Hacker News Reader

Comment by nickles | Hacker News Reader

> Discount rate is, by design, quite a bit higher than the fed funds rate, and is only tapped for credit in emergencies.

That's right. Open market operations were traditionally used to target the fed funds rate. The discount window does serve as the upper bound though in extreme cases. Typically, reserves would be loaned overnight between banks on the fed funds market. This market has been decimated by the floor system.

> IOER, by design, provides just a floor.

In theory, yes. It has also acted as a ceiling. Not all institutions with access to Fed's balance sheet are able to receive IOER (e.g. GSE's). These entities loan the money to banks overnight at a rate below IOER. The banks then earn IOER on that money. This arbitrage can lead to an effective fed funds rate below IOER.

View on HN

> This market has been decimated by the floor system.

You mean there has been so much cash around that nobody really needed to get cash via the fed funds market, for quite a while? It does seem like the spread between fed funds and IOER is generally just a few bp, maybe even 1bp sometimes, which I guess is where the "ceiling" you were talking about comes from:

https://fred.stlouisfed.org/series/IOER

https://fred.stlouisfed.org/series/FEDFUNDS

> You mean there has been so much cash around that nobody really needed to get cash via the fed funds market, for quite a while?

Prior to IOER, banks needed to lend money on the fed funds market to earn interest on any excess reserves. Doing so exposed them to the credit risk of their counterparty. Once it became possible to earn IOER risk free, banks had no incentive to expose themselves to the credit risk. Since banks (of all sizes) could not then rely on being able to access the fed funds market for liquidity, it became necessary for them to keep more reserves at the Fed. Perversely, as the quantity of reserves held at the Fed increased, interbank lending decreased (for more detail, see [0] pages 33-39).

> It does seem like the spread between fed funds and IOER is generally just a few bp

Until recently, it looks like the spread has been roughly 5-17bps [1]. For a floor system, it seems rather odd that this would exist at all, especially for a period of a decade (see [0] pages 18-19).

[0] https://www.cato.org/sites/cato.org/files/pubs/pdf/working-p...

[1] https://imgur.com/n3R366c

All makes sense.

Thank you for many details, you've greatly increased my (and hopefully other people's too) knowledge of the United States money markets.

Reply on news.ycombinator.com