Thanks to their superior liquidity, benchmark bonds are the primary instruments used by market participants to take positions and adjust prices in response to new information. Those prices then serve as a reference for the rest of the bond market, thus helping to improve the market's efficiency. Benchmarks seem to arise spontaneously in deep and liquid markets. But can government policy lend a hand in creating them where markets are too small? It turns out that some governments have tried. Authorities in Indonesia, Malaysia and Thailand have designated specific bonds as benchmarks and fostered their liquidity, with the aim of improving the functioning of their bond markets.
We start by identifying exactly which bonds were designated as benchmarks. We then compare the liquidity of these bonds against that of others with similar maturities. We consider superior liquidity to be a necessary but not sufficient condition for true benchmark status. To assess liquidity with sparse data, we propose a ranking approach that combines three simple measures. We interpret as success any time that the designated bonds are the most liquid ones in their market segment. We also identify factors that influence how successful authorities have been.
We find that efforts to establish benchmark bonds in three emerging markets in Asia have been generally successful. Designated benchmark bonds are the most liquid bonds in their market segments in around 60% of months in our sample. In Malaysia, the success rate is 78%. Factors that contribute to this success include choosing as benchmark bonds those are (a) are already liquid; (b) have already served as de jure benchmarks at longer maturities in the past; and (c) will be issued during the month.
Benchmark bonds help to improve market efficiency. They seem to arise spontaneously in deep and liquid markets. Can governments help to create them where markets are too small? This paper examines three emerging markets in Asia where authorities have tried: they have designated specific bonds as benchmarks and fostered their liquidity. We identify exactly which bonds were the designated benchmarks. We then propose rank-order measures of liquidity and determine the extent to which these de jure benchmarks end up as de facto benchmarks in the sense of being the most liquid bonds in their maturity segments. We find that this occurs in close to 60% of months in our sample, covering a range of maturities for Indonesia, Malaysia and Thailand. We identify three factors that make success more likely: (a) choosing already liquid bonds; (b) choosing bonds that have previously served as de jure benchmarks; and (c) choosing bonds that will be issued during the month.
JEL codes: G10, G12, G14
Keywords: benchmark bond, price discovery, market liquidity, informational public good, recycling, de jure, de facto, wannabe benchmark, probit model, inverse Mills ratio
Read the full paper at: https://www.bis.org/publ/work830.htm