Measuring Bond Portfolio Liquidity with Spreads and Risk Proxies

Summary

The discussion considers whether a portfolio bond-liquidity score can be built from bid-ask spreads and a published liquidity score. It distinguishes the relative spread, calculated against the bid price, from the absolute spread, and questions whether weighting observations by dividing aggregate spread by each bond’s spread yields a meaningful portfolio measure. The question does not establish that weighting method as valid.

The response notes that liquidity can be asymmetric: selling pressure may be greater during market declines than buying pressure during rallies. It suggests incorporating trading volume, adapting an illiquidity measure to account for this asymmetry, and considering spread-based measures such as DTS, since liquidity may be reflected in credit spreads. It also contrasts corporate bonds, which are generally less liquid, with developed-market sovereign bonds. These are qualitative suggestions rather than a tested comparison; the excerpt gives no data, calibration procedure, or evidence that one proxy is best for a particular portfolio.

Key ideas

• Relative bid-ask spread divides the spread by the bid price, while an absolute spread does not.
• The proposed portfolio weighting based on total spread and each bond’s spread is questioned but not validated.
• Bond liquidity can differ between buying and selling, especially under downside pressure.
• Volume-aware illiquidity measures and spread-based proxies such as DTS are possible alternatives.
• Corporate bonds are generally less liquid than developed-market sovereign bonds.

Tags

# Measuring liquiduity of a portoflio of bonds


# Measuring liquiduity of a portoflio of bonds












I'm currently looking into applying bond liquidity out of curiousity.

The Method i'm currently using is the Barclays LCS score (live.barcap.com/publiccp/RSR/nyfipubs/barcap-email-mkting/qps/LCS_In-brief.pdf)

which states that a possible way to grade a bond liquidity is simply

$Ask-Bid/Bid$

What I do is I measure the spread

$Ask-Bid $

This is the part where I'm doubting i'm going the correct way.

I take a total of the entire spread and divide it by the spread of each asset to get a weighted number.

This weighted number is then multiplied by the LCS Score.

$(Ask-Bid/Bid)*LCS$

My question is if this is a good way to determine liquidity with the limited means I have.

Thank you in advance for any comment/help you can provide.

## Answer by Vitomir (score 1, accepted)

https://quant.stackexchange.com/a/46818

It does not take into consideration the fact that liquidity is not symmetric, also in Fixed Income markets. Indeed, there is much more liquidity pressure on the downside than on the upside. I suggest you reading liquidity black-holes. as a general rule, corproate bonds are not very liquid compared to DM sovereign ones. Looking at Ahimud allows to take into account Volumes, as well as you can modify it to account for asymmetry above described. A general proxy is DTS, as liquidity is somehow priced into spreads.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.