC at Lake Mariner – which effectively ended the original Lake Mariner lease from less than four years ago ten years early. This is an interesting deal for a few reasons. First, TeraWulf CEO Paul Prager owns Somerset. Which could theoretically be seen as a conflict of interest, especially since TeraWulf paid Somerset’s parent company 20 million WULF shares and $12 million in cash to terminate the original lease. Per the deal, Somerset’s parent can only sell the WULF common shares after lockup periods of 12 months (5 million shares) and 18 months (15 million shares). Given Prager is already a large holder of WULF stock, this new lease agreement could also be viewed as an alignment of Prager’s interests. The justification for ending the original lease early has been that the first agreement had an escalation in payments that is no longer in the new agreement. TeraWulf’s new lease was reportedly negotiated by the Audit Committee of the Board of Directors with consultation from independent legal counsel. The Company’s financial advisor deemed the deal to be fair from a financial standpoint. Roughly two weeks after the new lease agreement, TeraWulf’s Board authorized a $200 million share repurchase agreement that would be effective through the end of 2025; of which $115 million has already been deployed via the purchase of 18 million shares. Finally, on October 25th TeraWulf announced the closing of a $500 million senior note offering at 2.75% due in 2030. As just mentioned, $115 million of this note offering was already used to repurchase WULF shares. November Investor Deck Slide 13 (TeraWulf) The company referenced slide 13 from the November investor deck a few times during the Q3 call, and it’s notable because it shows a $387 million unallocated cash position for 2025. TeraWulf is anticipating $100 million in 2025 cash flow with $347 million in prepay and project financing to grow out HPC/AI compute. Prager said on the Q3 call that TeraWulf is in advanced discussions for the company’s first HPC client and expects to make an announcement by the end of the year. This would be a big deal and I think highlights where TeraWulf is ultimately going. Bitcoin Production What seems quite clear at this point in time, at least to me, is TeraWulf is planning for a potential life after Bitcoin ( BTC-USD ) mining. We can see these signals through not just the strategic moves in October, but also through what leadership is overtly saying during conference calls. TeraWulf seems well aware of the mining economics of Bitcoin, and that’s why the growth area has been identified as HPC/AI. CFO Patrick Fleury from the Q3 CC: certainly the next 500 megawatts of expansion for us is in high power computing AI. And that will probably coincide with the next halving. So we’ve got plenty of running room and then time to decide whether we keep mining Bitcoin or not. But we’ve got plenty of time to play that out. For the time being, mining Bitcoin is potentially a means to end for the company. TeraWulf has one of the most efficient operations judging by BTC mined per average EH/s as of November 2024: November 2024 EH/s BTC BTC per EH/s MARA 46.1 907 19.67 WULF 5.9 115 19.49 HIVE 5.3 103 19.43 IREN 19.7 379 19.24 RIOT 25.8 495 19.19 Source: Company press releases, analyst’s calculations Readers should recall that TeraWulf has never adopted a “HODL” strategy, as the company has generally sold all production during the course of the month. That production has been slipping since the end of Q3: Monthly BTC Mining Production (Author’s Chart, TeraWulf Filings) In the month of November, TeraWulf mined 115 BTC at a reported energy cost of $41k BTC. The average operating EH/s in the month was slightly down month over month, from 6.8 EH/s in October to 5.9 EH/s in November. During the month, TeraWulf had a planned outage while upgrading machines and connecting high-voltage redundant power feeds. In the November investor deck, TeraWulf guided for 465 BTC to be mined in Q4-24. Given just 265 BTC mined between October and November, I question whether the company can get there: Data by YCharts With the recent surge in the price of Bitcoin over the last couple of months, global network hash rate has exploded to about 800 EH/s. The growth in network hash has produced yet another all-time low in BTC per PH/s per day – which currently stands at slightly above 62,000 sats. The good news for miners is the dollar-denominated hash price is near a 3-month high at $60: Hashrate Index It should be fairly obvious at this point why TeraWulf is positioning for a future that is more reliant on high performance compute rather than mining Bitcoin. The latter of which is a very challenging business and one that lacks stability in revenues. Hedge Fund Positioning & Valuation In August, I noted that Stanley Druckenmiller’s Duquesne Family Office had taken a stake in WULF stock and cautioned against fading the legendary trader. Since then, we’ve seen Duquesne increase its stake in WULF from 2.1 million shares to 2.9 million in just a single quarter. It’s important to remember that these are lagging reports, and we won’t know if any of these shares have been sold until well after Q4 is over. But we can also look at hedge fund purchases for WULF against similar trends in other Bitcoin mining stocks: Shares Held By Hedge Funds (millions) Q2-24 Q3-24 QoQ Bitdeer ( BTDR ) 5.62 8.96 59.43% Core Scientific ( CORZ ) 88.59 110.84 25.12% Hut 8 ( HUT ) 28.91 32.64 12.90% Cipher Mining ( CIFR ) 58.26 62.06 6.52% CleanSpark ( CLSK ) 66.45 70.41 5.96% TeraWulf 125.15 130.02 3.89% Bitfarms ( BITF ) 27.74 27.39 -1.26% MARA Holdings ( MARA ) 114.63 113.01 -1.41% Riot Platforms ( RIOT ) 89.12 86.35 -3.11% IREN Limited ( IREN ) 50.47 26.67 -47.16% 10 Stock Average 6.1% Source: HedgeFollow The table above shows the quarter-over-quarter change in hedge fund ownership of top Bitcoin mining stocks based on market capitalization. At 3.9% QoQ growth in shares held by hedge funds, TeraWulf wasn’t the biggest beneficiary of investment capital last quarter and actually slightly trailed the 10 stock average of about 6%. However, at 385.9 million shares outstanding at the end of Q3, hedge funds owned 33.7% of all WULF shares. This was well ahead of the 22.4% average of the stocks sampled in the table above. The question that I keep coming back to is what is the right valuation for WULF shares today? HPC/AI compute is clearly the direction the company is going, but there hasn’t been any realized revenue from that segment yet and, frankly, there is no guarantee there will be. To me, it still makes the most sense at this point in time to value the business based off what the company is producing from BTC mining. So let’s make some assumptions about what WULF can realistically generate based on Bitcoin prices. TeraWulf Market Capitalization $2.6 billion Estimated Monthly Production 250 BTC FWD P/S At $100k BTC 8.7 FWD P/S At $150k BTC 5.8 FWD P/S At $200k BTC 4.4 Source: Analyst’s table and calculations TeraWulf’s three-month average between September and November was 147 BTC mined, and it did so on average of about 7 EH/s each month. Let’s assume the company gets to a realized average hashrate of 13.2 EH/s in Q1 2025 and can continue to mine 19 BTC per EH/s. That would put TeraWulf at about 250 BTC per month. Shareholders are paying nearly 9 times forward BTC sales for WULF at a $100k BTC price. Even if Bitcoin doubles from here, WULF shareholders are still paying a premium over the 3.3x info tech sector median forward price to sales multiple. Closing Summary TeraWulf is a very interesting company, to say the least. While there are certainly some Bitcoin miners-turned HPC companies that are buying GPUs and going with more of the ‘compute as a service’ business route, that is not what TeraWulf is doing here as the company is focusing more on scaling power infrastructure for the AI hyperscalers. I think TeraWulf’s approach is probably going to be the better long-term strategy, but that’s purely a guess on my part. All this said, I think a lot of potential has already been priced in to the stock based on the current business model. I could certainly entertain taking a position in WULF again following any further progression in the company’s HPC business. But for now, I’m going to reiterate my ‘hold’ call on WULF. I suspect the stock is still marvelous for trading both Bitcoin and HPC narratives in the zeitgeist. But as a long-term investment, it’s still very speculative.

Source: Seeking Alpha