Stress testing better option than risk indicators: James

Under the Financial Markets Conduct Act (FMCA), managers must include a risk indicator to make it clear to investors how volatile a particular fund is. The risk indicator runs on a scale of one, indicating very low risk, through to seven.

It is calculated on the basis of the previous five years' returns, based on annualised standard deviations. The more the actual weekly or monthly return differs from the average, the higher the standard deviation and the higher the volatility.

But Richard James, of NZ Funds, said he was advocating for New Zealand fund managers to use a stress testing model instead.

"I think the [risk indicators] are problematic and backward-looking over a relatively short period of time. Risk is by nature episodic."

He said it was possible that in some cases no volatility would be reflected in the indicator for quite risky products.

A better model was to put together a portfolio and test it against major financial events of the past 30 or 40 years, he said. That would enable managers and advisers to tell clients what those events would have meant for their portfolios in terms of dollars and percentage loss and the time it would take to recover.

"Even the more sophisticated investors are not clear what standard deviation is. They don't care about volatility. They care about losing money and what it would take to recover. Standard deviation doesn't give a sense of that."

He said while it was comforting to follow global guidelines, there was an opportunity to do something cleverer in New Zealand.

John Berry, of Pathfinder Asset Management, agreed there were problems with the system. "I think the wider issue is that what is called a risk indicator measures volatility which is not how most investors define risk. Most investors think of risk as 'what is the probability of me getting my capital back?'"

He said the risk indicator measure could be misunderstood or wrongly interpreted by clients.

Comments from our readers

On 21 January 2016 at 9:44 am Richard James said:

Just to clarify, I actually said (and we do) stress test against the major events of the last 30 - 40 years. In our view, 10-15 years is not long enough to give a true sense of the types of events that can and do occur across the asset classes

On 21 January 2016 at 9:51 am Craig Simpson said:

Having run some preliminary calculations on over 100 KiwiSaver funds using available unit pricing and the formula proposed by the FMA it is possible for various funds over different risk profiles to have the same risk indicator, which I feel defeats the intended purpose of what the FMA is trying to achieve. Education of investors may help but I think this risk indicator measure needs some serious re-thinking.

On 21 January 2016 at 11:37 am Kimble said:

Because everyone believes and trusts back-testing done by every fund manager, right? Because people will have to trust the manager's analysis, they couldn't possibly do it themselves.

Even if the portfolio wasn't deliberately constructed to minimise the risk to the previous crises, there is still the likelihood that any honestly constructed portfolio that shows too much risk in those crises wouldn't see the light of day. Who is going to sell a product that would have imploded during the tech bust or Asian crisis?

"Yes, our portfolio would have lost 90% in 1987, but the reason for that is easily... hey, come back! Look! Look at this 12 page explanation of why... huh? its NOT a ponzi scheme, that doesn't even... HEY!"

So the portfolios that the public ends up seeing are nicely calibrated to survive all the swans that were once black. What does that tell us what about their positioning for the next still black swan?

Using Standard Deviation has it's (significant and numerous) problems, but scenario analysis doesn't overcome them all.

On 21 January 2016 at 11:46 pm Peter Urbani said:

Consider the following two funds A and B

Fund A has returns +3,-2,+3,-2,+3,-2,+3,-2,+3,-2,+3,-2

Fund B has returns
-2,-2,-2,-2,-2,-2,+3,+3,+3,+3,+3,+3

Both have Avg annual returns of +6 and
Both have annualised volatility of 9.05

Fund A has a Maximum Drawdown of -2
Fund B has a Maximum Drawdown 0f -11.42. Which one do you think investors would prefer ?

The fact of the matter is that volatility says that they are statistically the same. This is because volatility is not path dependant. It is also symmetric. Neither of these two thinks gel well with how people actually percieve risk.

See;

https://www.academia.edu/15672333/The_Good_The_Bad_and_The_Ugly...of_Risk_Statistics

and

https://www.academia.edu/1873890/Properties_of_an_Ideal_Risk_Measure

Stress testing better option than risk indicators: James

New Zealand could do investors a service by not following other countries' lead when it comes to talking about the riskiness of their investments, one fund manager says.

Thursday, January 21st 2016, 6:00AM 4 Comments

Comments from our readers

www.GoodReturns.co.nz