1. iii)c) Widening of credit spread. While considering the impact on available capital, we have considered the reduction in BEL due to increased matching adjustment. Just want to clarify few items practically on MA (Matching Adjustment). It is defined as spread on portfolio of matching assets less fundamental spread published by EIOA. Here, MA increases as the spread on matching assets is higher, right? Practically, would it mean that MA is adjusted whenever spread is affected? As MA is based on matching assets which are held till maturity, why would the spread affect the MA rate? Please help. Thank you.
1.iii) d) I could not understand how required capital will increase here. I have added a numerical example below only for counterparty risk module SCR (hence stressed assets lower). Can someone please suggest how will each of these components be affected on increased risk of default from reinsurance counterparties? Hence, how to justify the increase in required capital numerically. A. Unstressed assets Base scenario 200 B. Unstressed BEL Base scenario 80 C. Stressed assets Base scenario 180 D. Stressed BEL Base scenario 80 SCR =(A-B)-(C-D) 20
Simplistically: Credit spread = default risk premium + illiquidity premium The MA reflects the illiquidity premium part of this (the fundamental spread represents the default risk premium part) If the credit spread widens, this can be due to one or (more normally) both of these two components increasing. To the extent to which the widening is due to an increase in the illiquidity premium, this would be expected to translate into an increase in the MA. But the extent to which the credit spread widens due to an increase in the default risk premium component, would not be reflected in the MA. (In reality it's a little more complex than this, as the fundamental spread won't exactly equal the default risk premium part - but hopefully it clears up the basic principle)
Hopefully this now makes sense based on the answer above, but we can also step back and think about the matching principle to explain this: If the credit spread widens, the value of the matching bonds will fall and so the value of assets decreases. As they are matched, we would also expect the value of liabilities to decrease, and that can only be the case under Solvency II if the MA also increases (since liability discount rate = risk-free rate (unchanged) + MA).
Remember that, under Solvency II, assets include the expected present value of reinsurance recoveries, net of best estimate default losses. Therefore unstressed assets will reduce by a small amount, to reflect the best estimate expected default losses now being slightly higher. So let's say they reduce to 198. Reinsurance doesn't impact the BEL, so those figures are unchanged. Stressed assets will start from the slightly lower amount of 198, but then will be reduced further as now we need to deduct not the best estimate expected default losses, but the default losses that would be expected to happen under the 1-in-200 year adverse event. These losses would be greater for a reinsurer with a low credit rating than one with a high credit rating. So stressed assets here might reduce to say 170. Thus new SCR = (198-80)-(170-80) = 28. An increase. Stepping back from these figures, we are basically saying that: a lower rated reinsurer gives us greater risk of default losses, so we have to hold higher capital (SCR) to protect against that. Try to make sure that you have a good grasp of these key principles, rather than getting too absorbed into detailed figures or calculations - the former are much more important.
Thank you, this was very helpful! Regarding the SCR numerical example- I have been using Excel for such examples and it has been useful to draw conclusions on SCR stress related questions. At times, rationalizing with thumb rule/first principles hasn't worked in my favour. If you have any notes on how different types of shocks affect SCR (based on previous response in this forum or otherwise), feel free to share it here. Thanks again!
SCR is calibrated to Value at Risk, so basic rule of thumb = if the event increases risk, it increases SCR, and vice versa. So any deliberate action by the insurer that reduces risk (eg moving to a less risky / better matched investment strategy, or making more use of reinsurance etc) will reduce the SCR, and v.v.. And any event that causes the insurer's risk exposure to reduce will reduce the SCR, and v.v.. For example, if the event results in the insurer having fewer policies in-force (eg a mass lapse), then (all else being equal) it subsequently has less risk exposure -> lower SCR. If the event results in the insurer having a smaller equity holding than it did before (eg due to an equity stockmarket crash), then (all else being equal) it subsequently has less equity risk exposure -> lower SCR. And so on.
Can we also connect this to the risk margin concept in addition to SCR? For the non-hedgeable risks, usually when the risk increase, SCR increases and risk margin is expected to increase as well?
Circling back to credit spread widening and its impact on SCR. (Q1 iii c) As we know that BEL is also affected, why don't we talk about impact on SCR = credit spread risk SCR+ interest rate risk SCR? Since there is a matching adjustment in place, any widening of spread will surely affect BEL discount rate and then interest rate risk SCR will be affected. Typically for interest risk SCR, we do consider the impact on both fixed interest assets and BEL. Stress will be applied to both as both components are affected by change in interest rate (with or without MA). Is my understanding correct?
The interest rate impact is second order. Remember that SCR for a stress = {Assets before stress - BEL before stress} - {Assets after stress - BEL after stress} so there is a reasonable amount of offsetting. {BEL before stress} will be lower than it was before the spread widening (due to the higher MA), but so will be {BEL after stress}.
Yes. Here, it is mentioned that the scenario is- credit spreads widening on corporate bonds Why would the interest rate impact be second order? Credit spread does affect MA directly because of how MA is calculated. If MA wasn't applied, I agree that interest rate impact would be second order and I wouldn't have thought of including any SCR impact other than a fall in corporate bonds due to spread widening. Can you please clarify? The offsetting of BEL is surely there but when we think of interest rate risk SCR, as you agreed, stress will be applied to both fixed interest assets and BEL. However, lets say risk free rate doesn't change in spite of credit spread increase. In this case, increase in stressed BEL will offset increase in unstressed BEL and a fall in interest rate risk SCR would be expected. My main question is why just focus on credit spread risk SCR and not both credit spread risk SCR and interest rate risk SCR. The question just talks about total SCR impact hence I would have thought of both.
On second thoughts, are we saying that even when we think of interest risk SCR, this capital pertains to change in risk free interest rate only. Hence, increase in credit spread and matching adjustment will be second order impact as change in BEL discount rate is driven by spread increase here. But again, not sure how will we segregate interest risk SCR impact of change in risk free rate and change in spread as ultimately RFR+MA is applied for calculating BEL. Hence, how will the stress be able to distinguish between a second order impact and first order?
The stress doesn't distinguish between first and second order impacts. My point was that the interest rate stress impact is likely less material than the credit spread stress impact (because there is more direct offsetting) and therefore wasn't considered important enough by the examiners to give credit for (given the fairly low mark allocation).
Perhaps a numerical example will help to illustrate this. Let's say that: there are 100 of assets all in corporate bonds and BEL of 60 the interest rate stress is an increase in basic risk-free rates of 1%, which results in a fall for both assets and liabilities of 10% the credit spread stress results in a fall in value of bonds of 20% and a reduction in BEL of 5% (through the MA, which will only partially reflect the widening) the actual credit spread widening is an increase in spread of 2%, resulting in bonds falling in value by 20% to 80 this actual credit spread widening allows the insurer to add an MA of 0.5% to the basic risk-free rates, reducing BEL by 5% to 57 SCR = {A before stress - BEL before stress} - {A after stress - BEL after stress} Credit spread stress: SCR before credit spread widening = {100 - 60} - {80 - 57} = 40 - 23 = 17 SCR after credit spread widening = {80 - 57} - {64 - 54.15} = 23 - 9.85 = 13.15 So the SCR has fallen by just under 4 Interest rate stress: SCR before credit spread widening = {100 - 60} - {90 - 54} = 40 - 36 = 4 SCR after credit spread widening = {80 - 57} - {72 - 51.3} = 23 - 10.7 = 2.3 So the SCR has fallen by 1.7 = lesser impact
Alright. Thank you for clarifying! Sometimes it is just not clear to me how much depth is expected for SCR impact justification although it seems like not mentioning interest rate SCR will not be penalized here as you said.
I get what you are saying: keep a close eye on the number of marks available, as that should help with this.
