,
32 tweets,
5 min read
Read on Twitter
1/32 Two big overarching themes dominated the policy response to the global financial crisis. First, there was a move toward engineered asset inflation, an attempt to reflate the banks out of their solvency problem.
2/32 As discussed previously, this was highly regressive, and now there is a populist whirlwind to reap with respect to the long term implications of that policy.
3/32 Less talked about, but perhaps equally significant, was a movement to concentrate and control the banks. This was manifested in shotgun marriages (WB=> WFC, WAMU => JPM, ML => BAC, etc.) and nationalisations/rollups.
4/32 Philosophically, I understand the approach here. Competently operating and risk managing a banking footprint is an intensive job. I suspect the idea was that merging the wounded into the healthy would make regulating the system more tractable.
5/32 From a banking regulator's perspective, if they are big, you know where to look if you suspect there is a problem in the system, and its a simpler chain of command when you need to pull control levers to engineer outcomes.
6/32 And control them they did - alongside greater corporate concentration came much more aggressive capital and liquidity rules, greater conduct safeguards etc. This makes it harder for banks to undertake risky private sector activity including warehousing risk, etc.
7/32 And yet: "too big to fail" has now become "too *bigger* to fail." And that was a proactive choice, made by regulators, because they thought it made the system safer and easier to manage.
8/32 An aside: the choice to concentrate through "orderly" mergers and interventions also minimised volatility, thus preserving value for incumbent owners of financial assets, just like QE.
9/32 However, the supposed risk mitigant of the greater control implicit in the concentrated architecture comes with a cost. In the event of a problem with a single institution, the severity of that problem to the overall system becomes bigger.
10/32 In banker speak: the regulators accepted the tradeoff of assuming larger loss-given-default (LGD), in exchange for the perceived benefit of a lower probability of default (PD) arising from faith that the new more aggressive rules would successfully make banking "boring"
11/32 But is the PD really lower? Implicit in the existing liquidity and capital rules is an important assumption: that the sovereign debt concentrated inside the banks is both solvent and liquid. History shows this is not a great assumption.
12/32 The market knows this and does differentiate sovereign bond prices on the basis of perceived credit quality. But banks are allowed to hold domestic sovereign debt at a risk weight of 0% against their reg cap.
13/32 Why is this the case? Turns out differentiating sov debt capital rules by jurisdiction is too problematic politically to do otherwise.
14/32 Of course, if you are allowed to hold alot of risk against 0 capital, and the regulator and politicians are encouraging you into buying sov paper, why not lever up? Especially if new rules make your ordinary business with the private sector more difficult.
15/32 Meanwhile fiscal imbalances, unfunded liabilites etc. are only growing. So you can see the vector along which risk may be building now again in the system.
16/32 Meanwhile, in the fiat system, if you want to hold funds, you have to keep your money on a bank's balance sheet, where it is re-hypothecated, co-mingled and then re-lent, in many cases to the local sovereign in the form of government bond purchases.
17/32 In the event of a problem with your bank, the risk of haircutting above a certain balance is real. And even if deposit insurance is paying in a systemic event, its may very well come with a healthy dose of inflation, since any event may very well be sovereign-driven.
18/32 But why should your access to the payments system and savings infrastructure come with the implicit risk to the bank/sovereign's health? With crypto, there's no reason that the lending function and payments function need to be co-mingled at all.
19/32 Instead, a key-holder can hold/manage their own funds, *without counterparty risk.* This is a key differentiator that provides the ability to envisage a better, more resilient architecture for the system as a whole.
20/32 The process of matching savers with borrowers could be done on a pass-through basis using tokenised products. Instead of making a demand deposit to a bank, savers could self-allocate their savings across a menu of products of varying time-durations.
21/32 I am sure there will eventually be great UI/UX making this all happen quite naturally, with the benefit of full asset transparency downstream. All self custodied.
22/32 The vast majority of loan credit underwriting is mechanical and analytically straight forward anyway. Why not have software do it, and pass the risk directly through to savers on an asset-specific basis?
23/32 The potential for an efficiency uptick from this sort of thing is quite large. There's alot of disintermediation to do. As a saver I would get greater direct security over hard assets, plus an uplift in the rate I earn on my savings. Borrowers get a less frictional rate.
24/32 All at the expense of expense of the net interest margins of incumbent, intermediating banks. Win-Win-Win.
25/32 Put differently, whenever I hear someone mention "tokenising" cash-flowing assets, I just replace that word in my mind with the word "securitising."
26/32 Digital secrities tokesn (STOs) like these will represent another step in the long history of market pricing mechanisms growing and coordinating more and more human activity.
27/32 Asset-backed securitisation got a pretty bad rap post-crisis, and undoubtedly some pre-crisis products were problematic. But the vast majority of the simpler structures proved quite resilient, especially the ones with good data reporting.
28/32 Bottom line, if I'm going fund a box full of loans with my savings, I'd rather its a transparent pool of loans where I have direct ownership, rather than a bank doing who-knows-what with my money.
29/32 Blockchain tech will make the audit trails on these things much better, heavily reducing the risks for origination abuses that were prevalent in the last crisis.
30/32 The tech will also remove some of the need for reliance on third party rating agencies, replacing it with transparent, realtime, publicly verifiable performance data.
31/32 I suspect that in the future the tagline for this better, atomised, financial system architecture will be "assets, not institutions." Non custodial wallets and pass through tokenisations of secured lending will eventually make this easy and ubiquitous.
32/32 Assets, not institutions. Endpoint self custody, not intermediating risky balance sheets. Market pricing, not command-and-control. Atomised and resilient, not concentrated and fragile. I believe this is the path to a safer, more just, and better financial system.
