Description

We are going to make this post quick as we see it as very timely with catalysts that could come at any point and high beta stocks, which we believe could have a break-out.  Also, we think ondacapital's recent CORZ post on June 11th lays out the CORZ investment case well, which underlies the theses behind WULF and IREN, where we see 132% and 162% upside respectively. 

The Opportunity

As we are all likely aware at this point, there is a significant rush by Cloud providers to secure AI compute power.  This is manifesting itself through stronger demand for power and data center capacity.  We have invested in this theme across a number of names such as VST, AES, NEE, CEG, TLNE, etc but these stocks are becoming increasingly crowded, along with names like VRT, ETN, etc.  However, we think investors are just beginning to wake-up to how this theme impacts the Bitcoin (BTC) miners, who are sitting on MWs of tappable cheap power, data center build expertise, and very attractive evaluations.

To capture this demand, certain BTC miners such as CORZ, WULF and IREN are transitioning some of their asset base toward AI / High Performance Compute (HPC).  With WULF, we see the potential for a 100% transition. On the back of the recent CORZ / CoreWeave deal announced on June 4th, WULF, IREN and CORZ and sell-side research firms have all discussed a significant amount of increased inbound on the BTC miners from institutional investors. 

Importantly, a Cloud provider by leveraging a BTC miner can greatly accelerate their time to market, which Morgan Stanley estimated in a recent report to alone be worth $5-12 million per MW of data center capacity.  Importantly, this report was before the CORZ / CoreWeave deal, which we believe has repriced the whole space materially higher. 

We believe IREN and WULF represent the best way to play this theme given attractive assets for AI / HPC, higher-quality management teams, upside potential based on an asset build, WULF's insider ownership of ~1/3 of shares outstanding, and both IREN and WUF have discussed the value creation potential of pivoting toward HPC.  We also prefer the companies that are looking for deals like we saw CoreWeave versus buying their own GPUs and renting them out themselves - a more challenging and capital intensive biz model.

Ondacapital discusses this deal well and the implications for CORZ so we won't repeat those here.  We recommend investors listen to CORZ two hour Inv Day presentation on June 12th, which had a number of takeaways for us, including:

1. they discussed the 80% margin as fully loaded for overhead so we see this as the EBITDA margin vs gross margin;

2. they discussed this deal as a baseline for future deals ("so this is kind of the starting point for where we believe our contracts are, and we're going to continue to seek out contracts that look very similar to this type of deal");

3. It was an impressive presentation and hard to see why the valuation gap to a Company like DLR which trades at 23x NTM EBITDA shld be very significant (understanding that DLR's REIT structure is an advantage)

We think it's worth re-highlighting that CoreWeave is paying for all the CapEx above $1.50 / MW and CORZ at their investor day said:

"We ran a very competitive process on those 500 megawatts. You can tell by our deal economics that are relatively shaking to the entire data center industry in terms of what we were able to execute on. We're going to continue to build on top of what we did on our first 200-megawatt deal and continue to find clients that are willing to pay for the CapEx upfront."

A big issue for BTC miners has been a high cost of capital.  But by transitioning to HPC, there are a number advantages, such as: a) customers willing to fund the transition and growth, b) the companies will be able to lever up the EBITDA stream for HPC, which they couldn't with BTC mining, and c) of course the multiple the market will pay for an HPC EBITDA stream vs a BTC EBITDA stream is significant.  This allows for a very simple Financial Engineering 101 opportunity to massively increase the Enterprise Value by transitioning from BTC to HPC. 

Additionally, we've been working on the economics of what a CoreWeave cld see on a deal like they signed with CORZ.  Initial work seems to imply that Coreweave could have less than a 2yr payback on the CapEx.  Consequently, we believe the CORZ, WULF, IREN's of the world have pricing power given the demand backdrop, lack of supply, and attractive economics the customers see. 

We believe WULF and IREN are the two best ways to play the theme right now.  We evaluated them on an asset by asset basis using the CORZ / CoreWeave deal as a baseline.  While we are discounting the assets compared to the CORZ / CoreWeave deal, we actually think IREN and CORZ assets are comparable.  WULF also may have the most attractive asset with their 500MW hydro-powered 93% renewable data center in Mariner, New York.  Being located next to water, a renewables power source, bigger scale at 500MW, and a cooler location than Texas are all important factors.

We value the CORZ / CoreWeave deal at $2.5bn for CORZ or $12.4m per MW.  This is similar to ondacapital's original post's value of $2.1bn, but his post was before the CORZ Inv Day, after which we increased our estimates.  We assumed a 12x EBITDA multiple for the $232m per yr of EBITDA and netted out the $300m of CapEx.

WULF

WULF's has 500MWs available at their Mariner facility (200MW brownfield and 300MW greenfield) and 100MWs which is part of their JV with Talen at the same nuclear site that AWS bought.

Despite what we believe could be a more attractive asset in WULF's Mariner facility as discussed above, we haircut the CORZ / CoreWeave $ per MW value by 20% ($10m per MW) for the 160 brownfield MWs (we use 160MW vs the 200MW nameplate for our estimate of PUE).  These 160 MWs are already operational and similar to CORZ, this would be a BTC to HPC conversion.  For the brownfield Mariner MWs of 240MW (300MW nameplate), we haircut by 30% using $7.5m per MW (using 240MW).  We assume very little value for the 100 Talen MWs using $100m given their lease ends in 2033 and AWS is the landowner - so there is little terminal value here.  But we think WULF cld see more than $100m in a sale to AWS given the PV of WULF's attractive energy price alone is $36m and the facility generates good cash flow from BTC (anyway, this only represents 3% of our TEV).  We assume very little resale for their BTC miners, deduct $120m for extending their lease at the Mariner facility, and capitalize their SG&A at our EBITDA multiple of 12x.  This gets a $3.7bn TEV, after deducting ~$50m net debt we get a $3.3bn mrkt cap for $9.10 per share or 132% upside from today. 

Valuation: 

IREN

We value IREN on their Childress assets, the 1.4GW development site mentioned above and then low value for the BTC assets.  For the Childress site, we value the brownfield MWs at a 30% discount ($8.7m/MW) to the CoreWeave / CORZ deal, the greenfield assets at a 50% discount ($6.2m/MW) and the 1.4GW development site at a 75% discount ($3.1m/MW).  We actually think the 1.4GW asset could be very attractive given it's scale.  For example, NEE mentioned at their recent Investor Day that they were approached by a customer looking for 5GW!  Using the same methodology above for BTC miners, SG&A and just $1m/MW for the BTC MWs yields a $5.8bn TEV, with no net debt (they are net cash but we deduct for the cost of bringing the additional 250MW Childress assets online by the end of this yr), we get a $5.8b Mrkt Cap on 168m shares for $34.63 per share or 162% upside from today.

Valuation:  

Catalyst

We believe we could see more deal announcements like we saw with CoreWeave and CORZ at any point in the coming weeks / months with IREN and WULF very actively pursuing HPC contracts.  We believe an announcement would quickly point the market to the analysis above, if not higher.  We also expect the companies to more clearly illustrate the value potential from HPC at their upcoming earnings as well

Risks

Of course, BTC price is a risk but we'd point out over the past couple weeks the stocks have begun decoupling from the price movements of BTC - we believe a good indicator the market is beginning to recognize the story here.  As with any BTC miner, these are high beta stocks and an industry that has not had the best track record of governance.