With abundant reserves, do banks adjust reserve balances to accommodate payment flows?

Following the the global financial crisis (GFC), the Federal Reserve moved from a regime of from rare reserves to one of abundant reserves. In this article, we explore how banks’ day-to-day management of reserve balances with respect to payment flows has changed with this regime change. We find that the behavior of banks has not changed on average; under both regimes, banks increased their opening balances when they expected higher outgoing payments and similarly decreased these balances with higher incoming payments. However, there are substantial differences between banks. When the abundant reserves regime was introduced, the smaller national banks no longer adjusted their balances in line with changes in outgoing payments.

How do banks manage their balances given payment flows?

A central aspect of banking is fulfilling its obligations to other banks on behalf of customers as well as a bank’s own business. This is often achieved through the Fedwire® Funds department, which allows banks to make immediate, high-value payments to each other using central bank money or reserves. (“Fedwire” is a registered service mark of the Federal Reserve Banks.)

Prior to the GFC, the Federal Reserve maintained a scarce reserve regime, providing banks with less than $50 billion in total reserves. Most banks were faced with the problem that the total payments sent and received during the day were significantly higher than the banks’ opening balances, making overdrafts possible. To minimize overdraft fees, banks both carefully timed when they made intraday payments and strategically chose their opening balance. This latter decision involves a trade-off: larger opening balances reduce the risks of intraday overdrafts but incur opportunity costs because reserves can generate higher returns elsewhere (in the fed funds market, for example).

After the GFC, the Federal Reserve moved to a regime of abundant reserves and significantly increased the level of reserves in the system. He also started pay interest on account balances. With these changes, the question remains open whether the strategic management of balances in response to payment flows remains of paramount importance for banks.

How to compare balances and payment flows?

To answer this question, we use confidential account-level information about bank opening and closing balances, as well as the total daily value of credits and debits to that account by type of payment service. We focus on Fedwire credits and debits, grouping all other transfers into an “other net change” variable where a bank’s closing balance equals its opening balance + Fedwire credit – Fedwire debit + other change net.

We start by considering the ratio of opening balances to the value of Fedwire debits (or bank-sent payments). We define the pre-GFC scarce reserve regime as February-May 2007 and the post-GCF abundant reserve regime as the same four months in 2015, shortly after the Federal Reserve ended. third period of quantitative easing in October 2014. We find that the distribution of the ratio of opening balances to Fedwire debits is radically different over these two time periods. In 2007, these ratios were much more concentrated near zero and the median was 0.27, meaning the median bank held an opening balance of 27% of the total daily value of outgoing Fedwire payments. It is therefore clear that banks relied on incoming payments to fund outgoing payments, a point examined more closely in this 2014 Economy of Liberty Street Publish and this working paper. In 2015, by comparison, Fedwire opening balance to debit ratios shifted to the right, with the median value equaling 11.06. This significant change suggests that banks may no longer have to worry about funding their outgoing debit payments with incoming credit payments.

Banks differ significantly in how they choose opening balances. To show this, we first divide banks into foreign and domestic, then we further divide domestic banks into small, medium, and large based on their daily gross Fedwire transfers (credit plus debit). We classify the top ten banks as large, the next forty as medium and the remaining national banks as small. We then look at the distribution of the ratio of opening balances to Fedwire debits over the two time periods between the banks (see chart below). It is particularly striking that the small national banks have higher ratios in both regimes and that, in the abundant regime, their balances are much larger than the payment outflows.

Fedwire Funds Opening Balance to Debit Ratio by Bank Size

Percentiles
Year Bank type 25th 50th 75th
2007 little servant 0.055 0.289 1.394
Domestic medium 0.004 0.008 0.017
Great servant 0.001 0.002 0.005
Foreign 0.001 0.008 0.051
Total 0.048 0.267 1,330
2015 little servant 2.094 12.289 61.409
Domestic medium 0.235 1.000 1.442
Great servant 0.161 0.292 0.903
Foreign 1.000 1.849 6.783
Total 1.797 11.061 56.775
Source: Federal Reserve Bank of New York.

A more formal analysis

We explore the relationship between balances and payments more formally using regression analysis. For each regime and for all types of banks, we regress opening balances on Fedwire credits, Fedwire debits, and the constructed variable of other net changes. We include bank and day fixed effects to control for mean differences between individual banks and daily aggregate shocks, and standard errors clustered by bank.

Estimates of the Scarce Reserves regime coefficients demonstrate a strong and statistically significant association between changes in Fedwire payment flows and opening balances. Estimates imply that a one dollar increase in Fedwire debits is associated with a $0.509 increase in opening balances. Additionally, a one dollar increase in Fedwire credits is associated with a $0.505 decrease in opening balances, illustrating a symmetry in how banks assess incoming and outgoing Fedwire payments when making decisions about opening balances. Finally, a one dollar boost (or credit) from other transfer services (such as ACH Where Fedwire Securities Service) is associated with a $0.503 decrease in opening balances.

The estimated coefficients change only slightly between different types of banks, suggesting that banks may follow similar strategies when adjusting their opening balances in response to changes in expected payment flows.

Let us now turn to the estimated coefficients of the regime of abundant reserves. Contrary to expectations that opening balances would be less strongly associated with payment flows, the estimated coefficients for the specification including all banks show that opening balances remain strongly associated with expected payment flows.

Although there are several differences between the types of banks, the most striking result concerns small national banks. Likely reflecting their high ratios of balances to debits described earlier, smaller domestic banks no longer appear to be adjusting their opening balances in response to expected changes in Fedwire debits. However, they are decreasing their opening balances by almost 1 to 1 when it comes to increases in Fedwire credits.

Take away food

The 2015 results for small domestic banks are consistent with the intuitive idea that (much) larger opening balances can reduce the strength of the link between payment flows and balances. An abundance of reserves can therefore alleviate a central constraint on intraday liquidity faced by banks.

Our results should be viewed with caution, however, because even though a bank may appear to have an opening balance large enough to meet its payment obligations, these balances may not be usable. Instead, balances may be used by banks to meet internal or regulatory intraday liquidity requirements, or otherwise be reserved for other uses that prevent them from being used to fund payment flows (as discussed elsewhere , including in this 2021 Economy of Liberty Street Publish). Indeed, in 2015, many medium and foreign banks held balances on average greater than their Fedwire debits, and yet we continue to find a strong statistical association between payment flows and opening balances for both types of banks. .

Perhaps more tellingly, when we re-estimate the regressions detailed above for the same four months in 2021, when aggregate reserves are above $3 trillion, the estimated coefficients for all types of banks are statistically significant and imply that changes in Fedwire payment flows were strongly associated with changes in opening balances, results consistent with those of the 2007 scarce reserve regime. As argued in this recent articlea difference between 2015 and 2021 could be that during those seven years banks may have reacted to the Fed’s abundant reserve regime by taking on a variety of claims on their reserves, perhaps making those holdings unavailable for fund payment obligations.

Photo: portrait of Adam Copeland

Adam Copeland is a financial research consultant in the area of ​​money and payments studies with the Research and Statistics Group of the Federal Reserve Bank of New York.

Catherine Huang is a research analyst in the Research and Statistics Group at the Federal Reserve Bank of New York.

Kailey Kraft was a summer intern with the Research and Statistics Group at the Federal Reserve Bank of New York.

How to cite this article:
Adam Copeland, Catherine Huang and Kailey Kraft, “With Abundant Reserves, Do Banks Adjust Reserve Balances to Accommodate Payment Flows? », Federal Reserve Bank of New York Economy of Liberty StreetOctober 12, 2022, https://libertystreeteconomics.newyorkfed.org/2022/10/with-abundant-reserves-do-banks-adjust-reserve-balances-to-accommodate-payment-flows/.


Disclaimer
The opinions expressed in this article are those of the authors and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System. Any errors or omissions are the responsibility of the authors.