Download this article in PDF format
Choose your country :
Choose your country ?
In this document we describe how we build scores for corporate bond issues that are updated every day on the basis of timely characteristics. We also describe how the scores are used by the fund managers in the construction of corporate bond portfolios.
Since the level of market liquidity is dispersed and is generally low for corporate bonds, it is advisable to take liquidity into account when investing in such bonds. It is wise to weigh the effort it takes to wind or unwind positions into the bond selection process. To be able to do so the level of liquidity must be quantified in some way. To this end Barclays -among other data providers- has developed liquidity cost scores (LCS) for all members of their Global Corporate Bond index on a monthly basis since 2009, see Konstantinovsky et al. (2015).
The specification of Barclays’ model, given in equation (1) in next section, is straightforward: the liquidity score of a bond is the sum of its attributes weighted by the respective model parameters. The parameters are calibrated onto the set of bonds that trade regularly for which bid-ask spreads are registered. Estimates are obtained through standard cross-sectional regression analysis. The model is then run for the lowly-traded bonds making the assumption that the same relationships between liquidity and bond attributes apply.
1 See Ben Dor et al. (2007).
2 The liquidity scoring model
2-1. Model specification
Barclay's liquidity scoring model is specified as follows:
Region & currency
Coupon & outstanding amount
Age & time-to-maturity
Yield, price, DTS & peer status
Peer status is defined by three possible configurations: a bond is the single issue of a company, or otherwise it can be either ‘on-the-run’, meaning the most recent issue, or else ‘off-the-run’. To give an idea, in April 2015 5% of the bonds were single, 83% off-the-run and 12% on-the-run. A difference in liquidity level is detected between the latter two in the literature, by Helwege et al. (2013) for instance, no comparison is made with the single status.
2-2. Model estimation
The model has been estimated by running cross-sectional regressions of the bond attributes on the Barclays liquidity scores. We have carried out forty such regressions, one per month between January 2012 and April 2015. All bonds belonging to the Barclays Global Corporate Bond Index and being domiciled in North America or Europe have participated. We have retained the variables that are statistically significant, intuitive and relatively stable over time. We have run the usual validity checks necessary for a linear regression model, in particular we have verified that there is no co-linearity issue. The parameter estimates as for April 2015 are given in Table 3.
We recall that the LCS have the opposite signs of the liquidity levels. A positive sensitivity to age for example, with a coefficient of 0.15, means that the older a bond is, the higher the liquidity cost so the less liquid it is. Yield, DTS and price go the same way: the higher their levels the lower the liquidity. The coupon and outstanding amount have the opposite impact (negative coefficients). So the bigger the total debt amount, the lower the cost so the more liquid the bond, which is intuitive.
Interestingly we find that bonds denominated in their local currency, i.e. European bonds in euros and American bonds in dollars -denoted EU, € and US, $- are significantly more liquid than those domiciled abroad. Within that European bonds seem slightly more liquid than US bonds, with a coefficient of -0.26 versus -0.24. In accordance with the literature we find that on-the-run bonds are more liquid than off-the-run bonds on average, with a coefficient of -0.18 compared to -0.15. However this relative result is overshadowed by the third category, the single bonds which are substantially less liquid than multiple issues, though we find no reporting of this phenomenon in the literature.
We establish the frequency distributions of the liquidity scores over the test sample and period. They are given in Figure 4. On the left are the Amundi scores, derived by means of model (2), and on the right those published by Barclays.
The scores don’t appear to be normally distributed. A relatively small portion has high scores and are thus less liquid than average. Note that all Barclays scores are positive. This conditions must have been imposed, which indeed makes sense. We do the same: the few scores that end up negative in the estimation process are set to zero.
3 Managing the liquidity of a portfolio
The individual bond scores can be aggregated within a bond portfolio to get an idea of the general liquidity level. It may be useful for investment managers to monitor this level for risk control purposes. More interestingly the liquidity level can be proactively managed by integrating the scores into the portfolio construction process. We describe how this can be done.
An explanation for the positive result can be found by looking at the initial selection criterion, the DTS contribution. Both duration and spread are unfavourable for the liquidity of a bond. The longer the duration and thus time-to-maturity and the higher the spread level, the less liquid the bond tends to be. Adding the liquidity scores must offset this bias towards illiquid bonds to a certain extent.
4 Ongoing research
We are aware2 that individual bond scores derived through cross-sectional regressions are effective for making relative comparisons across bonds, but less so for analysing liquidity levels over time. There is some time-series information incorporated in the Liquidity Cost Scores stemming from the fact that they are anchored on the bid-ask spreads registered over time by the Barclays trading desk. However this information is remote and incomplete. We consider to improve the time content of the scores by adding a term to our model that captures a general market condition. We are looking at a difference in spread level that can
2 Guy Lodewyckx made the point.
Amundi has built its own liquidity scoring model
Amundi liquidity scores based on ten bond characteristics
No notable difference in liquidity level on average over sectors
Single issues substantially less liquid than multiple issues, wheter it be on-or -off-the-run
Liquidity scores inserted into the objective function of the protfolio optimizer
Portfolio liquidity level can be improved at little cost
Download this article in PDF format
Doubts are rising whether bond indices, in the way they are constructed, are effective in their role of representing the markets they are designed for. Since index constituents are defined on market shares –the larger the debt obligation, the larger the share in the index– it may be that certain risks related to a high level of indebtedness are being accentuated which are not necessarily representative for the market as a whole.
Marielle de JONG, Lauren STAGNOL