Unlock the Editor’s Digest for free
Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
You may have noticed that banks aren’t quite what they once were. More and more of the financial system’s heavy lifting is now being done by capital markets. At least in the US.
Here’s a new NBER paper that estimates that the market share of US banks in all private lending has almost halved from 60 per cent in 1970 to 35 per cent last year; loans as a percentage of bank assets has fallen from 70 per cent to 55 per cent; and that the share of household wealth held in deposit accounts has fallen from 22 per cent to 13 per cent.
The flipside is an explosion of lending by bond funds and other investment vehicles that are often collectively called “shadow banks”.
The numbers calculated by Greg Buchak, Gregor Matvos, Tomasz Piskorski and Amit Seru are stark and interesting, but not enormously groundbreaking.
It’s widely known that financial technologies like securitisation make it easier for markets to do more lending, and that more onerous regulations make traditional lending much trickier for banks. Meanwhile, people are stashing more and more of their money into investment funds rather than bank accounts.
The result looks like this:
One interesting thing is that the bank share of overall lending — what the paper refers to in teeth-grinding academese as “informationally sensitive lending share” — hasn’t declined further since the 2008 financial crisis, despite even tougher regulations in its wake. And there’s a lot of money that stayed in deposits despite negligible interest rates.
But the most interesting aspect of the paper is the argument that the radical reshaping of the lending ecosystem over the past few decades means that the banking sector could handle MUCH tougher capital requirements without massively curtailing credit extension — the biggest argument against such a move.
We consider a scenario in which capital requirements are increased dramatically, and evaluate how the intermediation sector would react in 2023 versus how it would react given the 1963 parameters. We find that while raising capital requirements to 25% modestly decreases lending in both economies, the effect in the 1963 economy is much more pronounced: Total lending decreases by 120 basis points, and the loan share of lending decreases by 8%. In contrast, the effect in 2023 is much smaller: Total lending decreases by only 60 basis points, with informationally sensitive lending decreasing by 5.7%.
In both cases, bank balance sheets contract dramatically. However, we only observe a fairly modest decline in aggregate lending, which becomes even more modest at 2023 parameters. This is due to the fact that while the increase in bank capital requirements results in a significant decrease in bank balance sheet lending, there is simultaneously an increase in lending through debt securities that substitute, albeit imperfectly, for informationally sensitive bank balance sheet lending.
One should obviously be careful about placing too much trust in simple modelling like this. The real world is messy. And this is just a look at the US financial system, where capital markets are much more advanced than anywhere else.
But it intuitively makes sense that as lending has migrated out of banks and into markets it become less sensitive to shifts in bank regulation. This might explain why the overall share of bank lending has basically flatlined since 2009 despite the whole Dodd-Frank jamboree.
This is interesting in light of the ongoing debates about the Basel II “Endgame” rules that many banks are unhappy with, as the paper points out.
. . . The 2023 bank failures underscored yet again the fundamental issue of banking vulnerability rooted in the high financial leverage employed by banks. The high leverage of banks is largely byproduct of safety nets embedded in insured deposit funding and ability of banks to issue money like claims.
The ongoing regulatory discussion including Basel III endgame aims to address this vulnerability by considering increased capital requirements for the banks. The critics of such proposals express concerns that the increase in bank capital requirements will have large adverse effects on aggregate lending and the broader economy.
In this regard, our analysis suggests that banks nowadays are much less important for provision of credit than they used to be. Our structural model indicates that increasing bank capital requirements would have only modest adverse effects on aggregate lending and it will mainly lead to reallocation of credit from bank balance sheets towards debt securities.
Of course, this “reallocation of credit” can have downsides, even if the overall quantum of credit doesn’t change much in the process. Squeezing risks out of banks makes that part of the financial system safer, but the risks don’t disappear. They just take on another form somewhere else.
