Capital project Appraisal

Hi, in the chapter on project appraisal it is said that the risk discount rate should be determined by the systematic risk of the project. From what I can understand the specific risks should be allowed for in the cashflows. If a life company were to launch a new product what would be the systematic risk and what would be the specific risks?

 

This is harder than I expected.

Specific risks would be:
Variations in mortality experience,
Getting the mortality assumptions wrong,
Possibility of poor sales, bad product design, misjudged demand,
Possibility of poor management decisions

Systematic risks would be:
Poor investment returns,
Poor sales/higher lapses due to recession,
Higher than expected inflation (of expenses, and possibly of claims)
Certain tax or regulatory changes

One question I have is about how specific risks get reflected in the price. We're supposed to use expected cashflows, so these don't really take into account the variability in profits due to specific risks. For example if this were a new and unique product, then the uncertainty as to its success (as a result of not knowing how customers will respond) doesn't get reflected in the price. The NPV will only be lower (compared to a minor variation on an existing product, with the same degree of specific risk) if we expect the profits to be lower. There might be more risk, but if we believe there is as much chance of the product taking off really well as there is of it failing, then the expected value could be the same or higher. Is this right?

If so, then why mitigate specific risks? By taking out reinsurance to reduce variability of profits due to mortality for example, we expect to pay out more in claims to the reinsurer than we expect to get back in claims. This reduces our expected profits. But since this is a specific risk, then a reduction in the risk doesn't decrease our discount rate, so there's no way that taking out reinsurance can increase the NPV of the project.

Have I gone wrong somewhere in the reasonsing?

Regards,



Sam

 

From what you seem to say it looks like the economic risks seem to fall under systematic risks and non-economic ones fall under specific risks.....

I think variability of cashflows cannot be accounted in the expected cashflows and as a result should be reflected in the risk discount rate. Am I correct?

With reinsurance, allowing for it in the cashflows should reduce the NPV as future profits should be smaller.

 

That's what causes me some confusion: variability of cashflows affects the NPV via the discount rate - but this only refelcts systematic risk. So what incentive is there to reduce specific risk? It seems that the only concern with specific risk factors should be to generate the highest expected profits with no regard to variability.

I understand why reinsurance might well reduce the NPV of a project. I just don't understand how you would explain (from a capital project appraisal point of view) why anyone would use it, since it has no impact on systematic risk and therefor no effect on the discount rate.

I have thought of some possible reasons:

It reduces risk of ruin, and thus actually increases some future expected cashflows.

It increases the amount of policies that can be sold, thus increasing opportunities for cross-selling.

It makes the outgoes more certain, enabling management to deal with them more effectively. This could increase expected cashflows. It could also reduce systematic risk and increase NPV that way - for example allowing better matching (less exposure to assets falling in value more than liabilities).

I feel like I'm getting the hang of this, so I hope this makes sense!

 

I think with specific risks they can be allowed for by reducing the expected cashflow amount. We can just attach a higher probability to the unfavourable outcomes which will result in a lower expected value. Variability is not always a bad thing and it is downside variability which should be more of a concern.

This thing can get quite complex and confusing. I am hoping the webinar on Thursday will clarify this whole risk discount rate issue

 

I may have confused myself once again, but here goes . . .

I'm starting to think that the difference between specific and systematic risk might not be all that well-defined. After all, we can't completely diversify away certain components of risk in practice.

And CAPM can be derived wthout making this assumption and without assuming that the (factors influencing the) returns on the market are the only general factors influencing returns on individual assets. I seem to remember that these concepts were introduced halfway through CT8s discussion of CAPM, presumably to avoid lots more matrix algebra, but the result of CAPM can be derived without them (ie. with only a few assets, or with any correlation matrix you like). I'll try to find the proof of this, which I think I've seen in some online lecture notes.

I think CAPM is really saying that (asset return - r) = beta*(market return - r) holds because the market reflects the specific risks of individual assets, rather than because it diversifies them away.

As we have more and more assets, the covariances between assets become more important than the variance of individual asset returns in determining the variance of the market returns. And these covariances are likely to be largely the result of exposure to general economic factors. So the idea that a share's beta reflects the share's "systematic risks" is a useful approximation.

Likewise, there is not clear line dividing specific or systematic risk in practice. For example factors that strongly affect one particular industry sector will seem more systematic if that sector makes up a greater proprtion of the market as a whole.

Does this sound right?



Sam

 

Hi Anna

Thanks for your explanation. It now makes sense. So the key is:
Systematic Risk - unavoidable
Specific Risk - avoidable

So for example in a project if the variability of cashflows is unavoidable then it is systematic. In some cases the variability can be avoided (eg using reinsurance) then it will be a specific risk.

The trick is to put it into context.

Your explanation was brilliant and thanks for helping us SEE THE LIGHT!!!

 

Thanks for the reply, Anna.

I think I understand it well enough to answer CA1 questions now.

I tend to think of the context as an investor's portfolio of shares. The systematic risk of a particular investment is the contribution it makes to the variability of his overall portfolio. The specific risk is the variability of that investment that isn't reflected in his portfolio.

In the first example, someone who only invests in annuity companies will obviously care a lot about longevity risk.

In the second example, someone who invests only in the domestic market will obviously care a lot more about the performance of domestic stocks than someone who invests in lots of markets.

One thing I still don't understand is why it would ever make sense to mitigate a risk that isn't reflected our investors' portfolios, and which they therefore won't care about.

I'm sure I must have missed something important here, but I can't explain what it is.

Could you please give me some idea how to answer SBS's original question about what the specific and systematic risks are in a life assurance contract? Or is it more complicated than that, with some risks being to some extent both specific and systematic?

Thanks,



Sam

 

posted twice by accident

 

Thanks again for the reply.

I'm confident that I understand this now. A company will still want to mitigate specific risks that don't significantly affect its shareholders, so can we think of the cost of mitigating these risks as an agency cost to the share-holder? Mitigating these risk will benefit directors managers and employees by making their jobs easier and more secure, but won't benefit shareholders as much.

I always thought it would make sense for a company to mitigate its specific risks, but it seemed counter-intuitive that the cost of doing so would be an agency cost to share-holders.

I guess there are other issues to consider too, such as protecting policyholders. And shareholders might benefit from the mitigation of specific risk if it reduces risk of insolvency or big losses (thus increasing expected profits), or if it makes it easier for managers to run the business, and thus future profits can be increased through better management.



Sam