partner content

Time, trust and credit-risk exposure in a trade war

Adopting a credit-risk model based on issuer-specific curves and relevant cluster curves allows asset owners to capture changes in investor sentiment as they occur. This is especially important during turbulent periods such as we face now.
Time, trust and credit-risk exposure in a trade war

The world of fixed income is inherently quantitative; it involves managing a three-way tradeoff between reward (interest-rate risk), trust (credit risk), and time (spread duration). For Asian high yields, managing that tradeoff can be especially tricky—but a granular risk model can help.

Let’s take some analogies. Say a close relative wants to borrow $100 until his next paycheck. He offers to repay you $105 at that time. Being a relative, you’re probably willing to waive the $5 interest (or not!), and given that he has been wearing the same threadbare sweater at every family gathering for the last five years, you trust that he will not incur any extravagant expenses within the next 30 days. The bank is only willing to give you $2 interest for your $100 and it appears unlikely they will be willing to pay more any time soon. Therefore this $5 is probably the best game in town for a while, and your exposure is only for a month. So, sure, let’s do it.  

Now say an old high school classmate you have not heard from in 20 years—and didn’t trust back then for good reasons—calls you to borrow $100 and promises to repay you $120 in a month’s time. You’re tempted by the potential for a $20 reward—which beats anything your bank, the US government, or any relative is willing to offer. But given your lack of trust on this borrower you’re probably only willing to trust him for a maximum duration of a week.

So, either you get him to accept your terms for a week’s duration or, if he insists on a month (and you are really desperate for the $20), you need an “out”. Meaning that if he suddenly goes back to his old high school ways, you can still sell that ‘IOU’ to someone else—maybe that close relative of yours? —for at least your original $100, and let him be exposed for another three weeks to get the $20. In short, time acts as leverage on trust. The longer you are exposed, the longer you need to trust the borrower. 


Fixed-income portfolio managers have developed a metric for comparing this credit-risk exposure between portfolios. They call it Duration Times Spread (DTS). Essentially it is a measure of the sensitivity of your portfolio to a sudden relative change in credit risk. So, in our examples above, DTS would measure your response to that close relative showing up at dinner the next day with a brand-new cashmere sweater or to seeing a Facebook post of your friend at the craps tables in Macau. Since the duration (one month) is the same in both cases, the relative change in your DTS for each “loan” can be entirely attributed to the change in trust (spread duration) and would be characterised as a sudden rise in your “exposure” to credit risk.

The reason fixed-income managers like DTS as a measure of credit-risk exposure is because over short periods of time, roughly speaking, the portfolio duration and the underlying spread volatility do not change all that much, but the actual credit risk itself can jump in a single day due to a sudden crisis. In this case, the DTS jumps immediately and so does the portfolio’s measured exposure to credit risk, whereas in traditional spread-based models the exposure hardly moves due to a single one-day spike in spreads. In this sense, DTS is like a daily indicator of trust jitters.

The more granular your fixed-income credit-risk model, the easier and more accurate the math behind DTS gets, which is especially important when it comes to quantifying your exposure to something as ephemeral as ‘trust’. In traditional spread-based models, only time (duration) affects your exposure, but in Axioma’s Fixed Income Credit Risk Model, the product of both time (duration) and the issuer’s credit risk (spread) will be used to accurately compute your exposure (DTS) and, more importantly, allow you to monitor the relative change in this exposure as time progresses. 

With this in mind, let us contrast the change in DTS for two portfolios on several recent key conflict escalation dates in the ongoing US-China trade war since May of this year. Our ‘close relative’ will be proxied by a portfolio of US investment-grade corporate bonds, and our high school classmate will be proxied by a portfolio of Asian high yields. Both are denominated in US dollars. The table below shows the summary statistics for duration (time), spread (trust), and DTS (exposure) for the two portfolios across four sets of time periods: May 6-13, Jul 30-Aug 5, Aug 13-15, and May 6-Aug 30 (start to finish). We note that during the entire period of May 6 to Aug 30, the S&P 500 was essentially flat (-0.2%).

Source: Axioma


  • The spread (trust) meter on the Asian HY portfolio is considerably higher (i.e., we have about one-fourth the trust) than on the US IG portfolio and therefore we’re only comfortable holding this loan for about a third of the duration (time).
  • At the start of this analysis on May 6, holding the HY portfolio for about 1/3rd of the time of the US IG portfolio, given its much higher spread (lower trust), gave us about the same credit-risk exposure (DTS) of 13.75 vs 13.39.
  • As the trade war between China and the US escalates, we see a rapid change in investor discomfort. At first, both portfolios are equally affected, but in the latter stages, investors are clearly more nervous about what this means for the Asian HY portfolio than the US IG one (about 2-to-1 more nervous).


The escalating trade war has made US corporate bond investors a bit more nervous about lending to these issuers. DTS for the US IG portfolio increased by 4.3%, 2.4%, and 7.8%, respectively, between the selected dates, and by 17.7% for the entire May-Aug period. In comparison, the S&P 500 index declined by 4.1%, 5.6%, 2.7% and -0.2%, respectively, between those same dates. But look at our Asian HY portfolio, where credit-risk jitters there increased by 3.6%, 7.3% 14.2%, and 19.6% for the same sub-periods.

During this time frame, the US IG portfolio’s duration and spread remained relatively constant, so we can attribute the higher credit-risk exposure to an increase in spread volatility (i.e., increased nervousness). As for our Asian HY portfolio, despite a shorter duration (time), the spread increased from 446 to 567, hence the rise in credit-risk exposure reflects a combination of decreased trust (higher spread) and increased nervousness (spread volatility). 

Having a granular credit-risk model based on issuer-specific curves and relevant cluster curves allows this math to be automated on a daily basis and to capture changes in investor sentiment as they occur. An accurate quantification of the drivers of increased credit-risk exposure is vital to making the right risk management decisions. The granularity of this model allows the portfolio manager to drill-down to the individual bond position and view the change in investor sentiment for each, in order to decide on appropriate trades to efficiently reduce the portfolio’s credit-risk exposure.

Author: Olivier d'Assier, head of applied research, APAC at Axioma

¬ Haymarket Media Limited. All rights reserved.