I was beginning to get really bored of reading about Microstrategy $MSTR but this is the most comprehensive technical write up of the Saylor ‘infinity machine’ that I have seen so far. Bonkers stuff. Excellent note Alex.
Thank you for sharing your insights into the situation. I’m confused with one bit. 3/4 in you mention that BTC per share has fallen. I’m not sure how you arrive at that conclusion. Did I misunderstand?
I wished that someone would tackle the most potent bullish argument - is MSTR unique in its ability to generate a BTC yield? I think the answer could be yes. Smaller copy cats will have a tough time issuing 1Bn issue size and capture institutional liquidity. Larger names with the cash such as meta will never be able to replicate the realised vol of a pure BTC treasury company. If it is unique, does it deserve some premium (to NAV) for its ability to tap into the cheapest form of financial engineering?
I don't consider borrowing money to buy bitcoin a 'bitcoin yield.' Just buy the calls yourself!
You are comparing it to meta, whereas you should be comparing it to IBIT or other pure play ETFs that do that MSTR is doing but without the extra burden of running an actual company.
Also as mentioned in the blog, yes they are doing a good job of financial engineering, but you would expect that to eventually get competed away. Which means you need bitcoin to go up AND for them to buy more to make up for the fall in the 100% premium over assets.
Is there no value in being able to issue shares at 2-3X the value of the net BTC on the balance sheet? The sky-high realised vol permits MSTR to issue 150% conversion convertible bonds between 0-1% coupon. Assuming the mNAV is 2.5X, via convertibles, MSTR is effectively issuing at 3.75X NAV. Under these conditions, the post-dilution cost of acquiring BTC is 26.6% (1/3.75). This year alone, the BTC per share has increased shy of 65% YTD. This is unique and hard to replicate because I can't issue overvalued stock to buy the underlying stock, and I can't issue convertible bonds with 150% conversion and 0-1% coupons.
Apart from BTC's price trend, the highest risk is that the premium gets competed away. This is why I focused on two competitive threats in my initial message. First are smaller companies with a so-so operating business desperate enough to venture into the Bitcoin treasury model. They'd fall short on the convertible bond issue size (USD +1bn) and have trouble capturing real money manager AUM. The second is larger companies with enough cash (e.g Meta or Apple). They still won't be able to match MSTR's stock price realised volatility and, therefore, won't be able to issue convertibles with close to 0% coupon and far OTM conversion prices.
Therefore, apart from the apparent direction of BTC price, I think this strategy's limit is MSTR's convertible bond TAM. Based on my rookie research, it seems the yearly issue of the USD convertible bond market will reach 100bn in a good year. Unless MSTR can convince traditional non-convert FI investors to invest in their converts, I don't see how MSTR can accelerate its convertible bond issuance much from current levels.
I think this is an interesting article but doesn’t really draw any meaningful conclusions.
Saylor’s vol is higher than BTCs, so his converts are cheaper. That’s true. But so what? He could sell converts that are at a lower conversion premium and the conclusion is the same.
The punchline is it trades at a significant premium to NAV which doesn't make sense. Even with no debt and current holdings it's overvalued. I guess there is some world they can issue enough converts that convert at higher prices and erode the premium that way. Would be curious to know what that path would look like.
I find it very entertaining to watch this one. So unique. I see some ridiculous hype from Mr. Saylor that dramatically reduces his credibility, but I also admire the success so far in this unique financial engineering that he and his team has used to reboot that moribund company into this current thing. The Convert buyers are still lining up for these things. Will be fascinating to see how long this can last.
Wanted to check your calculation of the correlation between the volatility of the premium and the volatility of BTC - how do you get the underlying volatility of the premium? Everything in that equation is known besides the vol of the premium.
I guess these are all realised vol numbers and hence they are observable?
And hence it is a realised correlation number.
(Even though it says "implied volatility of the premium")
My two cents: I understand the options math, but this one makes my head spin. Most good investors only get the first derivative of what is going on, but miss the eco-systemic limits. What I mean is you have $MSTU and $MSTX running out of swap counterparties to create their levered $MSTR ETFs, so they going into the options market to get get delta, raising call option premiums in the process.
Now, Saylor can take advantage of this to issue more converts, most of which will get delta-hedged. But are there enough buyers? What runs out first? Convert buyers, call sellers, margin loans on $MSTR common? We've already run out of long swap sellers (possibly limited by cost to borrow $MSTR for shorting). But really, I have no idea... I know the system has to have a limit, but where it is, I can't tell.
Ah great point! I hadn't considered the impact of levered ETFs
If retail is buying them and then they hold them outright, that could force the market short gamma and hence drive the realize.
In that case, that dynamics could work on the downside the vol stops when the flow stops. Not sure what kills the flow, outside of lower prices. Which is why we have so much upside vol i guess
Good analysis but the conclusion is flaccid. "Each new MSTR convert issuance further complicates the math: the effective Bitcoin-per-share ratio has fallen to 0.002 BTC, diluted by repeated capital raises."
MicroStrategy held approximately 0.001121 BTC per share as of year-end 2023.
As of 12/6/24 MSTR has BTC per Basic Share: 0.00171671 BTC
The dilution has come at the expense of people buying overpriced MSTR equity but has benefitted existing shareholders. If the company continue to close the arbitrage itself, it is increasing its BTC per share, which is ultimately what shareholders want.
Perfectly said. I've read a lot on this topic, this perfectly sums it up. Buy shares at your own peril but the company will keep acting to close the NAV on itself.
Ultimately, MSTR will continue to do the goal of the company - owning more BTC - at the expense of equity (dilution) and bond (opportunity cost) holders. As a BTC beta holder, there’s no cause for concern.
It’s been a while since my days on the convert desk at SIG. Any idea what coupon MSTR would pay for a straight bond? Ie they are funding at 0% in exchange for the upside equity stake. What is that worth in fixed rate interest OR what’s the implied vol?
Great summary. Seeing those zero coupon converts being “sold” on CNBC the other day by Michael Saylor reminded me of brokers in 1998 saying Internet stocks (Pets.com) could only go up and then in 2007 mortgage backed securities being “risk less”.
I love the positive business sentiment that has been brought about by the political sea change. I just hope a reset doesn’t blow up the party before it gets started.
I was beginning to get really bored of reading about Microstrategy $MSTR but this is the most comprehensive technical write up of the Saylor ‘infinity machine’ that I have seen so far. Bonkers stuff. Excellent note Alex.
Thank you for sharing your insights into the situation. I’m confused with one bit. 3/4 in you mention that BTC per share has fallen. I’m not sure how you arrive at that conclusion. Did I misunderstand?
I wished that someone would tackle the most potent bullish argument - is MSTR unique in its ability to generate a BTC yield? I think the answer could be yes. Smaller copy cats will have a tough time issuing 1Bn issue size and capture institutional liquidity. Larger names with the cash such as meta will never be able to replicate the realised vol of a pure BTC treasury company. If it is unique, does it deserve some premium (to NAV) for its ability to tap into the cheapest form of financial engineering?
I don't consider borrowing money to buy bitcoin a 'bitcoin yield.' Just buy the calls yourself!
You are comparing it to meta, whereas you should be comparing it to IBIT or other pure play ETFs that do that MSTR is doing but without the extra burden of running an actual company.
Also as mentioned in the blog, yes they are doing a good job of financial engineering, but you would expect that to eventually get competed away. Which means you need bitcoin to go up AND for them to buy more to make up for the fall in the 100% premium over assets.
Thank you for replying.
Is there no value in being able to issue shares at 2-3X the value of the net BTC on the balance sheet? The sky-high realised vol permits MSTR to issue 150% conversion convertible bonds between 0-1% coupon. Assuming the mNAV is 2.5X, via convertibles, MSTR is effectively issuing at 3.75X NAV. Under these conditions, the post-dilution cost of acquiring BTC is 26.6% (1/3.75). This year alone, the BTC per share has increased shy of 65% YTD. This is unique and hard to replicate because I can't issue overvalued stock to buy the underlying stock, and I can't issue convertible bonds with 150% conversion and 0-1% coupons.
Apart from BTC's price trend, the highest risk is that the premium gets competed away. This is why I focused on two competitive threats in my initial message. First are smaller companies with a so-so operating business desperate enough to venture into the Bitcoin treasury model. They'd fall short on the convertible bond issue size (USD +1bn) and have trouble capturing real money manager AUM. The second is larger companies with enough cash (e.g Meta or Apple). They still won't be able to match MSTR's stock price realised volatility and, therefore, won't be able to issue convertibles with close to 0% coupon and far OTM conversion prices.
Therefore, apart from the apparent direction of BTC price, I think this strategy's limit is MSTR's convertible bond TAM. Based on my rookie research, it seems the yearly issue of the USD convertible bond market will reach 100bn in a good year. Unless MSTR can convince traditional non-convert FI investors to invest in their converts, I don't see how MSTR can accelerate its convertible bond issuance much from current levels.
One comment - there are micro BTC futures, not sure when they were introduced but much more accessible to retail
yup, great point, I missed those!
I think this is an interesting article but doesn’t really draw any meaningful conclusions.
Saylor’s vol is higher than BTCs, so his converts are cheaper. That’s true. But so what? He could sell converts that are at a lower conversion premium and the conclusion is the same.
The punchline is it trades at a significant premium to NAV which doesn't make sense. Even with no debt and current holdings it's overvalued. I guess there is some world they can issue enough converts that convert at higher prices and erode the premium that way. Would be curious to know what that path would look like.
I find it very entertaining to watch this one. So unique. I see some ridiculous hype from Mr. Saylor that dramatically reduces his credibility, but I also admire the success so far in this unique financial engineering that he and his team has used to reboot that moribund company into this current thing. The Convert buyers are still lining up for these things. Will be fascinating to see how long this can last.
Incredible content. Thank you.
Hi there, great post. Thank you.
Wanted to check your calculation of the correlation between the volatility of the premium and the volatility of BTC - how do you get the underlying volatility of the premium? Everything in that equation is known besides the vol of the premium.
I guess these are all realised vol numbers and hence they are observable?
And hence it is a realised correlation number.
(Even though it says "implied volatility of the premium")
Thank you.
My two cents: I understand the options math, but this one makes my head spin. Most good investors only get the first derivative of what is going on, but miss the eco-systemic limits. What I mean is you have $MSTU and $MSTX running out of swap counterparties to create their levered $MSTR ETFs, so they going into the options market to get get delta, raising call option premiums in the process.
Now, Saylor can take advantage of this to issue more converts, most of which will get delta-hedged. But are there enough buyers? What runs out first? Convert buyers, call sellers, margin loans on $MSTR common? We've already run out of long swap sellers (possibly limited by cost to borrow $MSTR for shorting). But really, I have no idea... I know the system has to have a limit, but where it is, I can't tell.
Thanks for writing your erudite article.
Ah great point! I hadn't considered the impact of levered ETFs
If retail is buying them and then they hold them outright, that could force the market short gamma and hence drive the realize.
In that case, that dynamics could work on the downside the vol stops when the flow stops. Not sure what kills the flow, outside of lower prices. Which is why we have so much upside vol i guess
Good analysis but the conclusion is flaccid. "Each new MSTR convert issuance further complicates the math: the effective Bitcoin-per-share ratio has fallen to 0.002 BTC, diluted by repeated capital raises."
MicroStrategy held approximately 0.001121 BTC per share as of year-end 2023.
As of 12/6/24 MSTR has BTC per Basic Share: 0.00171671 BTC
The dilution has come at the expense of people buying overpriced MSTR equity but has benefitted existing shareholders. If the company continue to close the arbitrage itself, it is increasing its BTC per share, which is ultimately what shareholders want.
Perfectly said. I've read a lot on this topic, this perfectly sums it up. Buy shares at your own peril but the company will keep acting to close the NAV on itself.
Ultimately, MSTR will continue to do the goal of the company - owning more BTC - at the expense of equity (dilution) and bond (opportunity cost) holders. As a BTC beta holder, there’s no cause for concern.
It’s been a while since my days on the convert desk at SIG. Any idea what coupon MSTR would pay for a straight bond? Ie they are funding at 0% in exchange for the upside equity stake. What is that worth in fixed rate interest OR what’s the implied vol?
the pyramidal structure at work here is why Senator Lummis is writing legislation to get the American taxpayer to buy Michael Saylor's enormous bags: https://www.washingtonpost.com/business/2024/11/27/trump-strategic-bitcoin-reserve-plan/
Great summary. Seeing those zero coupon converts being “sold” on CNBC the other day by Michael Saylor reminded me of brokers in 1998 saying Internet stocks (Pets.com) could only go up and then in 2007 mortgage backed securities being “risk less”.
I love the positive business sentiment that has been brought about by the political sea change. I just hope a reset doesn’t blow up the party before it gets started.
Excellent piece.