When US safety nets undercut the financial system

Receive free Financial & markets regulation updates

We’ll send you a myFT Daily Digest email rounding up the latest Financial & markets regulation news every morning.

When Silicon Valley Bank collapsed this spring, it revealed ugly truths about the private sector financial world — including that investors and regulators had become “blindsided by risks which ex post seemed too obvious to be missed”, as a hard-hitting new book from NYU Stern business school notes.

Most notably, many financiers ignored rising interest rates — even though these were clearly visible — while taking crazy bets that later blew up.

Why? Excessively loose monetary policy was one culprit. Bad accounting rules were another. But some striking new research highlights two further issues that have hitherto been largely ignored: the role of government agencies such as the Federal Home Loan Bank system and the role of collateral in the financial system. Both deserve far more debate if we want to avoid another SVB-type shock.

The FHLB is arguably one of the oddest — and least-known — features of American finance. The entity was created in 1932 as a state-backed collection of 11 wholesale co-operative banks that jointly raise ultra-cheap debt to extend loans. Initially launched to help small savings banks and support mortgage markets, its mandate was expanded in the 1980s to cover mainstream commercial banks as well. And during the 2008 financial crisis, the FHLB furtively rescued many institutions by supplying loans when commercial markets closed due to the implosion of the subprime mortgage sector.

Indeed its unsung role was so significant back then that I often joked to European regulators that they had made a terrible mistake by letting their financiers import the US innovation of “free market” securitisation without mimicking the state institutions that quietly backstopped it. If the UK had had its own FHLB in 2007, say, Northern Rock might not have collapsed.

After the 2008 crisis, the FHLB’s lending shrank. But two years ago it quietly became hyperactive once again: between September 2021 and March 2023 its loan book exploded from $344bn to more than $1tn.

That surge was even more extreme than the one in 2008. But it went largely unnoticed, because the details of FHLB loans are not published in a timely manner. And in retrospect there are two striking points about this: the surge started long before the problems at SVB became widely known and data now shows that the three banks that imploded from March 2023 — SVB, Signature and First Republic — were heavy FHLB borrowers.

That was presumably because those banks’ funding costs were rising and their deposits trickling away. “Thus the FHLB borrowings . . . were critical in keeping the banks afloat,” the NYU book notes. “Even though they were undercapitalised . . . the FHLB advances allowed these banks to delay selling assets and/or raising equity” — and therefore keep “gambling for resurrection on the back of mispriced government-sponsored financing”.

The FHLB is certainly not the only government band aid for US finance; the Federal Reserve’s discount window, for example, also supports troubled banks. But the key point is that the Fed’s aid is only employed at times of visible crisis, while FHLB support is secretive and long term. It can thus potentially undercut market discipline by concealing stress in the banks — as the SVB saga shows.

The NYU team concludes from this that it is high time to overhaul the FHLB. I strongly agree. Although this strange beast has sometimes played a useful role in the past by backing small community banks, it should now either return to its original (limited) mission and dramatically tighten lending standards, or be abolished.

However, if regulators want to inject more market discipline into the system — as I think they should — they need to also look at a second issue: collateral.

This point is raised in a separate piece of new analysis from economists at the Bank for International Settlements. It notes another unsung detail of modern finance: that the use of collateral to back deals has exploded.

The BIS team is primarily focused on this trend in capital markets. But it applies to banks too. The FHLB, for example, has proudly emphasised that it has “over-collateralised” policies in recent years (ie by securing its loans against a plethora of assets).

In theory, this should make finance safer. But, as the BIS economists point out, collateral can also “reduc[e] incentives to screen and monitor borrowers . . . and increase aggregate leverage”. More specifically, there is less incentive to monitor counterparty risk — which was likely to be one reason why officials at the San Francisco FHLB kept lending to SVB.

Moreover, if the collateral is a security such as treasuries, it “heightens liquidity risk” and “an adverse shock to the value of government paper used as collateral could trigger destabilising dynamics”. This also played out at SVB — partly because regulators have encouraged banks to hold treasuries in recent years.

While a safety net can help individual players in the short term, it can also create more — not less — aggregate risk in the long term, by undercutting market discipline. This urgently needs to be discussed — not least because the NYU book ominously notes that a “current spike in FHLB advances to many other banks could well be creating similar distortions”. The saga has further to run.

gillian.tett@ft.com