How do you run a bank when a bank run is digital?
How do you run a bank when a bank run is digital?
Fractional reserve banking in a digital age
After the global financial crisis, banks have focused on optimizing maturity transformations to return equity to shareholders. Now, it seems like they should be focused on managing liquidity transformation.
First to get some definitions out of the way:
All financial products are just agreements about money now in exchange for money later, a series of cashflows and some risk.
Maturity Transformation: borrowing funds short term to invest in longer-term assets
Liquidity Transformation: Using liabilities that can be unwound every day to buy assets which cannot be sold every day.
Credit Transfer: underwriting a borrower with your own equity at risk, then selling that risk to a third party
Credit Intermediation: An institution which exists primarily to borrow funds and then lend them to others
Fractional Reserve Banking: A financial system where Credit intermediaries are only expected to reserve a fraction of deposits
Banks & Non Bank financial intermediaries.
Banks are different from non banks in two major ways
They are given access to different forms of leverage and financial products as members of a National or state level central bank and charter system, in exchange for high regulation.
they are allowed to market something called “deposits” which are usually insured by the economic and political system working in concert.
A Modern Treasury
The principal purpose of a bank is credit intermediation. Economies require healthy credit markets, where the median and aggregate investment is paid back at a return greater than inflation. This isn’t something every saver is going to be good at while being superordinate in their day jobs - so banks are supposed to do that.
One potential evolution for the central bank(s) could be to leave credit transfer and maturity transformation at the banks and take over liquidity transformation. I don’t mean centralizing and speeding up todays payments, but also having a clear ledger of tomorrows settlements of credit. The additional mandate mandate would be to further foster financial stability by keeping the total inflation adjusted present value of all assets and liabilities equal to the MTM CET1 ratio of covered banks?1
I’m not a financial stability expert by any means, but it seems we are in a regime where
deposits are very fluid and existing deposit outflow models are probably wrong2 about how fast they can move(more on this later)
Lots of banks own US treasuries(and fixed rate debt generally) at a loss
Central banks and regulators are expected to protect all deposits, whether they say so or not.
One practical implementation could be some sort of clearing price for banks who own things that cannot be unwound today and fund them with things that can be unwound today. Well capitalized banks should be willing to take on contingent deposit liabilities and get compensated through some rate higher than the short term funding rate. This is tricky for many reasons, but mainly because there are not really widely accepted quantitative liquidity metrics used at banks today, which rely on things observable in the market.3
US Banking Liquidity
This paper says that US banking system has $2tn of losses in their Held To Maturity portfolios4. That seems bad, but reserves of depository Institutions at the Fed sit at 3tn. The figure also doesn't take into consideration cash in the vaults.
Reserves
There used to be a concept of excess reserves , but they “simplified” this in 2020.5
Old Excess Reserves
There are enough reserves floating around in aggregate such that depositors could withdraw funds completely form the “bad banks” to the '“good banks” . The problem is that it is disorderly and many many banks would die along the way:
If one bank has 1bn deposits, invested in 1bn 10y US Treasuries held at a loss, then the system is probably fine.
If one bank has 40bn deposits, invested in 40bn 10y US Treasuries held at a loss, we know that this is not fine
If 100 banks have 2tn deposits, invested in 2tn 10y US Treasuries held at a loss, we are about to find out if that is fine…
The digital bank run age means there isn’t enough time for the market to clear. A central liquidity model with visibility into this liquidity mismatch could establish funding facilities in line with growth of liquidity imbalances, rather than wait for the bank run.
A Modern Currency
Mark Carney was on to something. Four years ago, the former BoE Governor gave a boring speech at the Jackson Hole symposium where he suggested that the world should rethink the international monetary and financial system.
He suggested that instead of a new "hegemonic" reserve currency, global central bankers should enable a synthetic reserve currency which was digital first, dynamic in the sense that its value was driven by multiple currencies, weighted by output and trade( I assume).
In plain english, this is a basket of currencies, available as a single number, accepted as payment and used as a denomination for debt of emerging market economies instead of USD. This would reducing the influence that gyrations in the US economy have on far flung countries.6
The US Dollar has dominated the world market for trade since the first world war.7 Local savers in the UK and investors in sterling assets who did not switch to dollar denominated assets have underperformed materially in the last century. Even further, the staggering wealth of the US saver supported the proliferation of USD denominated debt by foreign governments and countries. These countries have no choice but to live with a mismatch in their income/assets vs these liabilities - they default all the time once local inflation deviates materially from US inflation.8
In a de-globalizing world, this synthetic currency seems to me like a competitive alternative which the 7bn people who do not own BTC may trust quickly. An aggressive investment in technology by the top 10 central banks to create a currency that could deal 24x7 as a means of payment would be ambitious and probably impactful.
Liquidity and Funding in the digital age
Years ago, in one of my roles for UBS, I was a trader responsible for making prices in medium and long term rates across developed markets. I worked closely with the short term interest rates (STIR) trading team to manage the fixed income businesses’ funding and liquidity alongside our Group Treasury.
We utilized a framework, which I’m sure is common across many large banks, to manage liabilities principally by staring at a Cashladder everyday. Without getting into too much detail, the Cashladder showed the present value of the aggregate payments, in all developed market currencies, for the entire firm on a daily basis and teams were responsible for ensuring that those numbers were above a threshold based on how far in the future they were.9 They did this by borrowing and lending currencies in the FX and cross currency basis swap market.
Obviously, no cash balance could be negative for the next day, but it would be ok to be “short” some currency in 90 days, meaning the outflows on that single day were greater than the inflows, on an expected basis. There was more art than science in setting thresholds and apart from synthesizing the entire banks cash and derivatives portfolios, Group Treasury had to come up with a model for deposit outflow based on many factors.
This system was live for 24 x 5 and there no quantitative liquidity risk factor to measure the operation. Any group in the bank who could influence these numbers were not constrained by a dynamic observable risk, they were constrained by a threshold set by a team centered in Switzerland. On the flip side, every single trading desk could cause a breach of this limit, which would be resolved in short term FX and cash markets by the STIR desk.
Group Treasurers are probably working weekends these days. Some have suggested “narrow banks” are a solution for those who want their deposits to be available, and while the regulators seem to dislike this concepts, it could counterintuitively be a home for the least sticky deposits.
Every central bank in the world probably needs to stare at each GSIBs cashladder and help them discern the right thresholds and price for short term cash by both providing liquidity facilities and enforcing real daily penalties for running liquidity mismatches.
BTC vs CBDC
History suggests that reserve currencies will shift, that economic systems and the political systems which work alongside them will evolve and that there will be winners and losers. Smart people say history at least rhymes, and generally war comes alongside a shift in reserve currencies.
I wonder if a financial system for 8 billion humans can exist with unsecured credit. Maybe bitcoin “solves” inflation, but what is the point of not being exposed to higher future prices if there is no structure for getting money from the future(credit). Credit is one of the core ways that a healthy productive ideas come to fruition. But I do wonder - Bitcoin has many owners globally but there is no widely accepted concept of credit built into the system, or on top of it.
No-one seems to be lending bitcoin to others to build productive things in exchange for bitcoin later in an uncollateralized way. If you are lucky enough to be rich and have bitcoin that is great, but what are the rest of us going to do and why do we care to be saved from tyranny of fiat inflation when the reason we are sucking on the teat of the fractional reserves system (and specifically regional banks) is credit!
If someone out there manages to create a simple way to enable investment in bitcoin through unsecured credit, this would be the first true market measure of real inflation denominated in BTC.
Outside of the military might, credit may be the major advantage that CBDC’s would have over BTC as both concepts are in a race for hegemony.
This is the best thing I can come up with but clearly the point is the banks are not doing it well by themselves.
Recent FDIC outflow models suggested that banks would be able to raise interest rates to attract deposits below the FDIC limit. It would be interesting to see if this was the case with SVB.
I could be wrong here, but if you look at a bank like JPM’s 10k, they discuss Liquidity Risk by reporting regulatory ratios and quantifying the size of contingent funding scenarios like “downgrade triggers” - which refers to the risk of required additional collateral posting in the event of a downgrade of a particular entity. There is no real quantitative measure comparable to what they provide for interest rate risk - a 100 basis point shift in higher interest rates would cause a +5bn dollar P&L event.
Held to Maturity portfolios are just things that the bank chooses - obviously they can sell them anytime
If any of my readers know how this works or why it was done please let me know! I have to say that simplifying something in March of 2020 by not measuring it anymore is probably bad.
I have never understood how or why investors ignore structural issues and difficulty dealing in local currency debt of emerging market economies, somehow get comfortable investing in USD bonds and delaying liquidity issues in economies which are not inherently USD income driven.
The great depression was a small hiccup…
https://www.imf.org/external/pubs/ft/wp/2016/wp1637.pdf
We also obviously looked at traditional derivatives risk factors, but I always found that breaking down risk into the fixed cashflows and then the variable cashflows driven by some observable served me well.