Bitdeer - Post Q2 26 view

Conclusion is I would hold the position. If your view is AI capX stays strong in 2027, the aggressive move is to add to the position.

Who says life is fair ? Riot announced a deal. The stock surged. Bitdeer announced a deal, it promptly tanked.

First, let's figure out why did the stock tank 20% post result ? In order of importance, folks are fearful Bitdeer has to raise more money to fund its capX (it confirmed a bigger ATM shelf). Second, its site in Clarington is delayed as the judge ruled out dismissing the complaint summarily. And looks like it is going to position Clarington for BTC for now. Third, it now ranks Malaysia as the next project to watch. Fourth, Knoxville is delayed to 2H 27. Fifth, no news on Rockdale, Texas.

Basically, it perfectly fits the Wall Street image of Rodney Dangerfield i.e. it just cannot execute.

Now if we look under the cover, a few things are happening. See quotes from the call in Pic.

1. Now the elephant in the room... Does it have to raise more money ? Answer is yes..... But it is what it is.... In Q2, it already raised $3-400M from ATM if I am not mistaken.... So it has over $400M in cash at end of Jun. The remaining Tydal capX is around $500M. Importantly, in the call, CFO talked about majority of the Tydal capX will be borrowed via project finance. Guess it is not the end of the world if BTDR has to tip in $100-150M. With some luck, through 2H, dilution won't be too bad.

2. The big idea is BTDR is seemingly betting on itself to become a neocloud (or GPU rental, rather than co-lo). In Tydal, it deliberately held back just under 50MW from the co-lo deal with Volta. In addition, it signed a lease for 20MW in Malaysia for AI Cloud as well. So all up, it has set aside as much as 170MW for Ai cloud.... 170MW may not sound like a lot but capX alone can cost as much as $7bn or so.....

The intriguing question for me is why did BTDR choose to go after the neocloud opportunity ? Is it really a better business than co-lo ? This I don't have the answer..... In the call, CEO talked about for these GPU/AI cloud deals, it is looking at 5 year leases.... Presumably, capX + sum will be earned back by the end of the lease term.

Also of note, Knoxville is delayed because BTDR is modifying the design to one big data centre after feedback from potential customers.

Guess we can at least infer that management is sensible enough and it is lining up customers for these neocloud sites. And if these deals eventuate, they should offer OK economics to BTDR.

3. BTDR is now the largest publicly listed BTC miner. Its total capacity is 89EH..... Well ahead of Mara (70EH), Cleanspark (50EH) and Riot (44EH). It is producing close to 12K BTC a year. One day, when BTC recovers to over $120K, BTDR will be very profitable.

Here are the things to watch..... 1) Announcement on its funding package for Tydal, 2) when it buys land in Rockdale...

My gut feeling is stock is close to bottom... Announcement of the funding package for Tydal should signal the bottom i.e. it won't have to raise for some time until it has secured more customers for its neocloud.

How much upside ? Tydal should generate close to $300M in net profit a year... When BTC turns, won't surprise me if BTDR makes $500-1bn somewhere. Then throw in whatever it stands to make from its neocloud sites + co-lo sites.....

I can see one day in the future BTDR will be on 1-2X PE (if either the AI cycle keeps going or BTC comes back with a vengeance).