How can you make a bet based on current information, when that information is turned on it’s head by a large outside force? The trends are reversed in moments, and you take horrible losses as they shift vastly in a matter of moments, you cannot cut your losses before they have gone. Stop losses are even worse as in this volatility they get taken out all over the place ensuring that you will take at least your maximum loss (with price movements like the ones we have seen will ensure that you will get filled at worse the price of your stop)
Since this intervention has been accompanied by huge volatility and volatility is the enemy of trend following. If you take volatility based position sizes it mean that you will only take tiny positions and will get whip-sawed in and out regularly racking up further (although reduced) losses. Add to this the fact that you will find it hard to find any good short candidates left, then you and your system are in a place that they have probably never been to before – the long only portfolio in a falling market.
You cannot ‘overcome’ this intervention, trend following or other system does not ‘overcome’ anything. A system should just do the right thing in the circumstances based on probability. Since this is a set of circumstances that has not come up very often, if ever, how can you know of your system can handle it? How can you know what is right as you need to understand the probabilities and we do not have enough information to determine it? The only thing you can do is endure it and hope the inevitable losses don’t take you out of the game or sit it out until the market returns to parameters that your system works in – cut the losses and ride the profit simply do not cut it in this market I am afraid.

Cutting losses and riding profit works in any and all markets, always has always will. Volatility is not the enemy of anything except those who haven’t properly planned or adjusted for it. There’s more than one way to cut a loss. Using stops is one way, taking fewer positions is another. Matt if you are getting stopped out of even your tiny positions then just take fewer positions. Just keep cutting, cut, cut, cut!

BTW, I am using the DJI as the case study.

Volatility is not the enemy of trend trading. If there’s any “enemy” its lack of trends, sideways markets. Look at the period from 1998-2003, we had massive volatility but we also had great trends. You’re saying volatility is sky high now but you are up 15% for the year, Im up for the year, most trend traders are up this year, but the market is down 20%. We are crushing the market even though volatility is high. How can that be the enemy of trend trading?

Reply to Ken (mainly!)

Yep I am up on the year and very happy about it, but at every instance that volatility has been high, trends have been less profitable to trade.

I have tested it many times over and the returns are noticeably smaller when volatility goes up. The reasons for this seem to be two-fold.

The first is a matter of mechanics. If you use volatility based position sizing (which is pretty common amongst trend followers as mentioned by the good Mr Covel in Turtle Traders) position size must be reduced per position.

As a simple example using nice round figures, lets say you have a $100 000 folio and your model uses a 1% of equity at risk (or $1000). Now whatever model you use to determine volatility will give you an idea of how much the price moves around over a given time (eg 3 times the 20 day ATR or Average True Range)

So let’s say that under ‘normal’ circumstances the volatility is 10% so if we are prepared to risk $1000 and volatility is at 10% then a position can be up to $10 000 (10% of $10 000 is $1000, the amount we are prepared to have at risk) Now lets say that the volatility doubles to 20%. We still only want to risk our 1% of equity, so we can now only take up a $5000 position (as $1000 is 20% of $5000)

I use something similar for sectors too – for example I will not let my exposure to financials go over a certain amount of equity at risk. This means that if a sector becomes more volatile I will reduce my exposure to it (I tend not to close positions to make this absolutely exact, rather I let positions end naturally when the signals dictate and simply do not replace them. That is unless things get very out of whack, which happened recently with financials!)

If this occurs then you find that your overall exposure is lower to sectors that are more volatile. If all sectors are volatile, then you can find yourself less than fully invested. Naturally you can decide to put more at risk, but since the whole idea of being a systematic trader is to follow the system…..

The second issue is that you have to visualise trends as similar to a TV or radio signal. The more noise you add to a radio signal, the harder it gets to understand what is being said. Volatility has the same effect as noise to a radio signal – degrading the signal, making it harder to ‘see’ the trend.

What this actually means is that during higher volatility periods you typically see the accuracy of signals drop. For example, if your system is typically 60% right and 40% wrong, you will expect to see that rate decrease with volatility, lets say to 50% right, 50% wrong. As you can see this is going to have an adverse effect on your returns.

I have also modelled how these rates change with volatility and found that losers lose more on average and winners win less as volatility increases.

So as volatility increases not only does it mean that I will be right less often, it also means that I will lose more on the wrong ones and win less on the good ones, in other words ‘ugh!’

On top of this you need to have enough data to ensure that your results are statistically meaningful. For most circumstances I have enough data to do this but quite simply I do not have enough data to give meaningful results when volatility is as high as it has been recently. However the evidence I do have is that I am getting close to merely breaking even (as the win rate drops and winners do not win as much and losers lose more….I think I might have already mentioned this!!)

Since I can achieve that with near zero risk by simply keeping money out of the market, and I am not sure of my data, I have been doing just that until I get back to a set of parameters that I know work.

Add to this the additional risk from the unknown effects of having the SEC remove the ability to short an entire market sector, the effects of the current intervention and the possible effects of any further intervention and I think that this is the only prudent action.

However if you are running a hedge fund that has to be invested by a certain percentage and with very large positions that can’t be liquidated fast, then you are going to see some nasty draw downs because of this….

Hi Ken,

Having read your post, I think we are talking slightly at cross purposes here. I totally agree with what you say about volatility in the direction of the trend, it’s a nice bit of juice!

Where I think we are at cross purposes is where you mention the period 98-03. You are right, this period did have higher than average volatility. However, the volatility levels I am talking about only occurred during the Asian crisis in 98 and the bottom of the bear market in 02 during that time. In both the 98 and 02 spikes, the performance of the systems I tested dropped significantly (in fact to negative) which is why I dislike the current situation which has levels similarly high!

As we know trend following is all about having a system which gives you a statistical edge. I have gone through and tested pretty much all the common versions of trend following (and before you say it, yes I am very dull!!) and every single one of them has performed worse as volatility increases beyond certain levels, as I mentioned. You either hit your stops more often or get more phantom exit signals if you exit through those instead (MA crossovers, Darvas boxes etc) Add to this your reduced position sizes and you can quickly turn a positive edge into a negative edge or at least one that does not merit trading.

Over the long run trend following will give you decent returns as we both know. My point is that if you test your system under various conditions you will be able to find where it works well and where it tends to fall down. If you do this you can then use this information to amend your position sizes, or even determine whether your edge at that time is even likely to be worthwhile trading.

You seem to take what I say as being that trend following does not work with a little volatility. My argument is simply that performance tails off badly when volatility goes beyond certain ranges. This is quite easily demonstrated by the guy mentioned in this post, my own trading performance and I am pretty sure yours over the last month or two, where the volatility has been exceptionally high.

I should point out that these periods of high volatility are rare (luckily!) and so it is very hard to get enough data to be able to draw significant results from them, but this is what I have found to be true and seems to be borne out by the current happenings.

Add to this the additional unknowns like potential bailouts, the banning of short selling for financials and the effects this will have on the trends we are trading, at least in the short term. With this and the evidence of falling performance with those high volatility levels and you have to admit it might be worth considering sitting out until things calm down and get back to ‘normal’ conditions?

Oh and lastly, the numbers you refer to in your last post were purely there to illustrate what I meant rather than be anything exact, in the same way as the 1% of a $100k portfolio example.

One other observation ——- In my limited knowledge of trend following, isn’t the discussion of the current volatilty in the stock market a mute point? In other words, isn’t the idea to trade a diversity of markets so that there is more oppotunity to catch a trend?

Matt if you would just replace the words “high volatility” with “high volatility in the direction opposite of the trend” everywhere in your postings then I would mostly agree with you, haha. My research has shown that you can make more money overall in high volatility periods. The more the market moves the more money making opportunities there are. And yeah you get stopped out more and its a rough ride, but overall you make more. I mean think about it, if you want to make 100-200% gain on a stock then you’re going to have to ride out some 30-40% drawdowns along the way. The drawdowns are high volatility, but the gains are even higher volatility so overall volatility works in your favor.

And another point, I don’t think the volatility we’ve seen recently is that unusual. Its pretty much the same as we saw back in march and january.

Im not worried at all, you’re the one getting slapped around in the market not me. I never said I had 40% drawdowns, I was just giving an example of how risk and reward go hand in hand. You say my view of historical volatility only goes back only to the beginning of the year, huh? Didnt I refer to 1998-2003 period when talking about high volatility?? Pull up a 10yr chart of the market. How can you say historically this volatility is unprecendented? Have you done any system testing more than 5yrs back?

The bottom line is this: Trends exist in volatile markets, trends exist in non-volatile markets. Do not be afraid of volatility, simply adjust to it. And use good money management techniques to ride through the rough periods of losing streaks. In the long run, volatility is good. Volatility is just market movement. If markets didn’t move we wouldn’t make money!

And one more thing, just because one stock drops 40% doesn’t mean someone’s portfolio is down 40%. Actually it doesn’t mean anything, if someone is using a 50% retracement stop then they wont even get stopped out on a 40% drawdown. And if they do get stopped out and are using a 1% equity stop then they only lose 1% of equity. You could have a string of 10 stocks that drop 40% and still only lose 10% of your equity. So again, the key is to adjust for volatility.

OK last go banging my head against a brick wall….

Firstly, the draw downs are exactly what this article is about, they were caused by the intervention. If government intervention caused traders to make money do you think they would stop trading saying the markets are corrupt?

As for unprecedented volatility, VIX has recently hit an all time high by some margin….but for you this still seems normal?

My data goes back to the 1980s and yes I have tested it back to then through the 1987 crash etc. and not many periods have volatility this high I can assure you. There are periods where certain groups of stocks have got very high volatility but not this widespread.

From your post it sounded like you were implying that your portfolio got 30-40% draw downs chasing those 100%+ individual gains. Thinking about it though, if you have a hit rate of around 30% (as you stated above) and a stop of around 50% on each that loses then it still sounds about right…

On the position sizing point, I already mentioned I use volatility based positions sizes too, don’t know too many folk that don’t with any kind of trend following system.

This also demonstrates why one reason why volatility does have negative effects on returns.

As an example, assuming that as volatility increases you reduce your position sizes. Now let us assume we have a 100k portfolio. Also lets say we were taking 10k positions under normal circumstances and when volatility is high you take 5k positions (numbers are here purely to illustrate a point, not supposed to be what you would actually trade before you pull that one up again)

Lets also assume that on average you make 100% on a trade (yes it’s illustrative too…) So when it’s volatile then you only make 5k as opposed to 10k. This means 5% rather than 10% on your overall portfolio on an average trade.

So to make the same overall return on your portfolio you would need to take twice as many positions in this instance.

As we know trends do not turn up to order, you are often sitting around waiting for them (there is even that bit in the Turtle Trader book where they talk about how much table tennis they played while waiting!)

If you are unable to find any other trends to put your money in, then you are going to be making smaller overall returns in volatile markets as you will often not be fully invested.

There are other reasons why you also lose more on your losers and gain less on your winners in volatile markets too, but it would take far too long to explain and I doubt you would read it anyway

(if anyone wants me to post this just let me know)

As for volatility being market movement – it is in the same way as giving your granny speed gets her moving; she gets very excited and unpredictable but doesn’t go too far!!

Now if you put your granny at the top of a hill and pushed her wheelchair down it, then that’s a nice downward trend…

Well if you correctly comprehended what I said and what Michael said then you would save yourself alot of headbanging (and typing!). Michael says very clearly “The issue is one of trust”. Thats the issue! Its not drawdowns! The old pro trader mentioned drawdowns, not Michael.

You say the drawdowns were caused by the intervention. Were they really? You are not responsible for your drawdowns? The drawndowns had NOTHING to do with your system?

Volatility Schmolatility. Look at the currency markets. If you need liquidity on large amounts of money, trade FOREX, Trend following works on currencies very well, I’m riding the USD/CAD, now, and it is up 435 PIPs. My trail is 80 PIPs, so my profit is locked in.

If you don’t have the software to automate it like I don’t, create a spreadsheet of the last 20 day’s data on a few (it is a lot of work) currency pairs that are not correlated and update it every day to see short or long signals. A macro can change the running ATRs, etc daily.