Hi, Apologies if this is a stupid question. Is it ok to think of Diversification as negative correlations and "An accumulation of risks" as positive correlations? Thank you.
You need to be a bit careful here. Have a re-read of page 2 of Unit 22. Diversification effects arise because the various risks from a company’s operations are not perfectly correlated. Perfect positive correlations will give rise to accumulations of risk.
Hi Darren, Okay thank you. So we will have diversification benefits even when say 2 risks are positively correlated, but not perfectly positively correlated? I think that's what I was finding confusing but I think I understand it now. In other words, if say correlations between 2 risks was +0.8, we will have some diversification benefits. However, if say correlations between the 2 risks was +0.6, then the diversification benefits would be even greater. So the lesser the correlations between risks, the greater would be the diversification benefits. If say correlation = +1 and we have perfect positive correlations, then that would give rise to an accumulation of risks, and would therefore increase capital requirements (unsure though if we should say that diversification is zero at this point, as thinking of diversification as increasing capital requirements doesn't sound correct).
Thanks Darren, Apologies I just edited my response to add that last part on perfect positive correlations. I'd really appreciate if you could clarify that one please. Thanks very much
Yes if there is perfect positive correlation there is no diversification benefit. That means if class A requires say 500 of capital and class B 300 then with perfect positive correlation the overall combined capital requirement will be 800.
To add on to the above thread, and I'm afraid this might be silly, could correlations be > 1? Say 2 risks are related in such a way that if 1 increases by 1 unit, the 2nd one increases by even more than 1 unit. I do understand if this is expressed the other way round it'll be different (i.e. will be less than one)
Recall from your earlier studies (or a google search) that the correlation co-efficient is between -1 and +1.
Hi Jammy, No, correlations by definition ranges from -1 to 1. If we get a value greater than 1, we must have made a mistake in the calculation somewhere.
