Back to the Future? A Look Back at 1994

What We’re Watching

Key upcoming data points. A few key things we are watching in the upcoming week:

• Sometime this week: FDIC 1Q23 Quarterly Banking Profile, featuring updated industry aggregate data

• May 22: JP Morgan Chase’s investor day, which promises to offer useful read-throughs across the financial sector, in addition to incremental detail on the First Republic integration.

• May 25: Canadian G-SIB earnings; Royal Bank of Canada (6am release, 8:30am call) and Toronto Dominion (6:30am release, 1:30pm call).

• May 25: Fed H.4.1. release, ICI money market fund assets

• May 26: Fed H.8. release

This Week’s Additions to the FinReg Library

Now available for download. Our compilation of bank crisis data from wide-ranging government sources, including links to regulators’ reports, hearings past and future (with transcripts where available), legislation, and other documents is now available for download in a single file at this link. FactSet users can use the same link to access an updated version of the file each week. 

Among this week’s additions to the FinReg Library are: 

• FDIC annual report (containing time series of estimated uninsured deposits)

• Witness testimony from last week’s various Congressional hearings on banking matters

• Congressional Research Service reports on Crypto and Banking, and Short Selling

• House and Senate versions of Financial Technology Protection Act of 2023

2023 vs 1994

Some notable similarities. Last year we noted some similarities between the current interest rate cycle and that of 1994. In both cases, the Fed had held the Fed Funds rate low for an extended period following a financial crisis. The recession of the early 1990s was modest compared with 2008, but the effect on the banking system wasn’t; many more banks failed in 1989-91 than in 2008. As shown in Figure 1, the FOMC’s subsequent tightening wasn’t quite as steep in 1994 as in this cycle, but it was nonetheless among the steeper of the last four decades. The effect on bank securities books was exactly the same, however; yawning losses which, if recognized, would impair capital levels at many institutions.

Key differences. The lower/longer interest rate policy in the early 1990s wasn’t nearly as low, nor as long, as it was following 2008, with two key effects. Lower rates in the early 1990s were a huge benefit to banks; loan demand was flaccid but operating for a few years as levered bond funds enabled banks to recapitalize themselves at low risk. In contrast, rates were so low after 2008 that they weren’t quite as beneficial to bank margins since their costs of funds is subject to the zero bound. The “longer” component of policy was also much shorter, which did less to push markets out on the risk curve and offered a much narrower window for the misallocation of cheap capital.

Banks didn’t tighten credit sharply in 1994. One encouraging thing about current parallels with 1994 is that in 1995, despite the rate hikes, unrealized securities losses and derivatives blow-ups (remember those? Orange County does.), no recession followed. The soft landing is far from monocausal, but one very significant factor is that banks didn’t tighten their lending standards very much; credit remained available, albeit at higher cost. In contrast, as shown in Figures 2 and 3, banks have been and continue to broadly tighten credit standards this year.    

No big growthy banks in 1994. The list of the largest banks in the U.S. didn’t include a cohort of fast-growing niche-focused banks in 1994. Consolidation had a long way to go—the largest banks had total assets in the vicinity of $100 billion. Chemical Banking had recently digested Manufacturers Hanover, but Chase Manhattan, J.P. Morgan, Bank One, First Chicago, NBD, H.F. Ahmanson, and Washington Mutual—to name but a few of the predecessor institutions of the current JP Morgan Chase—were all independent banks (or thrifts), most of them taking up a slot of the list of the top 20 banks in the country. Each of these institutions had its vulnerabilities, but they generally weren’t out trying to grow at some multiple of the industry growth rate by concentrating in narrow verticals. Silicon Valley Bank was on investors’ radar in 1994 as an interesting niche bank, but much smaller (and as we now know, much better managed).

Uninsured deposits, then and now. Perhaps the most important difference between the two cycles (noticed only with the benefit of hindsight, and ruefully) is the share of uninsured deposits in the banking system. In 1994, just 18.7% of bank industry deposits were uninsured—with a percentage point of the all-time low. A secular increase was beginning; from the nadir of 17.9% in 1991, uninsured deposits climbed to levels not seen since the Kennedy administration; 46.6% in 2021 and 44.7% in 2022. Banks were simply far less vulnerable in 1994, because far fewer depositors had an incentive to participate in a bank run. It will be interesting to see whether we get an updated estimate from the FDIC in the Quarterly Banking Profile for 1Q23, which will likely be released later this week (though next week is a possibility).    

Weekly Federal Reserve Balances

May 17 Federal Reserve balances show slight uptick BTFP usage. Figures 5 and 6 highlight the key data points from last week’s release of Federal Reserve balances (Release H.4.1). Discount Window usage registered a slight decrease off a low base, while and Bank Term Funding Program usage rose $4 billion to another new high. The cost of BTFP funding remained 40bps below the Discount Window.

Money market assets eke out small increase. Money market assets rose $14 billion (0.25%) in the week ended May 17. Since March 8, that leaves money market assets up a total of $448 billion (9.2%). Overall, the trend seems to have reverted to its pre-crisis trajectory.

Total commercial bank deposits dip. Total deposits dipped slightly in the week ended May 10, falling $26 billion (15bps). Year-over-year, total deposits are down 5%, with just over half of that decline coming since March 8. As with money market fund assets, the trends in deposits have reverted to pre-crisis trajectories; the overall decline in deposits is slow and steady reflecting the interest rate environment, the decline is concentrated at larger banks (the starkest contrast with the events of March), and as shown in Figure 10, core deposits continue to slowly decline as a share of total deposits, reflecting the deteriorating deposit mix. This is the industrywide version of what we have seen recently at First Republic, PacWest and many other banks—outflows of demand deposits, typically noninterest-bearing, supplanted by hot money, typically costing 4-5%.  

Total loans dipped. Total loans were down $3 billion (3bps) in the week ended May 10, which left them up 45bps 2Q-to-date, which translates to about 4% annualized growth. C&I loans were down $4 billion (13bps) for the week, and down 2bps QTD. Commercial real estate loans were up $1 billion (2bps) for the week, which puts them on pace for roughly 7% annualized growth QTD. 

Interest Rates

Fed Funds futures turned slightly less dovish last week. As Figures 13 - 16 show, the yield curve remains inverted, and Fed Funds futures imply (as of this writing) that this tightening cycle, the steepest in over 40 years, is over. However, while futures still imply an overwhelming probability of a pause at the June 14 FOMC meeting, the outlook for autumn has turned a bit less dovish, with the first rate cut pushed out from the September 20 meeting to the November 1 meeting. Without reading too much into a relatively modest shift, it bears noting that the effects of “higher/longer” are likely to be asymmetrical; that is, the longer rates remain higher, the more likely we are to see discontinuities, both among banks and across the wider financial system, and the greater the magnitude of those discontinuities is apt to be. 

This blog post is for informational purposes only. The information contained in this blog post is not legal, tax, or investment advice. FactSet does not endorse or recommend any investments and assumes no liability for any consequence relating directly or indirectly to any action or inaction taken based on the information contained in this article.