Dear Don,

The eternal question of where a stop loss should be has been addressed by many contributions to the Worden Reports, and by the Wordens as well. As has been said before, stop losses need to be tied into the volatility of the stock, but how to do so is another problem.

For those traders who use a percentage stop loss, rather than a fixed stop loss placed on entry to the trade, I have found that the following method can help to keep me in a trade when there is a correction in the stock.

Obviously with any stock, the greater the volatility, the greater the gain (or loss) inherent in that stock provided you give the stock room to move. So the more volatile the stock is, the more risky it is.

Assume that you are prepared to lose up to 8% on a trade for a non-volatile stock. A volatile stock needs more elbow room, so assume a maximum of 12% for one of these. A bit of math below for those that are interested, followed by a formula that you can use as a Custom Indicator:

a=percentage stop loss for a non-volatile stock (8% in the example above)

b=maximum percentage stop loss for a volatile stock (12% in the example above)

v=volatility of the stock (which can be displayed in the left hand window on the standard TC2000 screen. This is provided as one of the fields available to users of the software).

To calculate the stop loss for any stock:

Percentage Stop loss for a given stock (S) = a + ( b - a) * v / 100, and then rounded to the nearest whole number

What this does is to calculate the stop loss for the stock as a value between the preferred stop loss (a) and the maximum stop loss that you are prepared to accept (b), dependent on the volatility.

To draw this on the price chart you will need to produce the following two Custom Indicators:

For long positions:

C * (1 - S / 100)

For short positions:

C * (1 + S / 100)

…where S is the calculated stop loss. These 2 indicators will then show you how the calculated stop losses have performed in the past. You may well find that in the current situation this value is too high, as the Volatility value is calculated over a 13 week period, so play with the indicator values to get a feel for an acceptable lower percentage stop loss.

Although this method is not the final answer to the problem of stop losses, it does at least give a reasonable method of tying together volatility and percentage stop losses, and making the exit from trades an impartial decision.

Regards,

Sir Ranter

Dear Don-

Imagine my pleasant surprise when Sir Ranter's letter on Stop Losses and Volatility appeared in The Worden Report this past Friday, February 20, 2003.

The previous day I had just finished drafting and sending my current "mission statement" to those members of my family whose trust assets I manage. At the heart of it is a focus on stops and their use in careful investment management. I am happy to share it with other TC2000 members if you feel it is appropriate. It is not the final answer to fiduciary investing. There is none. Markets and people do evolve while paradoxically they do remain ever the same.

I like Sir Ranter's well reasoned approach to the stop-loss problem and his mastering it into a measurement tool. Stops are always subjective and random. You yourself have tried to tame the beast by your inventive time-stops approach. I think my fellow knight's example that one has to pick his or her own preferred stop-loss level and then widen it to an upper limit based on the stock's volatility is a step in the right direction. But choosing both levels is subjectively arbitrary in his example. The market doesn't know or care about either of them. I think it is better to start with the stock itself and look at its specific volatility range, then base your stops on that. If that is outside your personal loss-tolerance limits, skip the stock and move on to another one until you find the stock that meets your comfort level. Anyway, here is how the business of stops fits into my current family investment mission statement.

INVESTING IN STOCKS

Background

Why do you buy stocks? To have them go up. There is no other reason. For income? No, bonds are for income. Stocks are for gains.

But the moment you buy, you have no idea whether the next price will be up or down, or next week's, or even next year's. It's a coin toss, 50-50, forever and always a coin toss. Are those really good odds to invest retirement money on? Most people would say no, but they do it anyway. I would say, absolutely not.

We cannot change the coin-toss odds. Can we develop some approach by which we can take advantage of stocks yet control their loss side?

That's the story of my life's passion. Here it is in a nutshell.

Fear, Greed, and Imitation

These are the major motivators of investors. They are mine, too. I don't like losses. When they come to me, and they are too frequent, or infrequent but too big, I become immobilized for a short or long period of time.

I am less prone to imitation but am as greedy, or fearful, or more so, as the next fellow. Thus, over years and decades, I have personally developed my own many trading systems, some good, some very good. That is, they make more money, some a lot more money than simply buying and holding or not investing at all.

Most of these systems occasionally score big or small losses. My goal has always been to build the best, that is, the most profitable, system over time. Within that goal I have been willing to tolerate losses. But when they occur in real time with real money, I am rankled and often immobilized as stated above.

Epiphany

I went to sleep one night last week pondering, as I have often done over the years, my most recent sets of trading systems and how to reduce or eliminate the loss problem. The goal has always been profits. The losses are something that is tolerated.

I awakened with a completely new thought that I should look outside the systems for my solution, not within them. I should have as a specific goal loss control. I should be thinking first how to control a trade so that the initial loss might be small if one occurs and that all losses be limited. If the trade shows a gain, then, and only then, add to the position. It is doing what you intended the stock to be doing. It is making gains. Reward it by getting some more. This approach would immensely reduce my fear factor upon entering a new trade.

Thus was born an improvement in my approaching the investment opportunity of investing in stocks. The strategy remains the same: use a profitable overall system. But now add tactics to control entry and management in a position.

Entry and Management of a Position

For example, I have say $10,000 to invest. Before, I would cross my fingers and invest it all. Now, I take only a quarter, a third, or even a fifth of that as my first commitment. And I set an automatic 'stop loss' limit of say 7%. Thus, instead of losing $700 if the stock goes against me initially, I lose only a quarter of that or $175. That's 1.75% of my total amount. I am comfortable with that and can try again later on.

If, instead, the stock rises 7%, it is showing me an initial gain--the reason I bought it in the first place. Now I move my stop-loss limit up to 7% below the new high price of the stock. If it drops from there, I lose nothing on this position. I break even.

At the new higher price I add an additional one-quarter of my total money which brings me now to one-half invested position. My newly raised 7% stop-loss applies to this position.

But if the stock drops from here to below my stop-loss, I lose only $175 of the total invested so far and can try again later on.

If the stock goes up 7% three times after first purchase, I am then fully invested (four times one-quarter each time). As the stock rises, if it does, I then trail my stop-loss order upward behind it so as not to let profits, if they occur, turn into losses.

I may widen the latitude of the stop depending on the price volatility of the stock. For example, if I start buying a stock at 70 and complete my four positions around 90 and the stock goes to 120, I may widen my stop to 15% to give it more room to swing without getting stopped out on a temporary 7% drop. And so on.

If the stock is 'stopped out' on first purchase, I must wait until it shows some evidence of turning upward. I will want to see the price rising once more before I start to buy a new initial position again.

A stop-loss order is placed the same day on every stock purchase and is moved upward as required, never down. Review is a daily process.

Sometimes the actual loss can be larger than the limit I set due to the mechanics of the market. The size of the stop-loss percentage used is a trade-off between a stock's characteristic historic price volatility and the magnitude of personal loss you are willing to tolerate.

Which Stocks?

My systems rank stocks periodically by the strength of their price behavior. That is, which stocks are newly rising the fastest, the greatest, or most steadily in price week after week? These are the best candidates for purchase--applying the entry and management tactics outlined above.

Of these, I want to use 'index' stocks (they're called ETFs, Exchange Traded Funds). These are single stocks, but they each contain dozens or thousands of companies. This causes them to be less volatile than the individual stocks which they hold in portfolio. They mirror an industry, a country, types of companies-like growth, or consumer goods, etc--or the broad stock market itself. They are like mutual funds which likewise hold many issues. But unlike mutual funds, they trade all day on the stock exchanges so that one can buy or sell anytime rather than just on the closing price for the day. This gives an investor much greater control over entry and exit. Their trading volumes are large which gives investors better price control when buying and selling. Their management fees are much, much lower than those of mutual funds.

Mutual funds or individual stocks may be used when conditions seem to warrant. Likewise options. But for the most part, straight index funds are the main investments of choice.

Reporting and Tracking

You observe everything I am doing by means of reports from BrownCo. You get confirmations when transactions occur. You can see the month-to-month changes in the value of your account when you receive their monthly Statements.

If you have questions on anything I am doing, please let me know. Although some of my research does attempt to forecast the market, none of my management is based on forecasts of either stocks or the market. To the extent possible, it is based on pure, rational mathematics to attempt to eliminate subjectivity of judgment which is so often influenced by the fear, greed, and imitation which I mentioned at the beginning above and has to controlled or abolished for success.

***

Don, you may notice in the above the absence of any reference to a stock's corporate fundamentals which many knights and TC2000 members include in their screening processes. I have nothing against fundamentals. You may recall I used them in one of my watch-list tabs in my very first letter to you back on May 25, 2001. I still do--occasionally but do not believer they are critical or essential to the 'mission.' ETFs and mutual funds don't have these data anyway.

I will write another report later on how to measure the unique volatility of each stock to meet one's own personal level of loss tolerance per trade.

All best wishes, and of course repeated thanks for creating and making available TC2000, without which none of the above, at least for me, would exist.

JM (Sir Copernicus)

Dear Don,

I enjoyed the note from Sir Copernicus (2/24/04) because his approach to investing was so close to the one that I am using. I agree with most of his comments and would like to add a feature that I have used to advantage.

There seems to be general agreement that good portfolio management requires the setting of stops to trigger the sale of an equity. However, when it comes to determining where the stop should be placed, there seems to be less than adequate information. This is especially so if a stock's variability is to be taken into account. From the far distant past it seems that a figure of "8% below purchase price" or "12% below purchase price" has become ingrained in investment lore, yet I have not seen any real evidence to support these apparently arbitrary values.

"Volatility" can be expressed in any number of ways depending on the question being addressed. When it comes to setting stops, what is important is the change in price that occurs from one day to the next. I use the term AVERAGE DAILY PRICE CHANGE or ADPC for this definition. If the stop is set too close, then the normal price change of the stock will trigger a sell signal. The ADPC adjusts for the daily price variability inherent in each stock.

Calculating the ADPC is simple. It requires the use of TC2000 and a spreadsheet program such as Excel. I strongly recommend that investors learn to use these two programs together since the results of the union are so powerful. For any stock TC2000 will provide historical price data going back a number of years. It is advisable to use at least 4 years of prices since you would want to cover up and down trends in the calculation. Using a large number of data points also minimizes the impact of the occasional price spike that occurs from time to time for any stock. The "Convert to Text File" feature of TC2000 will produce a file that transfers data to the spreadsheet. Only dates and closing prices are required in the file. Once this information is loaded onto the spreadsheet, it is a simple matter to calculate the PERCENT change for each successive date. Also, these percentages should be absolute values, that is, all values are shown as positive numbers. The average of these calculated percentages is the ADPC.

Examples of some values that I have determined: VSH 3.09%, ADPT 3.38%, AFCI 4.54%, EK 1.56%, GSF 2.77%, HPQ 2.47%, MERX 4.14%, NTIQ 4.14%, NOVL 3.17%, QCOM 3.48%, BLI 2.40%, VISX 3.34%, CTG 3.30%, FORG 4.54%, FEP 3.06%, IT 2.79%, GGI 2.00%, G 1.68%, IOM 3.05%, NR 3.20%, TGX 3.34%, TROY 3.79%, TRR 3.20%

To use the ADPC, it must be remembered that these are AVERAGES. An allowance must be made for the fact that it is not unusual for a stock's price to move in the same direction for one, two, three, or more days. Also, these values are averages. Individual daily price changes will fluctuate around this average value. The minimum individual daily price change is zero (0), but the maximum can be quite large. For example, EK is a stock with a low ADPC of 1.56% but 95% of the values fell between 0 and 4.3%. In the case of NTIQ where the ADPC is quite large at 4.14%, 95% of the individual daily price changes fell between 0 and 12.4%.

It is easily seen therefore that stops set by the arbitrary 8% and/or 12% drop in price rule will result in a lot of selling even though the decline in price was probably normal for the stock. For my investment protocol, I set stops at 6 to 7 times the ADPC and move the stop up as each new price high is registered.

This is the procedure I use while a stock is climbing in price but has not reached the point where it has established a significant profit. Once the significant profit point has been reached, I use a different method for setting stops that protects a major portion of the unrealized profit should the price pattern start a downward trend. This is a subject for another note.

-JL

Don,

One simple approach to tailoring a stop loss to the risk tolerance of the individual, as discussed in the 2/27 Notes, is to use a dollar stop.

The individual must decide how much money he is willing to lose if the trade goes against him, either in dollars or percent of his portfolio (which can then be converted to dollars).

This needs to be combined with a volatility approach (per Sir Day-to-Day, or using average daily range), or use of a nearby support point (my general preference), to set a logical stop price. The stop must be identified before the trade is initiated (always a good idea). The difference between the entry price and the stop price is then divided into the dollars placed at risk to determine the number of shares to trade.

As an example, if a trader wants to buy a stock at 35.00, and the logical stop is 33.00, and he is willing to risk $1000 on the trade, then he should buy

$1000 / (35.00 - 33.00)$/share = 500 shares

The dollar amount risked can be varied with market conditions or other factors. For example, when the market is overbought, but the stock and the industry group look very strong, it would be reasonable to reduce the amount at risk. On the other hand, if the market is oversold, and the stock and industry are both strong, the amount risked could be increased.

Regards,

Lanier Dodson

PS: The recent upgrades to TC2000 and TCNet are great. Many thanks to the folks that have implemented them.

Dear Don,

I have enjoyed reading the recent submissions on stop-loss. Over the years, I have experimented with various different methods to find the perfect stop-loss-- Average True Range (ATR), Parabolic Stop and Reverse (SAR), recent pivot lows/highs, trendlines, moving averages, 1-5 bar lows/highs, percentages of stock price, percentages of trading equity, just to name a few. At times, I found I could actually predict the direction of the market by noting where I had put my stop-losses.

The difficulty with stop-losses is that no one method fits all situations. Each trade is different. Each stock has a different personality, and a change in the direction of the overall market may ruin the best-planned trade. So for the newbies out there, perhaps frustrated with trying to figure it all out, I thought I might attempt to provide a logical framework for placing a stop-loss.

A stop-loss should be part of the overall trading plan. The plan should start with a "premise or a reason for the trade. Based on technical analysis, the premise represents what the trader expects to happen. For example, before entering a pullback in an uptrend, a premise would be: the pullback is not a trend reversal, and after the trade is entered, the trend is expected to resume its upward direction.

The stop-loss should then be set near the technical point where the premise becomes invalid. For example, if we decide to buy the aforementioned pullback in an uptrend, the stop-loss might be placed below the most recent short-term swing low. Why?

Trends are built on a series of higher lows and higher highs. If we suddenly get a lower low, the trend is no longer in effect. Our premise has been violated, so our stop-loss removes us from a trade that no longer makes technical sense.

Seen in this way, stop-losses become a more natural outflow of the thought process that goes into each trading plan rather than pure mechanical exercise. One can then make whatever adjustments are considered appropriate for a particular trade (volatility, risk/reward, percentages of trading capital, etc.).

Unfortunately, stop-losses also expose our flaws as traders. Based on my extensive experience in this area, here are some other points to remember:

1 The best stop-loss is one you'll use on all stocks, at all times. Any stop-loss is better than no stop loss. Be disciplined.

2 Determine your stop-loss before you enter the trade. This causes you to consider your entire plan while your mind is clear.

3 Your stop-loss method should be written. It is an integral part of your overall trading plan.

4 Move stops only in the direction of the trade. Never adjust your stop to prevent being stopped out. You're being stopped out for a reason.

5 A good stop-loss will not protect a bad entry. A stop-loss is only as good as your entry. If you get stopped out, analyze whether the trade or the stop-loss was the problem. Make adjustments.

6 Protecting profits is equally as important as protecting from losses. Some traders lose focus after they get into a successful trade and forget to raise their stops in a timely manner.

7 A triggered stop-loss is not always the end of the trade. You may find that you can re-assess the trade, re-position yourself, and make a better entry.

8 The market doesn't know about your premise. Trading against the broader market trends will stop you out whether your premise makes sense or not.

9 Smile when you get stopped out. You probably saved a lot of money.

10 Mental stop-losses are for mental traders. Re-read #1.

You can't control the market, but you can control your risk. The more I trade, the more I look for my stop-loss first; then I try to enter the trade as close to it as possible. It's better to take small financial losses instead of big mental losses.

Kindest Regards,

Charles (Squire Halloween)

Don,

I've followed the posts on setting stops with great interest, here's a system I use which uses Bollinger bands (which account for probability as well as volatility) which may be of use as it's so easy to set up in TC2000. -Nick

Dear Don,

Setting stop-loss targets is something I've been playing with for years. I've not come up with the perfect solution yet (and don't know anyone who has), but have found Bollinger Bands a useful tool to establish stops.

For those not familiar with Bollinger Bands (BB's) here's a quick and nasty overview:

BB's are bands, either side of the stock price calculated as + and - a number of standard deviations either side of the exponential moving average of the price.

Statistically then, the chance of the next day's EMA value being outside of the bands depends on the number of standard deviations applied to it.

The probabilities for movement outside of the BB's end points, for a given day, on the next day's EMA on close can be found in any book on statistics, for example:

The smaller the number of days in the EMA calculation, the closer these probabilities are to the chance of the closing price being outside the bands. TC2000 needs two parameters set to set up the BB's, the number of periods on which the EMA is determined and the width of the bands. Band width is expressed as standard deviations with 20 corresponding to +- 2.0 SD's.

Note that BB's work on closing price and the EMA of that. A stock may well trade outside the bands during the day but still not violate these probabilities so long as it closes inside the bands.

For those not particularly interested in the math, Bollinger Bands can be used to set stops as follows:

Set up a daily chart, Zoom 7, with the following parameters: Price - type Open/Bar Add Indicators: Bollinger Bands - Period 5, Width 10 Moving Average - Period 5 (to match the BB's period), Exponential (The EMA isn't really necessary for this exercise, but is a reference point which should fall midway between the bands.)

(This gives a running weekly, or day, average with an approximate 15% probability the next day's close will be outside the bands.)

Now try this out on some of your favorite stocks to see when it would have stopped you out of a long trade. If your tolerance to the probability of loss is lower than 15%, say 10 %, then reset band width to 13 and try that. You can leave the "Edit" window for the Bollinger band settings open, alongside the chart, and vary the parameters until you get something you are happy with.

Notice that as price volatility increases, that the BB's widen out, as volatility decreases, the bands tighten closer to the EMA.

The task is then, on a given day, to set the stop-loss to the lower band for the previous day's close. The chart has already calculated the approximate probability of the range of closing prices for you and accounted for the price volatility of the stock.

You can vary the period for the BB's to suit your trading horizon. Longer periods for longer horizons. But remember to re-set the EMA period to correspond.

Two notes of caution: The probabilities above are for the chance of the EMA moving outside the bands and are only an approximation on the price moving outside them.

Bollinger bands are useful as a trading tool as well, but not with these settings. I just use these to set stop-loss points.

Cheers,

Nick (Sir Aussie)

Dear Don,

I read your introduction to Sir Day-to-Day's submission regarding setting loss-cuts based upon average daily price change (See Worden Report dated 2/27/04). You asked to address the problem of adjusting loss-cuts for the "risk tolerance" of the individual trader or investor. After many hours of thought, my conclusion is that I would not adjust a loss-cut based upon risk tolerance. However, I would adjust my position size based upon the loss-cut. This would effectively relate one's risk tolerance with the loss-cut.

When setting loss cut levels on a particular stock, I believe that the market character of the stock should control. Thus, the two methods I use to set stops are support / resistance levels and volatility based multiple of a 10-day average true range. Whether one uses these methods or Sir Ranter's volatility based percentage methods (See Worden Report dated 2/20/04) or Sir Day-to-Day's average daily price change, is immaterial. The key is to dial in the appropriate loss-cut based upon some estimation of the "normal" fluctuations of the particular stock. That way, you will be "stopped out" only if something out of the ordinary happens, which would indicate a change of character and a reason to exit the position.

Therefore, I would question the usefulness of incorporating one's risk tolerance into setting a loss cut. Adjusting the loss-cut based upon one's risk tolerance takes the setting of the loss-cut point away from the stock's market based factors (i.e., volatility characteristics) and allows a person's individual personality to enter the equation. I don't think this would be advisable, because it interferes with this market based process.

Rather, I think the best way to account for one's risk tolerance is to modify the position size based upon where the loss-cut is set. For example, let's take a trader who has a risk tolerance of $1,000 (which is 2% of a hypothetical $50,000 account). Say he invests in a $20 stock with an average daily price change of $0.50. Using a multiple of 4 times the average daily price change, our trader sets a loss-cut at $2.00 below his entry point of $20. Since his risk tolerance is $1,000, he can buy up to 500 shares ($1000 / $2). If his risk tolerance is only $500 he can buy 250 shares. This way, the loss-cut calculations are market based and do not fluctuate based upon one's risk tolerance. As one's risk tolerance grows or shrinks, his position size would vary accordingly. I hope this helps. If it doesn't answer your question, it is a good answer to some question.

Thanks, David

Dear DW:

The recent terrorist massacre in Madrid prompts me to send you my meditations on the Stop Loss Levels submissions I've been reading in the "Report."

It is amazing the extreme sophistication some contributors on Stop Losses have achieved. I'm greatly impressed by the wide variety of techniques used to enhance their profits by minimizing losses: Bollinger Bands, Elliott Wave/Fibonacci Theory, Aussie Dollar Comparability, Sizing, Risk Tolerance Amounts, et al...It is just awesome. I'm happy it works well for each contributor.

You have profiled me as an architect, MBA, somewhat of a physicist. Yet I have traded all that for simplicity. Maybe experience has over-matured me, a nice way of saying 'burned out'. Personally, this is what I do. Before I say "BUY" I set a Stop Loss Level based on the chart pattern that is discernible to my eyes. I prefer it if it is no more than 10% down if target is 30% up. If Price rises, I trail (float) the Stop Loss to be 10% below new highest close. I've adopted DW's advice of not giving away all paper profits if market goes down. In such event I tighten the Stop Loss by using the suggested 66%, 50%, 33% of profit as Price rises over 10%, 20%, 30% over purchase price. If it is disciplined, it works nicely and it is simple. It may not yield me as much as those systems I've been reading and admiring; but I'm contented.

The key is to sell for profit in serenity and not out of fear. A fear from whatever catastrophe is unfolding. Panic can produce dire future consequences.

Such traumatic fear is what the terror massacre produced in Spain. Without objectively analyzing the enemy's logic, they feared worst things to come and had what we would call a Panic Sell. They caved in to the terrorist aim which was to separate Spain from the US led Coalition against global terrorism. Four days after the massacre were the general elections, and the incumbents were voted out in the 'hope' the Islamic terrorists will no longer attack them. Maybe they won't anymore, but they have set up a bad precedent: When under heavy attack, they weaken to the point of submission.

Predators can sense weakness. The Spanish domestic terrorists, the ETA are independentists who wish to split Spain into smaller entities as happened to Yugoslavia.

They know now that Spaniards weaken when terror is of dramatic proportions. Their future acts will no longer be "minor" but "meaningful" until a weak demoralized Spain accepts perpetual secessions and partitions.

The "moral" is Panic Selling is never good for Investors, Traders, or Voters. Logical Planning works best. An old investment adage is "In case of doubt, sell to a sleeping level", to which I add: "To a comfortable sleeping level, not to a nightmare!"

Best Regards,

Dear Don & Peter,

A lot of time is spent discussing entering stocks at the appropriate time and I've learned much. My biggest difficulty is determining when and how to depart a stock.

I've read and reread Worden notes on exits. However, maybe you could concentrate on some words of wisdom concerning exits.

More often than not it seems I set a stop that ends up being the low of a correction or if I don't set the stop, the stock continues lower. I hate to give up too much on a stop, but I'll state the obvious. If the stop is set let's say 10%-15% below the current price you might likely be stopped at the end of the correction or if the stop is placed too tight you can easily be stopped out on a volatile day (rumors or whatever) and get the low of the day.

Can you offer some of your ideas on this matter please?

Thanks,

Jack