volatility is not a true measure of risk

[Brendan Burgess]

I have always argued that volatility is not a measure of risk for a long term buy and hold investor like myself. It might be a useful measure for a day trader.

John Kay had a good article in the FT at the weekend about the new European regulations.

Risk, the retail investor and disastrous new rules


The measure of risk used in the Kid is essentially the historic volatility of weekly returns. Or to put it another way — if the fund managers had been stealing your money, an investment with them would be described as “low risk” so long as they were stealing it at a more or less constant rate.


But this model does not represent what ordinary people mean by risk. According to the Oxford English Dictionary, risk is “exposure to danger, the possibility that something unpleasant or unwelcome will happen”. This is very different from the financial economist’s equation of risk with volatility.



No one talks about the “risk” of winning the National Lottery, or the “risk” that their favoured horse will be first past the post. They don’t even talk about the risk of not winning the National Lottery, because sensible people don’t really expect to win. The modeller’s approach confuses certainty and security. The financial economist who knows he is going to be hanged tomorrow has certainty, but not security. He knows that something unpleasant and unwelcome will happen and it is more, not less, unpleasant and unwelcome because it is certain. For the intelligent investor, the unpleasantness or unwelcome outcome that they fear is that their investment strategy fails to meet their reasonable expectations. And that is the relevant concept of investment risk.

 

[jpd]

You can read the article on John Kay's website - https://www.johnkay.com/2018/01/22/risk-retail-investor-disastrous-new-rules/

This doesn't need a subscription.

 

[Duke of Marmalade]

The PRIIPs regulations are certainly a nonsense when it comes to performance illustrations, as is covered elsewhere in AAM and is addressed by Kay himself; the company of whom he is a director illustrating no less than 22% p.a. future after charges performance in a moderate (i.e. best estimate) scenario

I am not so sure of the emphasis he puts on the risk rating. He cheapens his case by the silly example of steady stealing. It is quite clear that the risk indicator is covering investment risk and credit risk. The risk of fraud is clearly excluded as it should be. The following is the wording of a sample KID.

The summary risk indicator is a guide to the level of risk of this product compared to other products. It shows how likely it is that the product will lose money because of movements in the markets or because we are not able to pay you.

The fund is rated a 2 which FF describe as a low risk class (2nd lowest). I am not sure a fund which has lost over 5.44% p.a. over the last 10 years is adequately described as being low risk but that is a side matter.

The risk indicator is in the context of a 7 year investment horizon. Historic short term volatility is not too unreasonable a basis for describing the "risk" as defined above over 7 years. It should be noted that pensions are not in scope.

But Kay is talking about a different definition of risk - the risk of not meeting your investment goals. But it is unreasonable to think there could be a regulatory formula for this hugely personalised assessment. Think of pharmaceuticals. Lots of bumph in the leaflet describing the risks. But the only risk the punter is interested in is will it sort her ailment? That is for her doctor to advise.

There is after all a role for the financial adviser.

 

[Steven Barrett]

I always ask clients what is there understanding of investment risk. The most common answer is that they will lose all their money. While this is not impossible in the worst case scenario of the robots taking over or Trump and Kim Jong-un seeing who has the biggest button, if you have a diversified portfolio of quality assets, it is unlikely to happen. So what we would be looking at is how much your money can go up and down.

I hate the ESMA ratings system as it assumes that everyone has the same understanding of what a 2 or a 7 is. And because it is measured by volatility, a fund that has the potential for big losses could have a lower rating because the markets have been relatively calm over the last 9 years.

2008 is still fresh enough in people's minds to use as an example. Explain to clients how any particular fund performed in that year and use euros as well as percentages (% don't have a real value to most people).

This fund fell by 40% in 2008, so if I came to you a year after making an investment and told you that your €100,000 was now worth €60,000, would you sleep that night? Would it have an adverse effect on your lifestyle?



Steven
www.bluewaterfp.ie

 

[Brendan Burgess]

This fund fell by 40% in 2008, so if I came to you a year after making an investment and told you that your €100,000 was now worth €60,000, would you sleep that night? Would it have an adverse effect on your lifestyle?

That is a very good way of presenting risk.

The problem with volatility is that it does not measure the risk of cash. There is almost no volatility in cash. But it has been wiped out by inflation in the past and could be again. Banks could go bust and the depositors could lose a lot of their cash. This risk is not picked up by volatility measures.

I have a portfolio of shares. I understand that volatility is low at the moment. But I think that the risk is quite high of a substantial fall as shares are probably overvalued at the moment. The value of my holdings could fall by 40% by the end of the year. But everything else is risky as well.

Brendan

 

[Monksfield]

The risk indicator is in the context of a 7 year investment horizon.

No - based on five year data, as are the potential returns being shown. It is absolute nonsense: both need to be based on longer time horizons if indeed any potential future returns (or past ones for that matter) should be shown.

 

[Brendan Burgess]

Rory Gillen has just written a good article on this topic.

Are Regulators Misleading Investors?

When investment risk is understood in this way it becomes clearer that volatility in markets is rarely actual risk, but mostly just noise. As an example, Kerry Group's share price declined 41 per cent from peak to trough during the Global Financial Crisis (GFC). Yet, business risks at Kerry were no higher at the end of the GFC than at the beginning, the company was never overborrowed and its shares were not particularly expensive prior to the crisis. The volatility in Kerry's shares back then was not the sign of rising risk at the Kerry Group. It was just noise and a damn good investment opportunity if you had any cash!

Yet, and most disappointingly, regulators in Europe would have you believe that this volatility in companies (and funds) represents the real risk. And the European Securities and Markets Authority (ESMA) has actually gone so far as to insist that retail funds be rated for risk according to how volatile their share or unit prices have been over the previous five years. So, in the eyes of European regulators, Kerry at €13.20 a share in late-2008 was a riskier investment proposition than Kerry at €22.50 a share 18 months' earlier because of this volatility.

 

[Duke of Marmalade]

So, in the eyes of European regulators, Kerry at €13.20 a share in late-2008 was a riskier investment proposition than Kerry at €22.50 a share 18 months' earlier because of this volatility.

European regulators to not set the price, the market does. The fact that the same company with the same business profile was valued by the market at €13.20 in 2008 and €22.50 18 months' earlier says precisely that despite the apparent unchanged business profile the market clearly perceived that Kerry was a much riskier prospect. The fact that this has turned out with hindsight to be incorrect is irrelevant. Is Mr Gillen arguing that in 2008 he would have had no problem in identifying that the market was overstating Kerry's risks?

On the general point though, the regulators are desperately trying to come up with a universal risk rating system. If we accept (and I know some don't) that risk is in the following order highest to lowest: equities/property/bonds/cash maybe they should simply have given each asset class a risk rating and then mixed funds could work out their weighted average risk rating. Instead they have tried to cover all possibilities by using market volatility as a proxy to risk and moreover they are using the recent past to assess the volatility. I don' agree with that approach. To me equities are risk rated 7, end of. A benign period of low volatility does not reduce that rating to 5.

The real reason KIDs are being withdrawn is not this volatility issue which is relatively minor but the ludicrous requirement to regard the last 5 years' performance as the guide to the future.

 

[Sunny]

I don't think you can say that volatility is not a true measure of risk. It is but is not the only measure of risk. Like all statistics, it is how and why it is being used that is as important as the actual result. I agree with Duke. Using 5 years' performance as a guide to the future is the most ridiculous requirement. If any bank went to the regulator and said they were using a 5 year data set to drive their VAR modelling, they would be closed down.

 

[Rory T Gillen]

"To me equities are risk rated 7, end of. A benign period of low volatility does not reduce that rating to 5."

I'm not sure the above is a valid statement. History and common sense teaches us that all companies are not equally as risky. Berkshire Hathaway comes to mind, as does Kerry. With no increase in business risk at end 2008, less debt than in 2007 and the shares offering substantially better value at end 2008, it is hardly a stretch to say that Kerry contained less risk at end 2008 than in early 2007. In that regard, having all equities at Risk Category 7 makes little sense (to me anyhow).

The fact that investors were selling Kerry shares over that period also tells you nothing about the risks in Kerry. Most likely, investors were selling for broader asset allocation reasons or for liquidity.

As frustrating as it is, the risk in each company has to be assessed separately and using the concept of business, financial and investment risk assists one to categorise the major types of risk in any company. But even using these headings still makes it a difficult task and it is subjective at best.

 

[Duke of Marmalade]

"To me equities are risk rated 7, end of. A benign period of low volatility does not reduce that rating to 5."

I'm not sure the above is a valid statement. History and common sense teaches us that all companies are not equally as risky.

Of course that is true. I am thinking more in the context of the regulator trying to convey information to retail investors who mostly invest in collective vehicles covering whole classes of assets or mixtures of assets. So in that context equities as a class are at the end of the scale which has been set at 7. What I am criticising is that the risk rating is required to be updated by reference to the past 5 years' volatility. I find that silly. I believe volatility has been relatively benign over the last 5 years. Does that mean we should advise retail punters that equities are a safer bet now than they were 5 years ago?

 

[Rory T Gillen]

Of course that is true. I am thinking more in the context of the regulator trying to convey information to retail investors who mostly invest in collective vehicles covering whole classes of assets or mixtures of assets. So in that context equities as a class are at the end of the scale which has been set at 7. What I am criticising is that the risk rating is required to be updated by reference to the past 5 years' volatility. I find that silly. I believe volatility has been relatively benign over the last 5 years. Does that mean we should advise retail punters that equities are a safer bet now than they were 5 years ago?

I can better see the point you are making with that last post. When I look back at 2013, Irish commercial property funds with no debt had halved in value between 2007 and 2012. They were well diversified across prime Dublin office property in the main. Under the Regulators rules, they were riskier in 2013 than in 2007. I think we both agree that that is nonsense. As the value on offer in 2013 was much better than previously the risk (valuation risk) had reduced significantly, so these funds carried much less risk in 2013 than in 2007. Such funds doubled in value from 2013 to 2017. Same for equities. Value is a key component - one of three components - in assessing real risk. It stands to reason, then, that when value improves it lowers the risk compared to the alternatives, including compared to non-risk assets. The other side of the coin is return and risk/return is tied at the hip. Long-dated government bonds may be classed as non-risk assets, but today they offer no value and contain considerable risk against potential inflation. On a risk scale from 1 to 7, they've probably migrated to 4/5 in my book. In contrast, Irish commercial property in 2013 had probably migrated from a risk scale of 7 in 2007 to 2 in 2013. So, it appears to me that risk - even in an entire asset class - shifts around!


That'll be it from me on this one!

 

[joe sod]

Long-dated government bonds may be classed as non-risk assets, but today they offer no value and contain considerable risk against potential inflation. On a risk scale from 1 to 7, they've probably migrated to 4/5 in my book.

warren buffet made this exact point in his annual letter this year about bonds and why he does not invest in them. I know you said you would not comment anymore on this but what are the low risk assets today in your opinion?, I know you cannot go too deeply into this as you have a subscription advice service and it would be unfair to your subscribers