CS01 and Relative Spread Shocks for Credit Risk Monitoring
Summary
The document explains two ways to describe credit spread exposure. CS01 estimates the first-order profit and loss from a one-basis-point spread move, making it convenient to translate a specified absolute spread change into an approximate P&L impact. The answer notes that gamma can supplement CS01 when a better P&L explanation is needed, since CS01 alone ignores curvature.
It also describes limitations of absolute spread bumps. Aggregating one-basis-point sensitivities across very different spread levels can be misleading, and large parallel widening shocks may produce implausible curve shapes or negative implied hazard rates. A relative spread shock scales each spread proportionally, so a given percentage change yields different absolute moves for tighter and wider credits. The response argues that this is often a more realistic approach for credit stress scenarios. It does not define a universal convention for CS01 sign or the relative-spread measure, and emphasizes that terminology varies; users need to be clear whether they represent tightening or widening.
Key ideas
- CS01 approximates P&L for a one-basis-point credit spread change.
- CS01 omits spread convexity, which may matter for larger moves.
- Aggregating absolute spread sensitivities across very different spread levels can obscure risk.
- Relative spread shocks scale moves to current spread levels and can form more realistic stress scenarios.
- Large absolute shocks may create implausible curves or negative implied hazard rates.
Tags
Full text
# the difference between CS01 and RS 1% # the difference between CS01 and RS 1% Please tell me the difference between CS01 and RST 1% (Relative spreads tightening by 1%) and how these two are used to monitor the credit flow traded product's exposures. Why would you use the spread tightening to monitor the risk? ## Answer by Dimitri Vulis (score 1) https://quant.stackexchange.com/a/73775 There is no standard nomenclature, but: Often "CS01" means the P&L impact of credit spreads changing by 1 bp - the credit spread delta. It's often used as a risk measure by credit trades. Some people prefer to look at spread tightening, others at widening, it does not matter as long as you remember which one you use. CS01 is convenient because if a paricular spread changes by some number of basis points, and you multiple the change by CS01, you immediately get a first-order estimate of the resulting P&L. If you also add in the credit gamma, you can have good P&L explain. The disadvantages of CS01 include: - it ignores the convexity (credit spread gamma), which can be quite material if the credit spread moves by tens of basis points. - if the book has both investment grade (e.g. 30 bps spread) and high yield (e.g. 700 bps spread), then aggregating their CR01s is like comparing apples and oranges. It's almost like having a penny stock and BRK and asking what happens if all stock prices change by 1 dollar. - this seldom happens, but if the spread is very tight, then perturbing one tenor by 1 bp may result in a curve that admits arbitrage (has some negative hazard rates). - if you use this methodology for stress testing - i.e. don't ignore convexity, but ask what happens if all spreads widen, e.g., 100bps, then you may end up with a curve shape that admits arbitrage; or a shape that doesn't admit arbitrage, but still does not look like something that would arise in the market. A relative, rather than absolute bump, works better for credit stress testing. Risk scenarios where every credit spread widens, e.g., 1%, 10%, 50%, i.e. a 100 bps spread becomes 101, 110, 150 bps, whereas a 200 bps spread becomes 202, 220, 300 bps, more realistically describe what happens in the market than everything widening the same 100 bps.
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.