Summary RIOT's recent out-performance may be driven by a combo of no more tax loss harvesting, significant BTC purchases, and excitement around AI/HPC developments. Despite a strong balance sheet, RIOT's reliance on dilution and speculative BTC holdings raises concerns about long-term shareholder value. I expect Q4 earnings to be lackluster due in part to minimal power credits and no realized revenue from BTC sales. However, BTC per share has increased year over year. Given the dilution concerns, RIOT shares are probably best used as a trading vehicle. But valuation multiples over BTC stack should guide entry and exit points. In late October, I reiterated a 'hold' call on Bitcoin ( BTC-USD ) miner Riot Platforms ( RIOT ). In the closing summary of that article, I left readers with the following: I'm struggling to see the bull case for RIOT relative to other opportunities. Perhaps the best thing RIOT has going for it is balance sheet strength with a large cash position and no debt. But the company now seems pot-committed to holding Bitfarms stock, which is eating up a considerable portion of Riot's capital. In the time since that note was published, RIOT has actually out-performed most mining stock peers: Data by YCharts Though it should be noted that this out-performance started in January rather than November, I still have to wear it here. RIOT has beat miners I myself am currently holding but has trailed spot ETFs so far this year. Following yet another serving of 'humble pie' from a market that occasionally makes my calls look bad, I am forced to reconcile how RIOT could be beating peers and only slightly losing against the Fidelity Wise Origin Bitcoin ETF ( FBTC ) over the last three and a half months. I have three theories. In this upstate, we'll look at what, I think, might be driving RIOT's recent out-performance and preview what, I believe, will be a lackluster quarterly report for Q4 later this month. Why The Recent Out-performance? As I said above, I have three theories regarding RIOT's performance so far this year. Each of which could be contributing to RIOT beating peers to begin 2025. The first is fairly simple; tax loss harvesting is over. RIOT's share price fell by over 30% in 2024. 2025 brings a clean slate and a new wave of speculation. Second, Riot has joined the bandwagon in using convertibles to buy Bitcoin . Show me on the chart below where RIOT started buying BTC: Company Filings, Author's Chart December resulted in an enormous spike in BTC held on the balance sheet from 11,425 BTC up to 17,722 BTC in just one month. Stripping out the 516 BTC that Riot mined in the month, the remaining Bitcoin was purchased with $579.2 million in net proceeds from a senior convertible note offering in December. By my math, Riot paid around $100k per coin for a little over 5.7k BTC, which though currently underwater, is actually cheaper than Riot's true break-even cost which was approaching $130k in Q3 by CoinShares estimates: CoinShares Regardless, at the end of December RIOT was able to grow its BTC stack by 70% quarter over quarter, and it did so mainly from convertible note issuance rather than hitting the common stock ATM. The third theory is excitement about potential HPC developments. If we look at RIOT against Bitcoin mining peers over the last year, it's very clear to me what had been driving the out-performance, high performance computing commitments during the AI hype cycle: Data by YCharts The chart above shows three obvious standouts in 12 month mining returns; Core Scientific ( CORZ ), TeraWulf ( WULF ), and IREN Limited ( IREN ). These happen to also be the three public BTC miners that, I think, have leaned the hardest into AI/HPC revenue as secondary or tertiary revenue streams. Following an announcement in January that Riot Platforms would begin evaluating the potential for 600 MW of AI/HPC power at Corsicana, the company has essentially confirmed these efforts will move forward by bringing three new faces to the board of directors who have experience in AI/HPC conversion, data centers, and real estate. Specifically regarding BTC mining to AI/HPC infrastructure conversion, joining the board would be Jamie Leverton. If that name seems familiar, Leverton was formerly the CEO of Hut 8 ( HUT ) when the company purchased TeraGo in an effort to bring HPC revenue into Hut 8 well before it was fashionable in the mining space. Q4 Earnings Preview Company Filings, Author's Chart During the three months ended December, Riot Platforms mined 1,516 BTC and held all mined-BTC for the company's HODL. A segment that had been a significant driver of top line revenue during the summer of 2023 - power credits - has come nowhere close to achieving those mid-2023 highs: Company Filings, Author's Chart With just $3.5 million in combined power/DR credits, HPC still being in an exploratory phase, and revenue from selling BTC effectively being zero since February 2024, Riot is essentially a BTC HODL play with stock performance likely being reliant on two things; the price of BTC and how much BTC each RIOT share represents. In the company's December production update, Riot shared a BTC per 1 million fully diluted shares outstanding figure of 44.3. I find this metric to be quite helpful. That said, it was not included in the January update and as far as I can tell it hasn't been featured before the December production release. Still, it allows us to assess RIOT's dilution expectations for the Q4 report: BTC Holdings BTC per 1 million FD Shares FD Shares (millions) Closing Share Price Closing BTC Price December 2023 7,362 31.8 231.5 $15.47 $42,241 November 2024 11,425 31.6 361.6 $12.65 $96,456 December 2024 17,722 44.3 400.0 $10.21 $93,371 QoQ 55.1% 40.2% 10.6% -19.3% -3.2% YoY 140.7% 39.3% 72.8% -34.0% 121.0% Source: Riot Platforms, TrendSpider, Author's Calculations At 17,722 BTC held on the balance sheet and 44.3 BTC per every million fully diluted shares outstanding, we get 400 million shares outstanding at the end of December. In my view, this number is probably more important than something like revenue or even net income because it's quite clear that Riot is funding operations through dilution rather than sales. The question is whether or not this will ultimately be beneficial to common stock shareholders over time and I think the verdict is still unclear on that. Take for instance the 140.7% year-over-year growth in BTC holdings against just 72.8% growth in fully diluted shares. Considering BTC has more than doubled in price over that same time frame, one would expect the price of RIOT to have increased year over year but it actually fell by 34%. Why? Multiples are important. Even if one takes the view that RIOT is a great asset, investors should still aim to buy great assets at decent prices. Data by YCharts While RIOT is far from the worst multiple expansion offender in the Bitcoin mining stocks, the 2.5 price to book ratio at the end of 2023 was elevated compared to how the market was willing to value RIOT shares over the last year. Furthermore, the current 1.5 price to book reading is indeed far from the infotech sector median. But RIOT is not a cheap stock based on earnings and sales multiples: Seeking Alpha We can reasonably deduce that this overvaluation picture won't change as a result of earnings later this month because we know the company didn't sell Bitcoin, didn't produce meaningful power credits, and is pre-revenue in an AI/HPC segment that may already be an uphill climb if DeepSeek has indeed changed the capex needs for AI and LLM training. Riot Platforms And remember, Riot Platforms is planning for expansion to 46.3 EH/s by the end of this year and 65 EH/s next year despite open market purchases seemingly being the more economical way for the company to acquire BTC given historical COGS and opex. At 18,221 BTC at the end of January and a $98k BTC price, Riot's Bitcoin is worth approximately $1.7 billion effectively backing a $4 billion market capitalization. Closing Summary Data by YCharts I'll reiterate something that I've said about Bitcoin mining stocks for quite a while; my view is that these are probably not great long-term investments and I think the long-term performance against Bitcoin shows that simple and clear. Since Riot first entered the cryptocurrency business back in September 2017, the stock has benefited from large booms and busts in line with Bitcoin's price gains. That is no longer the case if this halving cycle is an indication and I suspect ETFs are a big catalyst here. That said, RIOT shares can still be used as a trading vehicle. It is up to each individual trader or speculator to assess why they'd entertain a position in the stock should they choose to do so. For me, the only real fundamental reason to go long RIOT is because the company has a large stack of Bitcoin. Thus, if you're considering the trade, I'd be mindful of valuation multiples over that stack and time entries and exits based on those trends.