ABS Ratings = Milkshake NGU

Every passage, on the record.

1. 00:00Claim

   If you're an avid listener you'll remember the milkshake boys. Someone is losing here as Bitcoin continues to grow, because if you're a Bitcoin holder, your control over a certain percentage of the total accessible supply of energy available to the extended order at any one time is increasing.

2. 00:42Claim · condensed

   Someone has to lose in that situation, but not in the zero-sum physics sense, because knowledge is increasing, which means the physical transformations we are all able to perform are getting cheaper over time. In relative terms, if you bought bonds in 2010 versus buying Bitcoin, you are getting absolutely crushed and destroyed.

3. 00:42Analogy · condensed

   That is a milkshake, and getting drunk on it, the way you increase the size of the milkshake you can fit in your mouth, is to increase the size of the straw. Increase the width of the straw, that's how you're going to get it in there.

4. 01:30Claim

   This is a perfect opportunity to track why the knowledge we've been exposed to is actually going to produce real results in the wealth you're able to accumulate, wealth being the number of physical transformations you'll be able to initiate in the world. If you're a Bitcoiner, the number of physical transformations you can bring about at any one time has increased.

5. 01:30Event · condensed

   Six days ago, on the 20th of February, Ledn, the Bitcoin-backed lending company, pulled up a bunch of Bitcoin-backed loans and sold them in an asset-backed security. They pulled together 5,441 loans, initiated by 2,914 US retail borrowers.

6. 02:44Claim · condensed

   Almost 3,000 of those borrowers gave Bitcoin to Ledn, over-collateralized a loan, and got dollars in the last 12 months. Not a great move in the short term, but good for you long term, because you're short dollars and maintaining exposure to Bitcoin. Somebody out there is taking the other side of that trade, and they've been taken to the woodshed for the last 15 years.

7. 03:16Analogy

   If we go back to the jug analogy: someone has taken this lump of clay that was basically useless and, by inputting energy into it, constrained it into a shape that has a void inside that can now be filled. That causes new arrangements that were not possible before. There was a whole pool of milkshake that could not fit into the Bitcoin system, and someone has made an adapter that makes it significantly easier for your percentage of the TASOE to grow much larger, very quickly.

8. 03:16Idea · condensed

   This deal matters because it's a new configuration, a new arrangement available in the world that can now access other arrangements that weren't possible. There's a whole pool of milkshake, over 4,000 Bitcoin in this secure pool worth over 350 million dollars at the time of the valuation cutoff, that could not fit into the Bitcoin system before this.

9. 04:06Event · condensed

   The asset-backed security came out rated: of the 188 million dollars behind it, 160 million is rated triple-B minus by Standard & Poor's, which sits at the lowest rung of investment grade, still above junk bonds. The other 28 million is subordinate class B notes, non-investment grade, rated B minus.

10. 04:06Reference

    160 million of that is rated triple-B minus by Standard & Poor's.

11. 05:18Claim · condensed

    A pool of loans that are over-collateralized somehow comes back rated triple-B minus, which means lower rating requires a higher rate of return over the benchmark rate to compensate.

12. 05:49Claim · condensed

    The senior pricing came in at 335 basis points over SOFR, so if SOFR is 5%, that's 3.35% extra yield demanded by investors for the senior class note. The offering was two times oversubscribed, institutional demand exceeding the amount on offer, a sign of real appetite.

13. 05:49Claim · condensed

    The average weighted loan-to-value ratio on Ledn's underlying loans was 55%, a fairly large and conservative cushion relative to most lending, and the average interest rate borrowers were paying was 11.8%. If the ABS buyer is accepting roughly 8%, that's a four percentage point gap paying for admin and for Ledn to run the operation.

14. 07:19Claim · condensed

    That gap seems huge until you realize how safe these loans actually are, and the risk stacked on top of it, because there aren't just one or two layers here. There are seven different layers, and pricing them out is how we see that we just unlocked a huge pool of fiat dollars that couldn't get this exposure any other way, because their mandates won't let them buy and hold Bitcoin directly and stomach 80% volatility.

15. 07:19Reference · condensed

    This is in that 100 to 500 million dollar ABS transaction range, common in the consumer credit sector for things like auto loans and credit cards. That's how it's done.

16. 08:46Claim · condensed

    The 180 million dollar deal sits in the lower to mid-range of traditional ABS issue sizes, not some boutique micro deal. It's institutionally relevant, two times oversubscribed, and none of it rehypothecates the Bitcoin collateral. The pledged Bitcoin stays fenced in custody and can't be let out by any of the issuing trusts or partners, so that entire fear is removed from the equation.

17. 09:35Claim · condensed

    If you understand what Bitcoin is, you don't want to lower your exposure, you want to maximize it. If you're in this market for more than one or two cycles, nothing is going to beat this in terms of compound annual growth rate, and the smartest thing you can do is figure out how to stay as exposed as you can stomach.

18. 10:07Claim · condensed

    Rates are falling because as execution ability rises and the depth of the constraints verifiable by outside lenders increases, people realize these loans are incredibly difficult to default against: Bitcoin is over-collateralized, it's a 24/7 market, it can be liquidated at any time, and the whole process is automated.

19. 10:59Idea · condensed

    Is it impossible to default? No, that's the risk being priced in, that's why it's 3.3% over SOFR. But that spread comes down as the quality of the constraint increases. This is the cost of falsification over the cost of preservation, and this is where you can see how you make a fuckload of money understanding this stuff: if it's cheaper to verify than to falsify, that market or asset will increase in value and dig a deeper well in the informational manifold.

20. 10:59Quote

    This is where finally you can see how you can make a fuckload of money understanding this stuff.

21. 11:35Claim · condensed

    All you have to do to look out into the future and see where the flows are going to go is look for the deepest wells. But it's a mistake to think there's no risk on top of these loans just because they're over-collateralized, it's not all sunshine and lollipops.

22. 12:07Claim · condensed

    The volatility of the collateral is extreme. Bitcoin is over-collateralized, but it's extremely volatile compared to auto loans, credit cards, equipment loans and mortgages, because it's 24/7 and mark-to-market immediately. Even a conservative-sounding 55% average LTV can see that cushion disappear fast, since Bitcoin has done 50% drawdowns in a very short window in the last six months.

23. 12:44Claim · condensed

    Traditional collateral markets aren't set up to react to moves like Bitcoin's, and when there's fear, that fear gets mispriced as risk, which requires a larger return. There's real forced liquidation risk here: the whole ABS structure relies on automated liquidations, so if Bitcoin falls quickly the collateral has to be sold into a falling market, introducing slippage, liquidity stress and market impact that traditional triple-B auto loan ABS structures don't face on the same intraday basis.

24. 13:27Claim

    If the market goes down, we want to see liquidations. If we see liquidations and the system remains intact, it gets stronger. The anti-fragile systems are the ones we want, and it isn't unexpected that prices are high here because there's real operational and execution risk in a new asset class.

25. 13:27Reference

    The anti-fragile systems are the ones that we want.

26. 14:06Claim · condensed

    Traditional asset-backed securities have decades of performance history behind their pricing, whereas Bitcoin-backed consumer loans don't, so investors in this space demand compensation for that. If you're early to the game and this is all you can plug into, getting 8.5% versus 6% with significantly fewer defaults means you win.

27. 14:36Claim · condensed

    It's a tiny sandbox in terms of very few players, but huge in fiat dollar denomination. As people start winning, news spreads quickly, more people allocate fiat dollars to Bitcoin-backed ABS, rates come down, it becomes more attractive to Bitcoin holders, liquidity increases, and the speculative attack that was hard to run before becomes much easier.

28. 14:36Quote

    We want to grease the wheels of value and energy flowing from a system that is restrictive to one that is open and impossible to fuck with.

29. 14:36Claim · condensed

    Even with all these layers stacked on Bitcoin as a foundation, there are still opportunities for regulatory and institutional risk to be mitigated or priced in or out depending on the jurisdiction's rule of law. Bitcoin is trustless, but all the other layers built on top of it carry some amount of trust, which makes them easier to falsify.

30. 16:07Idea · condensed

    The cost of falsification versus cost of preservation ratio begins to diminish as you go up the stack, which gives us a framework to price these things based on that ratio. Even though Bitcoin has a huge cf-cp ratio, we're not getting rid of volatility risk, liquidity risk, model risk, execution risk, regulatory risk or secondary market risk on top of it.

31. 16:40Claim

    That triple-B rating is about expected loss probability, and the spread is about required return for all the perceived risks, not just risk to the Bitcoin. Those are not the same thing.

32. 16:40Claim · condensed

    If the system is constructed as well as it appears to have been over the roughly four years Ledn has been doing this, since around 2020, the numbers look good enough that these things end up getting plugged into new systems of capital that are increasingly difficult to get into.

33. 16:40Claim

    As you go further up the stack, more trust is required, and as more trust is required the cf-cp ratio skews in the wrong direction. There's only so high you can build this before it makes sense to get more cautious. Layer one is easy: Bitcoin, and who actually holds the keys.

34. 17:58Claim · condensed

    Layer one is collateral control and custody: not your keys, not your coins, so who has the Bitcoin, is each loan's collateral clearly segregated, and if something goes wrong, can the exact collateral be identified and moved immediately without confusion or delay. If a borrower defaults, can the system grab the exact Bitcoin tied to that loan without confusion.

35. 18:38Claim · condensed

    Building systems that answer these custody questions clearly takes energy and knowledge, and that's what you get rewarded for. If the answer is fuzzy, if you can't prove your work, that's risk, and prices have to rise to compensate the investors allocating there.

36. 19:09Claim · condensed

    Layer two is price marks: everything in collateralized lending depends on price. Which price feeds are used, how often they update, what happens if exchanges disagree, and is there someone who can override that, say, declare one exchange's price wrong after a hack, and determine the accurate reflection of the underlying collateral's value.

37. 19:41Claim · condensed

    If someone has the ability to override which price feed is used, that's decreasing the cost to falsify the system and increasing one individual's ability to mess with it, which makes it less useful. We're applying what we've learned about physics to this monetary network in order to increase our wealth.

38. 20:25Claim · condensed

    If the price mark is wrong or slow, the margin call gets fired late, and if losses grow quietly before anyone reacts, that's a problem. Good systems measure constantly, which is the depth side of things, and transparently, which is the breadth side, so more people from the outside can verify the system hasn't been tampered with.

39. 20:25Analogy · condensed

    It's like checking your blood pressure on a device that only updates once every six hours instead of constantly. You want to constantly be checking so you don't miss any spikes.

40. 21:18Claim · condensed

    Layer three is the trigger and margin rules for when things go south: at what loan-to-value does a margin call actually happen, how long does the borrower have to fix it, twenty-four hours, seventy-two hours, and at what point does actual liquidation start, with or without discretion.

41. 21:56Claim

    Automatic systems, this is what Bitcoin is, preserve discipline. Discretion-heavy systems protect appearances, they look anti-fragile but are in fact fragile, and that's the worst kind of system to be exposed to because it looks great right up until it's going horribly wrong. We want systems that respond immediately.

42. 21:56Quote

    Discretion heavy systems, they protect appearances. They look anti-fragile, but in fact are fragile, and that's the worst kind of system to be exposed to because it looks like everything is going great until it's going horribly wrong.

43. 22:48Claim · condensed

    Layer four is liquidation execution: when things have gone south, can you actually sell cleanly and quickly, are there predefined sell pathways, and if there's slippage, is it documented. Writing good rules is easy, executing them at speed in the middle of chaos is much harder.

44. 23:19Analogy

    If this liquidation execution layer is weak, everything above it is cosmetic. It's like having a great evacuation plan, but if all the doors you planned on evacuating through are locked, you're in trouble.

45. 24:07Claim · condensed

    Layer five is the legal structure: do investors have enforceable rights, is the special purpose vehicle holding the loans legally isolated from other assets, is it co-mingled, and who has authority if there's a default. Can those rights be exercised quickly, in a day rather than a year.

46. 24:51Claim · condensed

    Just because you have strong collateral doesn't mean it compensates for weak legal plumbing. You want jurisdictions that are clean, with legal pathways that are fast and unambiguous.

47. 24:51Analogy · condensed

    Credit is really a promise chain, the exact opposite of Bitcoin. You're making promises along the way, and that chain has to hold up in court if you go outside it and say, wait, you promised me this.

48. 25:46Claim · condensed

    Layer six is reporting and transparency: assuming all the layers below are true, how do you make the cost to falsify even greater than the cost to preserve. Can outsiders see what's happening inside, are there frequent reports from many people.

49. 26:24Idea · condensed

    This is the breadth and depth in K equals Ic squared. The constraint here is breadth: can lots of people see it. And depth: can they see it and track it over many state changes, with stable definitions, so every liquidation event, near-liquidation event and payout can be verified independently.

50. 27:00Claim

    If the reporting is thin or constantly changing, risk can hide and pool and grow until the fragile system collapses. We want to design systems so the cost of falsification is much higher than the cost of preservation. Good systems by definition make checking them easy, bad systems make you rely on trust.

51. 27:44Claim · condensed

    Once reporting improves and ratings move from triple-B minus toward triple-A, prices come down because it's deemed less risky, and risk can actually move between different investors, a secondary market for the bond.

52. 27:44Claim · condensed

    Layer seven is that secondary market and who owns the bonds. If one party owns a whole tranche of these Bitcoin-backed loans, that's not near as compelling in breadth and depth terms as a thousand different entities owning them.

53. 28:16Analogy · condensed

    As the buyer base proves itself over time it gains more depth, like pushing your finger into pizza dough: it starts going deeper, and then with two fingers it spreads out, and more of what was on the edge of the gradient starts flowing into the well, and that well gathers more mass.

54. 28:16Claim

    If only a few investors hold the paper, volatility can become self-reinforcing. If there are lots of independent buyers, the system can absorb stress significantly easier. This matters because we want to remove friction from the ability of individual investors and nation-states with Bitcoin to increase their odds of mounting a speculative attack.

55. 29:38Claim · condensed

    At each layer, you can either make lying expensive and verification easy, or you can make discretion easy and truth slow. This is the whole game, and it's the advantage you have if you understand it.

56. 29:38Quote

    At each layer you can either make lying expensive and verification easy, or you can make discretion easy and you can make truth slow. This is the whole game.

57. 30:26Claim · condensed

    This deal happened six days ago, not a time of exuberance in Bitcoin markets, but that doesn't make it bad for this new ABS asset class. If loan-to-value keeps climbing toward 90% and loans are forced to liquidate, that's proving the system works.

58. 31:05Claim · condensed

    Even Bitcoin's value going down and forced liquidations happening isn't inherently bad, liquidation is enforcement. If the structure says the collateral gets sold at a given LTV, selling it there isn't failure, that's discipline. Failure would be holding out past the trigger, letting trust erode, delaying action and letting losses accumulate.

59. 31:50Quote

    In credit, pain delivered early is way healthier than maintaining an artificial layer of calm.

60. 31:50Claim · condensed

    The thing to watch isn't whether liquidation happens, liquidation is inevitable. It's whether liquidation was executed exactly as specified. Rule-consistent liquidation increases constraint quality, and increasing constraint quality means increasing knowledge, which means we're able to perform more physical transformations in the world, which is exactly what you want if you want a bigger piece of the TASOE.

61. 32:45Claim · condensed

    Bitcoin as perfect collateral doesn't solve everything, it solves a very large problem as the base: it's transparent in supply, has public settlement, no committee valuation, and is globally liquid. But once you wrap it in credit, humans, contracts and legal pathways enter, along with custody risk, servicing risk, legal enforcement and reporting opacity.

62. 32:45Quote

    You can build a really shitty fragile wrapper around a pristine asset. What we want is a really strong wrapper.

63. 34:19Idea

    The difference is constraint quality. If falsification is cheap and verification is hard, securitization is very dangerous. If falsification is expensive and verification is easy, securitization becomes efficient risk distribution.

64. 34:19Quote

    If falsification is cheap and verification is hard, securitization is very, very dangerous. But if falsification is expensive and verification is easy, then securitization becomes efficient risk distribution.

65. 35:33Quote

    It is huge for S&P just to rate the thing. Now that just opens up a whole new world.