RDP 2024-03: Demand in the Repo Market: Indirect Perspectives from Open Market Operations from 2006 to 2020 1. Introduction

In some financial markets where trading occurs bilaterally between two counterparties (over-the-counter) rather than through an exchange, it can be difficult to obtain reliable metrics that allow analysis. Sometimes this is because when prices are bilaterally agreed between counterparties, there is no public record of the transaction price or volume traded. This lack of information is compounded in those over-the-counter markets where there is no common electronic platform that quotes executable prices. There might also be a significant difference in transaction characteristics between counterparties depending on the degree to which they are integrated into the overall market network. Peripheral participants that deal with larger participants at the core of the market might be subject to different pricing practices and unable to easily substitute to other forms of funding. Many of these considerations are relevant to the Australian repo market. Market conditions are therefore not easy to assess and are often dependent on anecdotal evidence.^[1] The Reserve Bank has been analysing the repo market by providing indirect perspectives using data from its own open market operations that are conducted using repos and by publishing data on repo market activity.

Prior to the COVID-19 pandemic, the Reserve Bank typically conducted open market operations to provide settlement balances for the smooth functioning of the payments system, manage liquidity in the interbank market, and for the implementation of monetary policy.^[2] Lending cash against collateral under repo is an integral part of these operations. The eligible private sector counterparties to the Reserve Bank in these operations have a variety of reasons for participating. We use the preferences they reveal through behaviour in the auction bidding process to draw inferences about the overall demand for repo and hence market conditions.

Arranging cumulative repo rate bids from highest to lowest is akin to estimating a downward sloping demand curve. This mimics the auction process where bids are filled, beginning with the highest, until the central bank's willingness to supply is exhausted. This is a simple way of assessing the relationship between the price and quantity outcomes observed in open market operations. However, there are some caveats when drawing inferences about the broader repo market. Importantly, the Reserve Bank only deals with those participants accredited to be eligible, and hence our analysis omits important segments of the market such as highly leveraged institutions. The private market itself also has more international linkages than the central bank requires for the conduct of domestic monetary policy operations. Finally, the terms at which the Reserve Bank operates are not very representative of typical market practice. The market typically contracts repos at much shorter durations than the central bank and has different standards with respect to eligible collateral.

Nonetheless, this framework contributes to our understanding of the repo market in three ways. Firstly, the properties of the estimated demand curve, such as the slope and direction of shifts, allow us to infer how attitudes toward repo and its importance as a funding source change over time. Secondly, the interaction between demand and supply indicates the market-clearing equilibrium price and characteristics of unsuccessful bids. Thirdly, we can address hypothetical questions related to repos. For example, the supply of liquidity required to compress the equilibrium price to a particular repo rate for an estimate of demand on a given day.

The remainder of the paper explains our approach to these points. Section 2 provides an overview of the Reserve Bank auction process for repos prior to March 2020, which is key to understanding the data-generating process. In Section 3, the methodology of aggregating bids is illustrated through a worked example. In Section 4, the methodology is applied to actual auction data, and in Section 5 the elasticity of demand is estimated. The final section offers some concluding comments.

Footnotes

For an earlier description of the Australian repo market see also Wakeling and Wilson (2010). [1]

In March 2020, the Reserve Bank announced a comprehensive package to support the Australian economy and made changes to the way monetary policy was implemented. At the same time, it also announced that open market operations would continue, albeit with a different focus. Aspects of operations described in this paper are mainly relevant for understanding repo market conditions in the pre-pandemic era. [2]