追踪信用违约互换（CDS）堕落天使的表现

Restricted - External Jigar Patel+1 212 412 1161jigar.n.patel@barclays.comBCI, US Despite the lack of fallen angels in recent history, the WHR downgrade and ongoing concernsabouttariffs(albeit less so recently), the strength of the consumer, and the growth outlook haverenewed focus on fallen angel performance in the CDS market.Using the 10 fallen angel constituents (ex-pandemic) of on-the-run CDX.IG from 2017 to April2025, we examine fallen angel performance relative to CDX.IG (using spread ratios) before andafterthe downgrade, and before andafterthe roll. We analyze how the average and medianspread ratios change over those periods, and we also look at how many of the 10 fallen angelsparticipated in the widening (versus CDX.IG) in the 20 business days before andafterthedowngrade and the roll.Performance into the downgrade: Broad-based wideningFallen angels tend to underperform CDX.IG into the downgrade to high yield, in line withintuition, as investors start pricing in the weakening in fundamentals. Both the average and themedian spread ratio move higher into the downgrade, with the median ratio rising more rapidlyabout 9-10 business days ahead of the downgrade (Figure 3).We look at performance as a ratio versus CDX.IG in order to control for overall marketperformance. Looking at performance in absolute terms instead provides similar results, withspreads widening 45-50bp on both an average and a median basis in the 20 business days priorto the roll. On a percentage basis, spreads widen 23% on average and 21% on a median basisover that period.To examine the pervasiveness of the underperformance, we compare the spread ratios on eachof the 20 business days before the roll with the spread ratios on the downgrade date (Figure 4).This gives us an indication of: 1) how many fallen angels are contributing to the aggregate2 FIGURE 3. Fallen angels in CDX.IG have tended to underperform intothe downgrade, which aligns with intuitionFIGURE 4. The underperformance into the downgrade has historicallybeen broad-based0%10%20%30%40%50%60%70%80%90%100%-1-3-5-7-9-11-13-15-17-19# of Business Days before Downgrade to HYPct of Time Spread Ratio is Higher by Downgrade DateSource: Barclays Researchperformance on any given date; and 2) if there are any particular dates when the performanceinto the downgrade date stands out.The underperformance versus CDX.IG is broad-based for most of the pre-downgrade periods,but starts to dropoffas we get inside of five business days. For example, positioning for thespread ratio to move higher nine business days before the downgrade would have worked 90%of the time (9 out of 10 fallen angels), but doing the same three business days before the rollwould have worked only 60% of the time.Performanceafterthe downgrade: Modest underperformanceWe also consider performance once the credit has been downgraded to high yield. As areminder, CDX.IG constituents are not removed from existing CDX.IG series. Once they aredowngraded, they are no longer eligible for any new CDX.IG series, and therefore are "removed"from the membership of the next series.The underperformance versus CDX.IG continues in the initial post-downgrade period, but isrelatively modest, as some investors presumably unwind long-protection positions once thedowngrade event occurs (Figure 5). Performance then appears to largely leveloffaftertwoweeks.Looking at how many fallen angel spread ratios are higher than their downgrade-date levels bybusiness date, the majority in our sample (60-70%) maintain their underperformance post-downgrade (Figure 6). Unlike the bond market, where fallen angels may encounter demandfrom high yield investors as they are added to high yield cash indices, no such post-downgradetechnical exists for CDS. 3 Performance into the roll: Tightening, then wideningUsing the same methodology as above, we evaluate fallen angel performance into the roll date(or, more precisely, into the day before the roll). As a reminder, all of the credits in our analysiswere members of on-the-run CDX.IG that had been downgraded to high yield while they were inthe index, and therefore we are measuring their performance into the roll before they areremoved from the next on-the-run index. Overall, performance is much more muted (Figure 7).Aside from a large drop in the median ratio at the beginning of our analysis period (due to onecredit), the median ratio moves modestly lower until 10 days before the roll, and then moveshigher. However, the average ratio is little changed, with a slight move lower until five daysbefore the roll, and then a slight move higher.We suspect that the tightening bias likely has to do with the typical buyers' strike ahead of theroll, as buyers of protection may be reluctant to add new risk ahead of the roll. The slightwidening in the five days before the roll could be due to some investors trying to get ahead ofskew-package rolls from the roll date onward. As a reminder, skew packages are owned by indexarb account