The Myth Of Tight Stops

TradingMarkets.com
The Myth=20 Of Tight Stops
Tuesday November 18, = 2:25 pm=20 ET
By Dave = Landry=20

Introduction

As a way to control risk, tight stops seem to be preached = universally.=20 This is either directly, by suggesting that you only risk a small=20 fixed-point amount per trade, or indirectly, through phrases like = "control=20 you losses," "cut your losers quickly," and so on and so forth. = The truth=20 is, in many cases, tight stops may actually increase = losses. From a=20 swing trader's perspective (2-7 day holding period), we will look = at why=20 tight stops often do not work, how to adjust stops to accommodate = a=20 market's normal fluctuations and how to compensate for this added = risk.=20 Finally, we'll also look at cases where tight stops can and = possibly=20 should be used.

Background

Years ago, I remember testing my first mechanical systems. When = I wrote=20 my first profitable one, I began to think, OK, I'll put in stops = and it'll=20 really make some money. To my surprise, the system actually began = to lose=20 money. Why? Because the positions were stopped out long before = they had=20 time to work. The normal noise of the market alone was enough to = hit these=20 stops before the predicted move ensued.

Noise

I recently received an email from a swing trader who had seven = losses=20 in a row. What frustrated him was that soon after he was stopped = out,=20 these positions went in the direction of his intended trade. Under = an=20 ideal situation, you enter your position, place a tight stop and = wait for=20 the market to move in the direction of your trade. The reality, as = this=20 gentleman painfully discovered, is that the market often moves in = the=20 direction of your trade but only after some chopping around--also = known as=20 noise. This noise is often enough to take out tight stops and is=20 illustrated in the figure below.

Calculating And Adjusting For Noise>

Obviously no one knows exactly where a stock will trade = in the=20 future. However, judging by past performance, you can get a pretty = good=20 idea where the stock will likely fluctuate over a given = period.=20 This can be measured with statistically based methods such as = historical=20 volatility and average true range. It can also be measured by a = simple=20 "eyeballing" of the chart.

Average True Range

As the name implies, the average true range considers the = true=20 range of a stock by accounting for gaps. Looking to Juniper = Networks=20 (NasdaqNM:JNPR = - News), a very = volatile=20 stock, we see that over a five-day period the stocks has an = average true=20 range of over 6 points (6.54). Longer-term (50-days) the stock has = an=20 average range of over 7 points (7.21). Therefore, if you intend on = swing=20 trading (holding two to seven days) a stock like Juniper which a = travel=20 range of at least 6-7 points per day, you're kidding yourself if = you think=20 you could use a tight stop (say 1-2 points) and not be stopped = out.

Historical Volatility

Historical Volatility (HV) is the standard deviation of = day-to-day=20 price change expressed as an annual percentage.(1) Whew! = Essentially, all=20 that means is that historical volatility is a measurement of how = much=20 prices fluctuate over time. Suppose a stock or commodity is = trading at=20 $100 and its historical volatility is 10%. At the end of the year = the=20 market will likely (statistically a 66% chance, assuming a = normal=20 distribution and that volatility remains a constant, see appendix = if you=20 are interested in the math) be trading somewhere between $90 ($100 = - 10%)=20 and $110 ($100 + 10%).

If you take the longer-term historical volatility reading and = reduce it=20 down to the intended holding period, it gives you a good idea of = where the=20 stock has the potential to trade (2). I have plotted this=20 calculation below on Juniper Networks based on the assumption that = you=20 intend on holding the position for five days. As of 04/30/01, this = gives=20 it a potential range of 46 1/2 to 71 1/2. This means in order to = help=20 ensure (but not guarantee!) that you won't be stopped out during = this=20 holding period, your stop would have to be at least 12 1/2 points = away.

Eyeball It

As a chart reader, a lot of my analysis is done by simply = looking at=20 the chart. A simple "eyeballing" of a stock that trades from 40 to = 50 one=20 day, down to 39 the next day and then back up to 50 is a volatile = stock.=20 Therefore, I know that I'm kidding myself if I think that I can = trade such=20 a stock and use a 1-2 point stop.

Don't Forget To Adjust For Risk

When trading more volatile stocks, you obviously can't just = loosen your=20 stop out without compensating for risk somehow. You must look at = total=20 risk and adjust your position size accordingly. Risking 1% to 2% = of your=20 trading equity per trade is a good rule of thumb. Therefore, in = general,=20 you can trade more shares of a less volatile stock such as a = consumer=20 non-durable or a utility that normally fluctuates 2 points = in a=20 week and maybe only 10-20 points in a year than a wild-and-crazy=20 technology stock that fluctuates 10-20 points in a day.

Avoid Anthills

Keep in mind that I am NOT suggesting you run out and use = looser stops=20 without a complete money management plan in place. Specifically, = you must=20 make sure that, on average, you are taking much more out of = positions than=20 you lose. For instance, you can't risk 5-points on each trade and = only=20 make, on average 1-point. Mark Boucher refers to such strategies = as=20 "Anthills." Ants can make a significant mound with little pieces = but all=20 it takes is one big footprint to knock it all down. Therefore, = make sure=20 your profits are large enough to justify the looser stop.

A Case For Tight Stops

Now that we have discussed the pitfalls of tight stops, there = are=20 instances where tight stops can and sometimes should be used. This = depends=20 upon the technical pattern and/or if you are willing to re-enter = positions=20 if your tight stop is taken out.

Pattern Based

In many cases, a stock should not break a technical = pattern. If=20 it does, it may suggest that the stock is failing. Therefore, in = these=20 cases, the stop can and likely should be placed below that = pattern=20 (for longs) even though this is often well within the normal noise = of the=20 market. An example of such patterns are low short-term = volatility=20 situations such as a high-tight bases. For instance, notice in the = figure=20 below that the hypothetical stock makes a narrow consolidation = after a=20 strong up move. The stock should not take out the bottom of this = base. If=20 it does, then it suggests that the stock may have topped out. = Therefore,=20 this provides a logical place to put your stop.

Keep in mind that in trading, everything isn't always "cut and = dry."=20 Sometimes, markets will make a false move below a consolidation = before=20 resuming their longer-term trend--a sort of "shaking of the tree" = or "head=20 fake", if you will. However, you must honor your stop because at = the time=20 it is occurring you have no way of knowing if it will turn out to = be a=20 false move (a) or a failure (b). This brings us to our next=20 point--re-entering.

Re-entering

If you are willing to carefully monitor positions and have the=20 discipline to re-enter if stopped out, then tighter stops can be = used,=20 even on volatile stocks. This is especially true in low = short-term=20 volatility situations such as the one described above. However, in = general, keep in mind that many small losses can and often do add = up to=20 much more than one large loss, especially when you factor in the=20 frictional cost (commissions and slippage) of all the additional = trades.=20 Also, keep in mind re-entry strategies are probably best suited = for those=20 with day trading experience who are used to taking numerous small = losses=20 and have the discipline to re-enter. If you can't watch a screen = or=20 psychologically aren't prepared to buy a stock that you just got = stopped=20 out of (sometimes over and over) then you're much better off using = a=20 somewhat looser stop and letting the position unfold as you think = it=20 should.

A Happy Medium?

The tighter the stop, the more likely you are to be stopped = out.=20 However, placing a stop where the statistics suggest is often too = far away=20 for most. Therefore, you can strike a happy medium and place your = stop=20 somewhere in-between. In this case, you know that your stop is = within the=20 normal noise of the market but it's less likely to get hit than a = very=20 tight stop.

In Closing

Tight stops, which many assume reduce the risk of trading, may = actually=20 increase risk as they are more likely to get hit due to the = normal=20 noise of the market. This reality must be factored into your = trading plan.=20 Above we have barely scratched the surface of the subjects of = volatility,=20 money management and technical patters. If you are to succeed as a = trader,=20 I strongly urge you to dig further.

Q&A

Q. How do you manage your protective stops? = Statistically?=20 Pattern based? A happy medium? Tight stops and re-enter?

A. I use all of the above. I mostly eyeball a stock to = get an=20 idea of where it has the potential to trade. I then look for a = logical=20 place that a stock should not violate. For longs, this can be the = bottom=20 of a pullback, the low of a setup, below recent lows, the bottom = of a=20 base, etc. In short-term volatility plays, I often use a tight = stop and=20 look to re-enter if it looks like a fake-out move.

Q. Is there a maximum percentage risk that you adhere = to,=20 regardless of what the statistics suggest?

A. As I wrote in my book, I normally don't like to risk = more=20 than 5% of a stock's value. However, there are cases were I might = trade=20 fewer shares and loosen that parameter.

Q. Such as?

A. If the overall market and sector conditions are = choppy, if=20 the stock is very volatile and/or if I know I won't be able to (or = simple=20 don't care to) watch a screen closely.

Q. OK, looser stops for an extremely volatile stock and = for when=20 you don't want to watch a screen makes sense. Can you explain = market and=20 sector conditions further?

A. Sure. If things are choppy then, I know that there is = a high=20 likelihood that stocks will fluctuate more before heading in the = intended=20 direction. On the other hand, if the market and sectors are really = in=20 gear, then I tend to use tighter stops.

Q. Define "in gear."

A. If the market and sectors have been trending, and/or = I have=20 market bias signals, then I might even use a tight stop.

Q. For instance?

A. Last week (late April), the overall stock market = began coming=20 out of a cup and handle. Biotech was on fire. In cases like these, = I use a=20 tight stop knowing that if I'm stopped out, I can either re-enter = or=20 probably find another stock in the sector that's moving.

Appendix

Calculating Historical Volatility

As with any indicator, I strongly urge you to purchase software = with=20 the indicators already programmed and would never suggest anyone = attempt=20 this by hand. With that said:

Let Length =3D length of volatility to be calculated and = ln=20 =3D natural logarithm

Historical Volatility(length) =3D standard = deviation(ln(close/yesterday's=20 close),length) * 100 * square root (256).

In English:

Divide today's close by yesterday's close.

Take the natural log of #1.

Take the standard deviation of #2 for length desired (the = number of=20 trading days, i.e., 50)

Multiply #3 by 100.

Multiply #4 by the square root of the number of trading days in = one=20 year (around 256).

HV is based on a one-standard deviation move

For those familiar with the bell curve and standard deviations, = statistically, this suggests that approximately 66% the market = should=20 trade within these ranges. If you double the HV, which gives you = two=20 standard deviations, then this gives you a statistical probability = that=20 95% of the time the market will trade within this range. A three = standard=20 deviation (3 * HV) would suggest that the market would trade = within that=20 range 99.7% of the time.

Keep in mind, however, that these calculations assume a normal=20 distribution and are based on present volatility (volatility is = constantly=20 changing). Natenberg said it best in Option Volatility and = Pricing=20 Strategies: "(You) shouldn't confuse unlikely with = impossible."

Using HV to set protective stops

In Connors on Advanced Trading Strategies, Larry Connors = shows=20 that historical volatility can be used to determine where stops = should be=20 placed. Calculation of these stop points are as follows:

Divide 260 trading days by the number of days you intend to = hold the=20 position.

Take the square root of #1.

Divide the historical volatility by #2.

Take the stock price and add (for shorts) and subtract (for = longs) #3=20 from it.

Footnotes 1,2. Connors On Advanced Trading Strategies, = Laurence=20 A. Connors. See the appendix for the formula for HV.

Additional Reading

Option Volatility and Pricing, Sheldon Natenberg

Dave Landry On Swing Trading, Dave Landry

Connors On Advanced Trading Strategies, Laurence A. Connors