Hi all, I have 3 queries around the 2019 April Question 1 iii This question is how required capital (SCR) is affected Whenever I am required to discuss the impact on SCR, I use this rule of thumb: SCR = Normal NAV – Stressed NAV SCR = Normal(A-L) – Stressed(A-L) This is the explanation given in the 2021 SA2 CMP, Chapter 11 page 18 - Standard formula calculation. So for whatever given scenario, I check what happens to the Stressed (A-L). Because whatever happens is considered a “stressed” event Query 1: Question a. I consider increase in equity as a stress event (Negative stress) I interpret this as improved investment return rather than higher equity value at a starting point ie: something happened that moved my equity level, rather than I simply start at a higher level There are two ways to analyse this: Approach (a) Thought process: My starting point is normal(A-L), something happened (increased equity). How does it impact my new position (Stressed(A-L)) ? From this perspective, increased equity increases fund value hence annual management charge (AMC) A increase, L decrease (Non-Unit Reserve decrease), Stressed(A-L) increase therefore SCR decreases Which makes sense, if your equity investment do good, then you are at less market risk Approach (b) However, it is also correct to say SCR reflects the capital required for a given risk taken, ie: Risk Based Capital If I analyse from this perspective, increased equity means I have more Equity backing Ratio, so I need more capital for this higher risk Therefore equity increasing means increased SCR I think they are both valid explanations, but there can only be one answer. Which one is correct? The examiner report takes (b) Query 2: Question c I realise this question is answered in another thread. I understood and agree with the reply there, however there are further follow up questions that contradict that idea. Approach (a) If I think about it logically, if bond investments are more likely to default (dropped credit rating), we should be worried about losing all our bond investments (Similar to default risk SCR). This argument leads to saying widened credit spread causes increased SCR Approach (b) However, Credit spread widening means bonds are more likely to default (or less liquid) The bond yield increases, bond value/price drops. So there is lesser bond value exposed to bond value, lesser risk exposure hence lower SCR This is the explanation in the other thread and this examiner report, which I don’t disagree. Again, approach (a) and (b) contradicts although both are logical. They give different conclusions. Query 3: Question a and c Based on the SCR analysis by: SCR = Normal(A-L) – Stressed(A-L) We know that the Required capital, SCR is dependant the BEL component through the L. So, (i) In question a, I discussed increased AMC will decrease SCR, reducing required capital (ii) In question c, if spread widening is due to increased illiquidity risk premium, the discount rate increases, BEL decreases, stressed(A-L) increases, therefore required capital reduces. However, the examiner report and ASET solution “parks” (i) and (ii) under available capital instead of required capital. Therefore I didn’t give any credits at all even I explained them perfectly under required capital. I think the reason they aren’t included under required capital is because SCR wasn’t interpreted as SCR = Normal(A-L) – Stressed(A-L) In the ASET, the commentaries for this question reads “the technical provisions, and so will not impact required capital but only available capital” This contradicts with the solution given for question (b), where the required capital used references on BEL and expense assumptions So I am really confused what is the correct approach. Can anyone explain this to me? I think this is a core SCR concept to be understood, many questions revolve around this. Thanks, Trevor
I think you are getting confused between the events that are stated as having happened within the question, and the stresses that are applied to determine the SCR. You have to tackle this question by thinking: what is the SCR like AFTER this event has happened, compared with how the SCR is BEFORE this event has happened. And when determining the SCR, you need to think about the standard formula SCR stresses, NOT the events that have just happened. So, for example, in terms of an increase in the value of equities: Yes, the non-unit part of the BEL will be lower than it was before after the equity values have increased. But this is true in both the Normal(A-L) and the Stressed(A-L) bit of the SCR calculation, and they will (broadly) cancel each other out when the company now does the SCR calculation (ie after the equity value increase). So that isn't going to have a significant impact on the SCR. The stress that is applied in the SCR calculation will still be a fall in equity values. As the company now has a higher equity holding than it had before, in absolute terms, the SCR for equity risk will be higher. [If Normal(A) was 100 before and the equity stress is a 40% fall, Stressed(A)=60 so a difference of 40. If Normal(A) is now 150, Stressed(A)=90 so a difference of 60.] I think that separating out the events that have happened from the SF SCR stresses should help you to see what the issues are here.
Hi Lindsay, From what you explained, did you mean I take these as the "new SCR after these changed", rather than "what is the SCR like under these stress event"? For example, I assume the company intentionally increase the equities, and then how is the SCR after this change. So normal(A-L) will be based on the higher SCR: SCR = Normal increased equity (A-L) - Stress increased equity (A-L) Rather than the company did not change the equities. "Increased equity" is a stress event itself. So only stress(A-L) will reflect the changes: SCR = Normal normal equity (A-L) - Stress increased equity (A-L) I understood it as the latter version because I am recalling what I learnt from September 2018 question 2(v), where they explicitly mention "increase in equity" as a sensitivity. I interpret a "sensitivity test" and "stressed scenario" as broadly the same thing, so considered only the stress(A-L). The Normal(A-L) is the current position without any changes at all.
For the SF calculation of Stress(A-L) the stress that is applied, in relation to equities, is a FALL in the value of equities. [For this type of product, this is the downside risk direction.] What you are referring to in the September 2018 paper (assuming that you mean Q1 rather than Q2) is a question asking about sensitivities and scenarios applied TO THE SCR. In other words, what happens to the SCR when these events (such as an increase in equity values) happens. It isn't talking about how the SCR is calculated.
You are right, I meant question 1. I am aware that an increased in equity is more of a positive event rather than stress. But if I were to strictly treat it as a "stress event" to calculate SCR, then I would look only at the stress (A-L). In this case however, I totally understand your explanation. But just to confirm, if the question asks "What is the impact on the SCR amount under the ____ stress scenario". In this case, it will be correct to look only at the stress (A-L), is this correct? Applying this logic back to Query 2 of my original question, is it implying credit spread widening is NOT a stress event (stress testing to calculate SCR), but rather what is our new SCR after this event happens. ie: bond value already decreased, how much more can I loss if the stress event were to happen on top of this.
It is asking about the IMPACT ON THE SCR of an event actually happening (in this case credit spread widening), not how to calculate the SCR. This again sounds like asking about the IMPACT ON THE SCR of an event happening, it isn't asking about how the SCR is calculated. So basically: Impact = SCR after event - SCR before event = {Base(A-L) after event - Stressed(A-L) after event} - {Base(A-L) before event - Stressed(A-L) before event} However, bits of the above are likely to cancel out with each other and I personally feel that breaking it down like this could well be over-complicating things. You need to be able to step back and have that high level of understanding of what is happening.
Thanks Lindsay, I think I understand them now. It is more of me needing to understand the question correctly and then thinking through the right way.
Hi @Lindsay Smitherman , Apologies for bringing back this topic. I understand the earlier discussion about how these changes will impact the SCR. However I tried to re-apply this knowledge in another question (Mock Exam 3, question 2 iii b) The scenario is, what happens to the solvency position when the backing asset is switched from an AAA government bond to a BBB corporate bond? Considering the Credit spread submodule: Using our discussion points earlier, the corporate bond is at higher risk of default, higher yield so reduced market value. Shouldn't reduced market value of bond (hence asset level) lead to reduced SCR, due to lower asset value exposure? Considering the Default risk submodule (Although we know in actual fact, corporate bond defaults should be under Default risk: Higher default risk simply means higher SCR to cover for higher default risk taken. Both perspective leads to different conclusion. Am I confusing myself, or misunderstanding the scenario?
Why a lower market value? If you have $100m worth of AAA bonds backing the portfolio and you sell them and then buy BBB bonds with the proceeds, you are going to have roughly $100m worth of BBB bonds now (minus a bit of transaction cost). You basically have the same overall size of portfolio that you had before, just in riskier bonds. The risk of spread widening and default on BBB bonds is higher than on AAA bonds, so the SCR goes up.
Ah right, I totally missed the point that selling and buying the bond isn't going to change the asset value (ignoring transaction costs). So SCR will be higher because in the event of spread widening, the we now suffer a "greater fall" (from the same position) since the yield is higher for a corporate bond. Thanks Lindsay.
That's right. Also, the spread widening event that EIOPA requires to be applied under the SCR will be a bigger change in yield for the riskier bond. Glad it has clicked!
Hi, I am still confused about the difference between April 2018 Question questoin 1 part (vi) and this question. What is the difference in approach? I mean why there should be a difference? an eg, increase in the mantainance expenses, would’t it decrease the asset value? Impact = SCR after event - SCR before event = {Base(A-L) after event - Stressed(A-L) after event} - {Base(A-L) before event - Stressed(A-L) before event} Above approach is for currency question? Thnaks
Hi - I'm not sure what you mean by there being a difference between those two questions (assuming you mean September 2018 rather than April 2018) - they are both asking about the impact on particular balance sheet items of specified events happening (albeit different scenarios and slightly different balance sheet items). In terms of the expenses example, bear in mind that we would be interested in the immediate impact on the balance sheet of each event happening. A scenario whereby maintenance expenses are now expected to be higher would not impact assets immediately.
Hi, Yes I am talking about the Sept 2018 paper. I was earlier getting confused, are the scenraios given in the question are the stresses to be applied to calculate SCR, but it seems event has already happened and now we need to calculate the impacts. It make sense from a point of view that stress scenario to calculate the SCR are prescirbed by the regulator so that won’t be change only the base position chages and then SCR and other items are calculated. I have one more query, in such question we get a scenario of increase/widening in spread Corporate bond spread = illiquidity premium spread + credit risk spread Which one they usally refer to in such questions ? Thanks.
The scenarios given in those exam questions (Sept 2018 Q1 part (vi) and April 2019 Q1 part (iii)) are events that can happen to the company, and we are being asked what the (immediate) impact would be on the company's balance sheet if that event happened. For example, if equity markets dropped significantly, what would happen to the size of the SCR that the company has to calculate (ie how would the SCR immediately after the event has happened compare with the SCR immediately before the event has happened)? That is a different concept from the stresses that are applied to calculate the SCR. As you indicate, if an SF is used then these stresses are prescribed by the regulator. If an IM is used, the calibration of the stresses is specified. Sept 2018 Q1 part (i) is asking about these stresses. Hope that helps to clear that up.
If the question refers to 'the spread' widening then it means the whole spread increasing. Since this could be due to the illiquidity premium and/or the default spread increasing, our answer should consider each of these possibilities.
So in a scenario where the spread has widened Asset will fall as they are values on the MV (and MV will reduce if the spread widens) And for the BEL, it depends on the matching adjustments used in the discount rate In case there is no MA used and risk free rate is not changed, we will not change the discount rate and hence there will be no change in the BEL. Anywhere we can fit in Volatilty adjustment here ? Thanks,
Yes - could do. If a company is permitted to use the VA then: - widening spreads in the general market should be experienced in the representative portfolio - hence this should be (partly) reflected in an increased VA, as set by EIOPA - so higher discount rate and lower liability (BEL) value But because the VA doesn't represent the full spread change, the liability fall wouldn't be expected to be as large as the asset value fall [The above covers the impact on assets and BEL if the spread widening event actually happened. If we were instead thinking about the calculation of the SCR under the spread widening stress, then I understand that the increase in VA for the 'stressed BEL' could only be assumed to happen if the company was using an internal model, not under the standard formula approach.]