30 Years After The Crash - How Low Volatility Shares Can Offer Protection From Turmoil

25th Oct '17 by Ben Hobson

17 comments 13207 reads

Thirty years ago this month Black Monday saw stock markets crash around the world. After several months of dramatic gains, the collapse unfolded on October 19, 1987, sweeping through Asia, Europe and the United States. By the end of that month, the FTSE 100 had fallen by just over 26 percent.

For some, the current market conditions are reminiscent - and perhaps a timely reminder - of how stomach-churning market turmoil can be. So it’s worth having a think about what action can be taken to prepare for periods of volatility, and even a crash.

Drivers of the ‘87 crash

While ‘87 was before my time (as far as watching markets goes), it’s clear that prices had been on a blistering run that year. It’s also certain that the crash caused a huge amount of justified fear. Over long enough periods, price charts will show events like these as mere bumps in the road - yet it was much more serious than that at the time.

Chart: Trading Economics

But what this chart does make clear is that the ‘87 crash wasn’t a prelude to a bear market. Rather than falling further, prices generally recovered within a couple of years.

As an aside, a major feature of the recovery was that the Bank of England cut interest rates (from 9.8 to 8.3 percent between August and December that year) to reinvigorate the market. The backdrop now is very different. Rates of sub- 0.5 percent since 2009 have been a massive boost to equities. But the Bank wouldn’t be able to help in the same way if the market crashed now - although few are predicting that will happen.

Stretched valuations cause concern

There’s nothing quite like a major anniversary as a chance to draw (sometimes spurious) comparisons. But thirty years on, there are at least a few echoes of ‘87 in the market right now.

For the most part it’s because valuations appear high - and stretched in places. The US market has been unstoppable for eight years, and we’ve seen strong gains in the UK too (although nowhere near as extreme as in ‘87). In parts, expectations are very high, which means that stocks - especially small caps - are being crushed for missing earnings forecasts. That’s exactly what we’ve seen recently in shares like Low & Bonar,…

Unlock with Email

Unlock with Google

Already have an account?
Login here

brucepackard

25th Oct '17

Great article. I once explained this low volatility anomaly to a professor from Warwick Business School. His response was interesting: " That can't possibly be true. It contradicts Efficient Markets Theory. You can't get higher returns for less risk."
I'm sure if it was 500 years ago, he would also have told Galileo that his observations couldn't possibly be true, because it contradicted the Catholic Church.

Rusty2

Hello Ben, the problem with most of those shares, you listed above, is they took a hammering in the last recession, many went down over 50%, I would be more inclined to invest in companies such as Unilever (LON:ULVR), Associated British Foods (LON:ABF), Relx (LON:REL), Imperial Brands (LON:IMB), GlaxoSmithKline (LON:GSK), Diageo (LON:DGE), Dignity (LON:DTY), Tesco (LON:TSCO), Jardine Lloyd Thompson (LON:JLT), Legal & General (LON:LGEN), Royal Dutch Shell (LON:RDSB), National Grid (LON:NG.), H & T (LON:HAT). Many of these pay a decent dividend also. Of course it depends what causes the crash. If it's a UK only crash then investing in companies that have earning overseas will offer the most protection, we saw that when the Brexit vote was announced.

mgallear

Just had to sell a stock that I thought was trending very well upwards before I went long in but then had results much worse than expected and nosed dived on me. Although I am feeling a bit humble, the bottom line is that if you reduce volatility you are just inhibiting the possibility of how high your shares can go up. Maybe a balance of risk in your shares is a good idea.

In a Black Friday situation everything will go down and you have to be in cash before it happens to avoid the damage. (How you do that is of course another matter.) Even the shares that fall the least will still have you wincing in great pain and I am not sure the risk/volatility system will pick them.

Crashes occur when the market thinks the economy is going into a recession. If there are enough individual bad results then that will happen.

Graham Codd

Hello, I agree with Herbie47 and prefer to invest in companies returning a yield of over 4.5.%. These are more commonly found among the builders, BDEV, PSN ,TW all of which have also shown decent growth. There are though others including some of the Insurers and banks also AZN which could be bought for 40.07 in Dec 16, and this offers a yield of 5.4% and growth 10.1% to date. In my experience smaller companies rarely pay a decent dividend, and while there are profits to be made, they are far too volatile. For every one that has made me money, 2 have left me out of pocket and worse still many have been liquidated. The only way I would invest in them is via a unit trust and there I have usually have made decent gains.

Thanks but to clarify I'm not just advocating buying high yielding shares. I agree that in a recession smaller companies will tend to suffer more as they generally have less reserves that large caps. Low
volatility is only part of it, think you need to look at sectors as well, adn invest in companies that will still be in demand in a recession, such as food producers, supermarkets, utilities, insurance, drink and tobacco seem fairly unaffected by recession. If you take Costain (LON:COST) this company has very low margins, in a recession it will probably be hit as contracts are often cancelled or postponed, in 2007 shares peaked at 516p in December 2008 shares were down to 183p. Construction sector is one to avoid in a recession, next recession I will be considering shorting some companies.

Re small companies I have made good profits on some but that's in the last few years which has been a bull market. A few I will name to give you an idea Ab Dynamics (LON:ABDP), Bioventix (LON:BVXP), Tristel (LON:TSTL), Boohoo.Com (LON:BOO), Burford Capital (LON:BUR) (no longer small), so I would not write off small companies but you need to be aware of the risks. I'm going more towards large caps with high dividends and also buying some European shares.

Nick Ray

The low volatility "anomaly" doesn't work by providing you with some protection against market downturns. It is much more subtle than that and in fact it is probably not an "anomaly" (in the sense of being a market inefficiency) at all.

First, back to basics. The reason that there is an assumed linear relationship between excess mean return and volatility (the CAPM) at all is a consequence of Modern Portfolio Theory. As you combine stocks into a portfolio the volatility reduces more than linearly whereas the mean returns combine linearly. The result is that you get a curve called the "efficient frontier" and there is an optimal portfolio called the tangent portfolio which has the best trade-off between return and volatility (or risk/reward as it is called colloquially.)

The tangent portfolio appears to correspond with the market-weighted portfolio and the conclusion is that you cannot outperform the market in the long-term and you should use a mix of cash, stocks and leverage to adjust your return/volatility profile. When you juggle the maths and turn it on its head you find that stocks cannot stray too far from the return/volatility ratio of the market which results in a linear relationship between a stock's return and the "non-diversifiable risk" of the market (non-diversifiable = you can't get rid of it) plus an error term (alpha.)

However there are some problems. Firstly the CAPM prediction is not seen in practice. Stocks do not lie on the security market line. Secondly, the efficient frontier seems to continue to rise forever to the right indicating that as you increase volatility you can find stocks with higher rates of return indefinitely. Hence the idea of a "low volatility anomaly" - that a portfolio of lower volatility stocks performs better than the model predicts. But note: there is no implication that low volatility stocks somehow have lower volatility than the overall market. That is not possible and it does not happen.

The best explanation I have seen for the effect though is that the basic MPT model does not correctly model compounding in the portfolio. When you adjust for this the excess return acquires a new term of -0.5*volatility^2. So now when we plot u-0.5*s^2 against s instead of u against s the efficient frontier will droop as you move to the right (e.g. see here), so there is a now an optimum volatility beyond which you will get diminishing benefits. But this is no longer an "anomaly". It is an effect predicted by the model. This then tells you how to form a well-performing portfolio but it does not offer any special protection against high market volatility or market crashes.

Richard Goodwin

Housebuilders are historically amongst the most cyclical stocks in a housing market bust or general recession as they have to build on expensively purchased land bought in previous years but get lower sales prices. They are definitely not recession or share price drop proof and often have to have rescue rights issues at the depths of the share price trough.

That's not to say that it won't be different this time, the stock ranks are in their favour at least.

The best way to understand the Low Volatility Anomaly is to read Pim van Vliet's excellent book (which is pretty much non-technical) https://www.paradoxinvesting.com/
My big concern is that these types of stocks (many of those listed by @Herbie47) have been become so fashionable that this time may be different and they may underperform until they become less trendy. I have no empirical evidence to back that up though!

Alanjames999

26th Oct '17

We have to take so much risk to try and get 5-10% a year return in this environment, where as in the mid 80s you could of had the money in the bank and return almost 10% with low risk. Really makes you think..

Ken Lowes

You are not factoring in inflation!! In the eighties interest rates were high but in the main still below the real value of money. The crash in 87 was really scary as it locked out sellers and the market fell dramatically in one day followed by a substantial rise in the morning only to collapse in the afternoon and again the following day--then it was all over- with lots of huge losses and shares with high PE's turned into junk. Don't for one minute think things are different today, the occurrence may develop in a different way but the result will be the same, successful, but foolish traders, turned into bankrupts.
Ken

27th Oct '17

I appreciate Ben Hobson efforts to try to use new financial investment theory and put into a stock screen. I was expecting these shares to be fairly flat but that is not the case in Ben’s screen. I am far from expert on technical analysis or charting so welcome any insights if what I am saying is wrong to them.

I have just had a look at Ben’s Stock Rank Low Volatility screen. First thing I have to say is that with a Stock Rank greater than 80 and Altman Z score removing a large number of potential bankruptcies in the future that are strong fundamental reasons why this screen or sample of stocks should do better than the general index.

Beta has been criticised by much cleverer people than me in past as a measure of risk as it makes shares moving up just as risky as shares moving down

The balanced and conservative risk ratings apparently use a stock’s three year standard deviation of price movement from the mean. Is that a long enough period or is to long after all I am not interested in what share did in a stock market crash three years ago but what it will do over the next 6 months to a year.

I looked at the long term graphs of the shares on their Stockopedia front page that come out of this screen. Then I like the look of the Greg’s screen, it has had a fairly smooth advance in a 45 degree line upwards. This is my idea of a trending share with low volatility. Unfortunately, there are not that many share in the screen that have charts like this.

What about Qinetiq? It had a steady climb then a fall which I suppose a stop loss what have kept most of the gains. More recently it had a choppy flight path down then a sudden rise and appears to be climbing up again or is it? Bloomsbury Publishing recent past has consisted of a choppy sideways move followed by a sudden jump up. Bitter experience has taught me that the next most likely thing to happen is more sideways movement followed by a fall.

My problem with low volatility anomaly is that the maths used to determine this - mean that it is not identifying shares that are moving up and the low volatility shares identified can be moving sideways for long period, that are followed by a sudden jump up, but this is likely to be followed by more sideways price movements or a decline.

I am puzzled as to how you can use this phenomenon to pick individual shares or even why the shares as a group did as well in the study as they did.

(Just had a go at trying to devise my own composite technical analysis ranking model to find trending shares. The first results are that it is statistically very good at finding shares that will nose dive in the next month but a few have trended up!)

I've posted this screen before, but if you want to screen for smooth "45-degree" trending stocks try using a Sharpe-Ratio based screen like this one.

HI Nick had a look at your screen it does indeed find 45 degree trending stocks and I am impressed with it but again the maths is such that some stocks in the screen are not in my view trending up at all, so you could not just buy them in all and believe you had portfolio of trending stocks.

30th Oct '17

I created the diagram below to clear up two misconceptions about volatility:

That low volatility on its own somehow means that the stock will perform well
That volatility is not a good measure of risk because it includes the upside as well as the downside.

The diagram above ("right click; view image" to see a bigger version) shows graphically the measured mean return and volatility for two stocks ( green: Intertek (LON:ITRK) and red: Plus500 (LON:PLUS) ) chosen to illustrate the points. The historical mean return occurs at the point where the curve has its highest point and the volatility controls how spread out the curve is. (Because the area under the two curves is exactly the same, when the volatility is larger the peak itself is less high.)

In essence we are using the historical mean return and historical volatility to try to estimate how the stock might perform in the future. The graph can be used to compute the probability of the various outcomes. The most likely outcome is the mean return, but the spread of possible outcomes is much wider if the volatility is higher.

To estimate a probability you have to measure the area under the curve. So for example, to estimate the probability that the return will be less than zero (i.e. we make a loss) we can look at the area of the curve to the left of the zero line (shown in grey).

So for the two stocks chosen over this particular time period, although Plus500 (LON:PLUS) has a higher mean return (and hence potentially a higher payout), the greater volatility means that the probability that Plus500 (LON:PLUS) could in fact return a loss over the next year is actually greater than for Intertek (LON:ITRK) (because the area under the red curve beyond the zero line is greater than the area under the green curve). You have to decide what you care about most to make a choice between the two stocks: bigger average return or lower risk of drawdown.

And thus we get the two conclusions we were looking for:

The ability to perform probability calculations of the downside risk does not require that volatility itself is somehow only measured as downside performance.
You need a decent positive mean return as well as low volatility for good outcomes. If the mean return (peak of the curve) is not greater than zero for example, the probability of a positive outcome (given by area under the curve to the left of the zero line) will be very low regardless of whether the volatility is low or high.

Caveats: Historical measures of mean return and volatility are "estimators" for future return and volatility. In that role they are not particularly accurate! In addition the probability distribution may be skewed or have fatter tails than in the diagram above. (I have in fact used fatter tails than the Gaussian distribution for the plots above, but they might be even fatter than that!) Also, I am in no way recommending Intertek (LON:ITRK) or Plus500 (LON:PLUS) here. There are in fact many stocks with better historical performance. These two were chosen purely to illustrate the point as clearly as possible. Finally, this is just the tip of the iceberg. Once you understand the concepts of mean return and volatility, you are ready to leap into efficient frontiers and tangent portfolios, and so on.

GrahamF

31st Oct '17

Thanks Nick. A nice reminder of that probability and statistics stuff we learnt at school and had mostly forgotten and in particular how it applies to share prices. I suppose that this means we need to put on the wish list for Stocko to give us skewedness figures for each stock's distribution now to go along with the volatility categories.

Nick Ray has done a well thought out piece on the different risks between two shares. I don’t think you need to do the maths to understand why one is probably more volatile and likely to break its trend than the other.

Internek has moved steadily up since mid-2015 and before then it had gradually trended down. So there is not much reason to believe it will stop trending. In a way it is much less predictable than Plus 500 as there no way to work out when it will stop trending. (All shares will stop trending upwards at some point either to do with fundamental factors in themselves or market or due to a wider stock market upheaval.) In contrast, Plus 500 has moved up to peaks then fallen back a number of times – 2014, 2015, 2016 before climbing up again. It does seem to have broken this trend in 2017 moving up higher than before but the suspicion has to be that at some point the trend will change and it will fall back again.

(I thought technical analysis and charting was just a lot of gibberish and still largely do) until I read Stan Weinstein - Secrets For Profiting In Bull And Bear Markets.)

My suspicion is that it is only a few shares that have this low volatile and steady rise in price characteristic which is causing the low volatility anomaly.

Don't get too hung up on the two stocks chosen. I picked them because they cross over at a handy point!

The real use is to analyse all the stocks in your chosen universe and pick the 5-10 with the most desirable characteristics.

1st Nov '17

All good advice I will shut up and take a look at them. (Just watch yourself - consumer spending on the high street seems to suggest a recession is a possibility, no doubt caused by Brexit. The headings from Ben seem to suggest he is warning us as well.)

30 years after the crash - How low volatility shares can offer protection from turmoil

Drivers of the ‘87 crash

Stretched valuations cause concern

Unlock the rest of this article with a 14 day trial

About the Author

Ben Hobson

17 comments

Unlock the metrics for all these stocks and more...

Unlock the metrics for all these stocks and more...