I read the US rule change on the supplementary ratio the week it came out, and again this autumn when the Basel Committee published new data. It is the Basel-style ratio that ignores risk weights. Tier 1 capital is divided by total exposure, so a Treasury bill and a credit card loan count the same.
On 25 November 2025 the Federal Reserve, the FDIC and the OCC issued a joint final rule that cut the enhanced version, the eSLR , for the US global systemically important banks and their bank subsidiaries. The stated aim was to "reduce disincentives a banking organization may have to engage in lower-risk activities, such as intermediating in U.S. Treasury markets". It took effect on 1 April 2026.
Banks could opt in from 1 January. I want to explain why the headline numbers on this rule point in opposite directions. It took me two readings to see it.
Start with the bank subsidiaries. These are the insured entities that hold deposits, and the OCC's bulletin says their old standard was 6%: the 3% minimum plus another 3%, with anything below it ruling out "well capitalized" status.
The new standard is 3% plus the lesser of 1% or half of the parent's method 1 surcharge. In the agencies' words, that makes the requirement "no more than four percent" for these banks. It also stops being a prompt corrective action test.
It becomes a buffer instead. A bank that dips below it faces limits on dividends and discretionary bonuses, but it no longer falls out of the well capitalised category under the prompt corrective action rules.
The one-percent cap was new in the final text. Governor Michael Barr objected that it was a change "for which the public did not have an opportunity to provide feedback". I hadn't noticed that detail on my first pass.
The rule only reaches the very largest groups. The OCC says it applies to banks whose holding companies have at least $700bn of total consolidated assets or at least $10trn of assets under custody. No community bank is touched by it.
Here is the split that matters. The agencies said the rule "will reduce tier 1 capital requirements for affected bank holding companies by less than two percent". Barr put numbers on both levels.
"At the holding company, the decline is 1.4 percent ($13 billion)." At the bank subsidiaries, he wrote, the cut is 28%, a "$219 billion decline in bank capital". Total loss-absorbing capacity requirements fall by 5%, or $90bn, and long-term debt requirements fall by 16%, or $132bn.
The cut is small for the group and very large for the bank inside it. I had to read that twice. The agencies' answer is that the bank-level capital "generally would not be available for distribution to external shareholders due to capital restrictions at the holding company level". In plain terms, the money can move around the group but cannot leave it.
This is the twist I keep returning to. Barr also ran the exposure ratio on its own, ignoring the risk-weighted rules that sit beside it, and on that view the holding-company cut would be "about $209 billion for GSIBs", with $289bn at their banks. The actual group-level cut is $13bn.
Risk-based requirements explain the difference. For most of these firms they already bind harder. The agencies wanted the exposure ratio to work as a backstop "rather than as a constraint that is frequently binding", which I read as an admission that it had been binding often enough to change behaviour.
Barr's worry runs the other way. If risk weights fall, through policy or through what he describes as firms' incentives "to game their risk-based requirements", the $13bn could drift toward $209bn. I cannot tell how far, and I don't think anyone outside the banks can.
He also doubts that the Treasury market will benefit much at all. "Treasury market intermediation primarily happens at the broker-dealer, but the overwhelming bulk of the capital depletion under the final rule happens in the bank." I find that sentence hard to argue with.
I wanted a global yardstick for these figures. The Basel Committee's latest monitoring report, published in September 2026 with data to 31 December 2025, covers 149 banks, including 106 large internationally active ones and 29 global systemically important banks. Their average fully phased-in Tier 1 ratio on the exposure measure was 6.0%.
For the 29 G-SIBs it was 5.9%. For the 43 smaller Group 2 banks it was 6.8%. The large banks' average CET1 ratio was 13.9%.
I checked one more line in that report. Across the whole sample, banks showed a cumulative capital shortfall of only €1.4bn under the final Basel III framework, which by that measure means the large banks are hardly short of capital at all.
Those are group averages across many countries, so they don't map neatly onto a US bank subsidiary. Still, a 6.0% average sits 2 percentage points above the new 4% ceiling for American banks. That gap is the cushion Barr worries about, and it is money that can be spent.
The same day, the agencies proposed a second change that matters more to fintech deals. The community bank ratio, the simple Tier 1 to average assets test that small banks can opt into, would fall from 9% to 8%. The statute allows anything from 8% to 10%.
The Fed staff memo says the change "would enable 475 additional community banking organizations" to use the framework, on top of the 1,659 already in it. The grace period for a bank that slips out would double, from two quarters to four. A bank below 7% could not use it.
That denominator is where fintech lives. A fintech that buys a small bank for its charter, or a sponsor bank that takes on fast-growing programme deposits, adds assets with every deposit it parks. At 8% instead of 9%, the same capital supports about 12.5% more balance sheet.
I have no settled view on whether the November rule was wise. I have not found a public estimate of how much of the $219bn has actually moved between bank and parent since April, and the bank-level filings are where I will look for it first. The agencies say overall capital stays "broadly unchanged", and the holding-company figure backs them. Barr points to Silicon Valley Bank, writing that "there are lots of examples of firms failing without supporting their bank subsidiary".
What I do know is how I now read a bank's capital page in an M&A file. I look at the bank subsidiary apart from the group. I check which ratio binds, and I note how much of the cushion sits above the new 4%, while for a fintech buying a small bank I look at the community bank ratio against the proposed 8% line instead.