I've been trying to develop a mechanical trading strategy for the portion of money I'm not using for a COT system. Just about everything I've tried to backtest has not worked out. I'd be very interested to hear from others about what trading systems you use if any if you're willing to tell us.
Howard

Howard,

For strictly long plays, there are a few with proven results. Henry Ford from Pitbull.com has a few I can personally recommend. The Earnings Trader and Pitbull Trader System. These systems are purely mechanical and are based on fundamental and technical indicators.
My opinion is that these strategies plus the COT and others learned on this forum is all I need.

Peter, I have seen the "reported results" of the pitbull system off of the pitbull site, as well as a few other sources. I am curious how well you have done by using the system either on paper or for real since the site says returns over 200% year after year. I ask, because I wonder if it would be worth checking out (I love strat 4 COT, and my silver stocks have done well but I am always looking for more ideas )

Thanks,
Heath

I really like these systems and unfortunately in the past few months not too many selections are available which in the past has shown a market direction change. Like all successful systems, if you follow the rules you can make money, but I must say I have not always followed the rules and sold too early. I am trying to become more disciplined. Always protect yourself with stops and this system can be profitable.

Hello Peter, I just purchased the Earnings trading system about one week ago. I have just started to read it.
How long have you had it.
Also do you watch Mr. Fords Crash index ?

Howard,

When you say that what you have backtested so far has not “worked,” if you would tell me what you’ve backtested and over what period you tested it and I can probably tell you why it didn’t work. I’ve backtested hundreds of systems and strategies.

Here is what should help you. Most profitable systems will fall into one of two categories – trend following systems and reversal systems. Trend following systems are reactionary and seek to capture a trend that is already in place. Reversal systems are anticipatory and seek to find the turning points in markets.

Examples of trend following studies and systems would be moving averages, MACD, Aroon, Wilder’s DMI, Donchian breakout systems, volatility breakout systems, CANSLIM, Pitbull Investor, Parabolic SAR, Dow Theory, Point and Figure.

Examples of reversal studies and systems would be Stochastics, RSI, Williams % R, Commodity Channel Index, sentiment strategies, 2B reversals, divergences.

There are some that seem to try to do both. I would probably put Bollinger Bands and Elliott Wave in that category.

When you say that what you have tested hasn’t worked, my guess is that you’ve found that it works some of the time but not all of the time. That’s because any trend follow strategy will only work in trending markets. And any reversal strategy will only work in sideways markets or markets that are at the end of trends. Alas, there is no Holy Grail.

Some people feel more comfortable with trend following strategies and some people prefer reversal strategies. So that’s one question you need to ask yourself – what kind of strategy most suits your temperament and style? Let’s say that you prefer tend following. There is always something that’s trending. So the way to make a trend following strategy work is to diversify into various markets. That way you are much more likely to catch a trend. And with proper money management, the markets that are trending should more than compensate for the losses you would be taking in markets that are not trending.

Van Tharp says that there are probably thousands of systems that you could come up with that have a positive expectancy. I believe that. The real reason that people don’t make money with systems is they don’t follow them. It’s a simple as that. Unfortunately, nobody has ever figured out a way to save people from themselves.

But here’s the best kept secret of all – you could enter a market by a flip of a coin and still make money if your money management skills are good enough. In Van Tharp’s book, Trade Your Way To Financial Freedom, money manager Tom Basso tested that proposition. He just flipped a coin to determine whether to go long or short. But by spreading his risk over several markets and concentrating on position sizing and risk management, he made money.

I don’t know if you’ve ever heard the story of Market Wizard Richard Dennis and his “Turtles.” But to make a long story short, back in the early 80’s Dennis had a bet with a fellow trader that he could take a group of ordinary people and turn them into successful traders. He literally put an ad in the Wall Street Journal looking for candidates. He chose some (he called them “Turtles”) and had them trade just a simple breakout tend following system that if a market made a 20 day high, they would buy. If it made a 20 day low, they would sell. That’s it. But because he trained them in money management skills, most of them were very successful.

Here’s another gem from Van Tharp’s book that I have always found to be very enlightening. One of the most successful Market Wizards of all is Ed Seykota, a trend follower. I’ve written about this before, but his system was to pin a chart on the wall and walk to the other side of the room and look at it. And if it was obvious to him that it was trending from the other side of the room, he would have no problem with trading that market in the direction of the trend. But, again, he put a lot of emphasis on position sizing and the old truth – cut your losses short and let your profits run.

I think in 2004, we should all repeat to ourselves over and over again until it’s second nature…

It’s not the system that is the key to my success. I am the key to my success!

Larry

Heath,

You mentioned that the Pitbull site cited returns of over “200% year after year.” The only serious market related dispute I ever had with the Pitbull folks is in the way they report annual returns in their manuals and in their systems. They have a very weird and misleading way of doing it.

For example, in the “Smart Money S&P Proxy Trader” – the manual that first got me interested in the COT strategy – they cite the annual returns for what we call Strategy No. 1 as 88.75%. A 2 to 1 leveraged approach to Strategy No. 1, they annualize at 789%. Of course, I ignored that part of the manual and computed the real annual returns. But I can see how a novice would except that as fact.

I could get into the details of how they compute annual returns, but the point is that it’s not accurate. It’s a shame. Their systems are good and they don’t need to do that.

Larry

Larry,
Thanks for your thoughtful post. I agree with you 100%. I don't think I'll have trouble following a system, I'm having trouble finding one that on backtesting proves to be positive in expectation. Frankly, after reading Van Tharp, I thought this part would be the easiest.
Here's why I agree with positive expectation and sticking to the rules. There is a psychological test called the IOWA Gambling Task. In it there are 4 decks of cards. The first 2 decks are the same and the second 2 decks are the same. The subject is told to turn over any card from any deck. He has 150 chances to pick cards. Each card has a money reward and a money loss. The first 2 decks are big wins and even bigger losses, but the losses are spread out so it isn't obvious. The second 2 decks have small wins and smaller losses. Basically the first 2 have a negative expectancy and the 2nd 2 have a positive expectancy. People are told to find the best cards that make the most money.
Obviously, the best strategy would be to pick 150 cards from decks 3 and 4. You would get about a 20% return. Nobody does that. But people can be separated by how many picks they take from decks 1 and 2 compared to 3 and 4. For instance methamphetamine addicts pick many more from the 1st 2 than non addicts. Not only that but while doing the test while in a PET scanner, it turns out that the part of the brain called the ventro medial prefrontal cortex is turned way down in the addicts. People who have frontal lobe damage from car accidents respond this way to. The VMPC is where we recognize changing rewards from changing environments and recalculated the most rewarding strategy based on risk/reward. Without it we use a more primative part of the brain to go for the biggest reward and damn the risk. So we'd look at day trading with our intact VMPC and figure we'd really have to know what we were doing and even then it would be too risky. But without our VMPC functioning well, we'd see only the possible rewards and go full steam ahead.
Frankly, I wouldn't trust my money to anyones, VMPC, even my own. Maybe least of all my own. So I have to figure that "discretionary trading" that requires and very good VMPC is not my bag. I'm looking for something as simple as: using stocks over a market value of $1B only buy when the 20 day MVA crosses over the 50 day moving average. Put stops at the last low and get out when the 20 day MVA is violated or when the SAR is hit. Position size so that no position is greater than 1% of trading captial.
I could then take those stocks and using stockcharts.com back test that through a lot of markets and convince myself that if I followed the rules it would have a positive expectancy. I think I could then just do it mechanically. I don't mind boring and mechanical. But I'm spending a lot of time trying different systems that over time are not positive in expectancy. I'm using Van Tharp's steps and they take a lot of time, but I'm not taking short cuts.
So my wish may be unrealistic, but I'd like to hear about a testable system I can try. I know what you'll probably say: The COT. And you're right. I really would be happy with $1M from each $10K in 17 years much less $5M so I don't have to put everything in the COT and I'd like to be more active with some thing else.
If I'm way off base, please let me know,
Thanks,
Howard
PS sorry for the long post

Howard,

That's fascinating about the Iowa Gambling Task. I had never heard of that. Very interesting.

Nope, I'm not going to say COT . I agree that you need to have something else going for you. Even I have never had more than 65% to 70% of my liquid assets in COT strategies. The rest is always doing something else. That's why even though the Model Portfolio is currently 60% committed to the COT it is only down about two or three percent while the market has been going straight up. But much of what I do is not strictly mechanical and I can appreciate your desire to have something that is mechanical.

If you're willing to be the guinea pig, Howard, I'm going to use you as a case study and, hopefully, it will help others as well. Let's start off this way - can you give me some particulars of some of the systems you've tested? If I know the types of things you've tried then I think I can lead you in the right direction. This process may take several posts over a period of time. So be patient.

Larry

Check out ‘fundadvice.com’ for 4 free systems that work on both the SP500 and Nasdaq 100.

there current position is 100% long - see
www.fundadvice.com.

A full explanation of their equity trading models is available @ www.fundadvice.com/modelsexplained.html#sys4

The track record of these systems from 1972 to 2002 can be viewed @ www.fundadvice.com/modelsexplainedtable1.html.

They will even e-mail you each time a new signal is given by on off these models @ www.fundadvice.com/newsubform.html

These systems do not seem to return the same as the COT approach but I think they are still worth looking at.

Regards
Andy

Larry,
I've tried this on large companies (orcl, ibm, jnj etc) that seem to have good moves up and down by just looking at the chart.
1. MACD above .25 or below .25 and starts to turn. (about 50% right at 1 week) so I went no further.
2.RSI over or under (33% right at 1 week so I went no further.
3. Full stochastic simple crossover buy and sell. Was pretty good so I added initial stop at 5% and position sized accordingly. Once the parabolic SAR got below (or above) the trend I used that until the price went on to the other side of the 20 or 50 day moving average. I ended the trade when the price crossed back over the moving averages by .10 or hit the sar. Did okay on orcl but not on any other stock i tried so I figured I didn't have enough markets to trade it in.
Those are the ones I didn't throw away. I shouldn't have thrown any away now that I'm telling you this.
Thanks for your help,
Howard

Howard,

You have a trend following test (MACD) and two reversal tests (Stochastics and RSI). Any of them can be turned into a system with a positive expectancy.

Let's take the MACD study as an example since trend following is probably a better way to go to compliment the COT. Here's a simple way to greatly improve the performance of your MACD strategy. Use two screens - a daily chart and a weekly chart. Put the MACD indicator on both of them. Then just take the daily MACD crossovers that agree with the weekly MACD direction.

For example, if the weekly MACD is up (the fast line above the slow line), you would only take the daily MACD crossovers that are also pointing up. And maybe you would get out when the daily crosses over to the downside but you wouldn't go short unless the weekly agrees. When you have time test that a little bit and see if it doesn't improve performance. Also, tell me what you don't like about it and we can fine tune it more. I'm not suggesting it to you as a system. I'm just taking you through some steps to turn what you were doing into a decent system. I suggest you test it with an index like the SPX or the NDX before you test it on individual stocks. It will save you some time.

By the way, forget about percentage profitability. That's irrelevant. A system that only produces 30% winners can be very profitable and a system that produces 70% winners can lose money.

Larry

Larry,
Thanks, I'll try it. I stopped because of low profitability because according to Van Tharp, when you add in commissions, stops and slippage the reliability goes way down so you want to start with something higher. By the way, when I test this, should I do what I've been doing as the steps Tharp lays out or test it with stops, position sizing and commsions in place from the beginning. I wasn't using MACD crossover, I was using change in direction from an extreme with the low or high as the stop with some room for error. What should I use as a stop with a crossover system?
H

Larry, great example of different time frames.
So many books discuss different time frames to make trading decisions, but they really do not state it the way you just did.
Could you elaborate a bit more, besides using it with MACD.
For example. Do you just look at daily and weekly. If during the day, do you look at 60 minute as well, etc.
Thankyou

Howard, I would just test it at first as a crossover – buy when the daily fast line goes over the slow line as long as the weekly agrees. And then get out when it crosses back down. Only go short if the weekly is short. Right now I wouldn’t worry about stops or commissions or positions sizing or anything else. The main thing you’re trying to do with a backtest is to see if you’re comfortable with it. If you are, then you can tweak it. But you can do that later. Your stop is the crossover.

Van Tharp wrote his book when commissions were much more of a factor than they are today. Slippage can be a problem but it is less so if you stay with liquid markets and you don’t make too many trades.

Larry

Scuttership,

MACD is just an example. The idea is to confirm a signal with the next time frame up. So you would confirm a weekly signal with the monthly chart. Or if you are a very short term trader you could confirm a 60 minute signal with the daily chart. You use the next higher time frame to confirm, not the next lower one. In other words, you wouldn’t confirm a daily signal with the 60 minute chart, but you would confirm a 60 minute signal with the daily chart.

Larry

Larry,
Let's see if I have this right:
Fast line crosses over slow line and weekly is also positive: buy. Sell when daily signal reverses.
fast line crosses under slow line and weekly is also negative: short. Buy when daily signal reverses.
If daily gives an entry signal and weekly doesn't confirm do nothing.
Question: do I end the trade when positive goes to negative or to zero. I'll test SPY from 1999 on and there is a two week period when the two lines merge and travel together. Is that an ending signal?
h

What if you used the MACD crossover with daily and weekly signals for buy entries and used a trailing stop method to let the market take you out of the trade?

Larry,
I tried 1999 using only moves to negative or positive as the signal to leave the trade.
There were 12 trades initiated in 1999. 4 were profitable. I bought and sold at the opening of the day after the signal was clear. The whipsawing in an out between Jan 4 and May 24 (when the weekly MACD went negative) turned a potential 10 point gain into a cumulative 3 point loss. I think a way to stay in a trade longer would help. Should I do the rest of the years first or can we tweak it now?
h

Howard,

I spot checked it for 1999. The only difference is that I used crossovers above and below the MACD Histogram simply because it’s easier to see the switches. It did do poorly that year. I show it lost about 12%. One question I would ask at this point would be why did this particular trend following system lose money in 1999 – a year that the market up about 20%? The answer is that even though 1999 was an up year there were a lot of wild swings along the way as the market was beginning its topping process. And this particular system is rather slow in reacting. Therefore, there were a lot of whipsaws and, at the same time it wasn’t quick enough to catch the fast moving moves that occurred that year.

So do we tweak it or do we trash it? I personally would trash it (I had tested this one a long time ago and although, as I recall, it had an overall positive expectancy the drawdowns were too much for my taste). But the reason I would trash it may be very different than the reason you may want to. The fact that it had a low percentage of winners would be irrelevant to me. I would just care about drawdowns and whether or not it makes enough money. If I thought it was close to being something that I’m comfortable with I would tweak it and keep testing. Peter, for example, has a suggestion for that in his post.

The main reason that we’re going through this exercise is that we’re trying to design something for you, not for me. That’s why I don’t just give you a system with a positive expectancy. I firmly believe that it’s important that you design one yourself. So to go to the next step, I have a couple of questions. You seem to want a system that has a high percentage rate of winners. If so, we’re going to have to discard most systems that are designed to cut losses short and let profits run. Because, by definition, those type systems are going to take more losses than winners. For example, I mentioned a few weeks ago that I tested a system where I bought the Nasdaq when it crossed above the 20 day EMA as long as the COT was bullish. And I sold when it crossed below the 20 day EMA. I went short when it crossed below the 20 day EMA when the COT was bearish and covered when it crossed above the 20 day EMA. I tested it over a lot of years and a lot of market conditions and, as I recall without getting out my notes, it made about 25% annually on an unleveraged basis and the drawdowns were very reasonable. But the percentage winners were only something like 38% and there were a lot of trades. I think you were concerned about the low percentage of winners and the fact that maybe it was too many trades for someone who doesn’t have the time.

So here are my questions – is percentage of winners important to you? If so, what is the minimum that you would be comfortable with? Because we may have to design one where you set target prices in order to increase the percentage of profitable trades. Or if the percentage is not important then we could tweak my 20 day EMA example by using a longer moving average. Also, approximately how many trades per year are you comfortable with?

So let’s try to find a trend following system that you’re comfortable with. And then, if you want, we’ll try to design a reversal system so that you can have one of each that you want to use.

Larry

Larry,
You make a very good point. I need to start at the basics of goals, availability, tolerence and skills.
In general my tolerence for draw down is high if I have strong faith in the system. For instance with the COT, I don't feel a need with that money to use any of the low drawdown scenarios. For this other money with which I want to be more active my drawn down tolerance is less. That is because my goal for this money is to produce income to free up time to spend more time developing more skills and areas of activity in trading. Right now I have one maybe 2 half day periods when I can be available during market hours so I need a system I can look at at night and respond to at opening. For resources: I can put aside 25K for this part of my trading, I can't program but I can do some nice things with excel, I have some knowledge of stocks and options, very little about bonds, and almost none about the commodity markets. So I feel more comfortable in liquid stocks and indexes. I really don't mind many trades a year as long as I can do them after hours at least for the first few years.
Thanks for taking your time with me,
Howard

Howard,

When I have time to think about it (no later than tomorrow) I'm going to suggest to you a very simple moving average system for entries. Of course, entry is the least important part of any system. And then I'll give you several options to consider in terms of exit. I think I have a better feel for what you would be comfortable with.

Larry

Howard,

I apologize for my tardiness but I want to continue with what we were doing. I have a couple of suggestions.

The first one is very simple. Choose an exponential moving average that fits your preferred time frame and simply enter new long positions when the price trades above the average and the average is pointing up (shorts would be just the reverse). Or you could enter anywhere in the vicinity of the average if the average is pointing your direction. You are assured of catching a trend. Experiment with the length of the moving average until you find one you’re comfortable with. Don’t be restricted to the same ones everyone else is using like the 50 day and 200 day simple moving averages. One I have found that seems to work well in a variety of markets is the 70 day EMA. It is approximately three months of data. Or if you wanted to make it almost exactly three months of data you could use a 66 day EMA. If you want to trade a shorter term trend you could use a one month EMA – 20 to 22 days. Or you may prefer to trade on a longer term trend and use the weekly chart. A good time frame would be about six months on the weekly. So you would be looking at an EMA in the 25 to 30 week range.

In terms of exits you have a variety of choices. The obvious one would be if the price trades above or below (depending on whether you’re long or short) you moving average, get out. But there will be times when you are whipsawed a lot if you do that. Other alternatives would be to place a trailing stop at some multiple of ATR. Or a Parabolic SAR stop. Or a stop right below a recent low (or high). Another way to do it would be to measure what percentage the price has violated your moving average in the past. And then you could place a stop right outside that percentage. For example, if looking back over a certain time frame the price had not violated the moving average by more than 1.5 %, you could place a stop at 2% above or below the moving average.

I know you like to take partial profits after the market has moved your direction by a certain amount. There’s nothing wrong with that. A lot of very good traders do it that way. So what you may want to do is set objectives. Take a third once the price reaches your first objective. Take another third at a second objective. And then let the market take you out of the final third on a trailing stop.

The other suggestion I have for you is Elder’s “Triple Screen” system (see Alexander Elder’s book, “Welcome Into My Trading Room”).

Here’s the way it works. The first screen is your weekly screen. On that screen you want to make sure that your weekly EMA is pointing your direction. Or to put it another way, you’re using that screen to give you the direction of the longer term trend.

Your second screen is your daily screen. You want to use oscillators to look for trading opportunities in the direction of the longer term trend. You could use several, but the main two are probably Stochastics or RSI. Don’t use them both. Choose one or the other. If you decide on RSI, I suggest you change the default on StockCharts to 5 or 9 days. 14 days is too slow. So let’s say the weekly moving average is pointing up. On the daily chart you want to use oversold conditions on Stochastics or RSI to buy. And, of course, you would do the opposite for shorts.

The third screen helps you pinpoint entry points. If you are looking to go long you could use a close above a recent high as your confirmation to enter the trade. The reverse would be true if you are looking to go short.

If you decide to try any of the above, Howard, let me know how it goes. Also, if you need further clarification I will be happy to help there as well.

Larry

Larry,
You've given me lots to try. Thanks for your time. I'll probably look at most if not all of them.
I got revved up to try something and what you said about SAR in a previous post got me to thinking. I looked at the weekly chart of QQQ from March 1999 to Dec 2003 (that's when stockcharts.com had SAR and MVA data). First I tried: Buy on open of the 3rd week after price moves above SAR and sell when it hits the SAR again. Opposite for shorting. No stops or position sizing. I used the third week because there were bunches of one and two week moves that would have whipsawed me in and out.
Results:23 trades; largest drawdown was 16.5% and over all the system was not very profitable: started with 10K and had $10,525 without commissions.
That didn't feel like a very good base for tweeking so I tried it again with only taking trades if the price was on the "right" side of both the 20 and 50 MVA. Results: 12 trades, only one large drawdown of 16.6% and an end result of $24,247 after the 4.75 year period. I tried to tweak that one with a stop of $2 adverse move from opening. I based that on what the maximum adverse moves had been in the winning and losing trades (5 losers, MAEs: 1.25,4,4,3,6; 7 winners, MAEs: 0,7,1,1,0,1,1.5). That stop took out a big winner so the results were worse $20,885 at the end of the period.
That was just practice unless you think there's something to it. I have a feeling I was over modeling and would need to check it independently on another market before I go further. Your ideas, having been thought of and tried by others before, seem more likely.
Thanks again,
Howard

larry,

since you’re looking at systems see what you think of this. it’s basically system #4 with a wrinkle.

1. When the COT report indicates that commercial traders are net long and the S&P is BELOW its 40 week simple moving average, take a non-leveraged long position.

2. When the COT report indicates that commercial traders are net long and the S&P is ABOVE its 40 week simple moving average, take a 2 to 1 leveraged long position.

3. If you have a non-leveraged long position and the S&P closes ABOVE its 40 week simple moving average and commercial traders are still net long, increase your exposure to a 200% leveraged position.

4. If you entered into a leveraged long position and the S&P closes BELOW its 40 week simple moving average and commercial traders are still net long, decrease your exposure to a non-leveraged long position.

5. If you are in a long position (leveraged or non) and the COT report indicates that commercial traders SWITCHED to net short the S&P – BUT, the current price is ABOVE it’s 20 DMA, maintain the long position.

6. If you are in a long position (leveraged or non) and the COT report indicates that commercial traders SWITCHED to net short the S&P – and the current price is BELOW it’s 20 DMA, switch to the short position (leveraged or non according to the short sell rules below).

7. If you are in a long position (leveraged or non) and the COT report indicates that commercial traders SWITCHED to net short the S&P – AND the current price CROSSES BELOW it’s 20 DMA during a week, switch to a short position (leveraged or non according to the short sell rules below).

For short positions:

1. When the COT report indicates that commercial traders are net short and the S&P is ABOVE its 40 week simple moving average take an non-leveraged short position.

2. When the COT report indicates that commercial traders are net short and the S&P is BELOW its 40 week simple moving average take a 200% leveraged short position.

3. If you have a non-leveraged short position and the S&P closes BELOW its 40 week simple moving average for any given week and commercial traders are net short, increase your exposure to a 200% leveraged position.

4. If you entered into a leveraged short position and the S&P closes ABOVE its 40 week simple moving average for any given week and commercial traders are net short, decrease you exposure to a non-leveraged short position.

5. If you are in a short position (leveraged or non) and the COT report indicates that commercial traders SWITCHED to net long the S&P – BUT, the current price is BELOW it’s 20 DMA, maintain the short position.

6. If you are in a short position (leveraged or non) and the COT report indicates that commercial traders SWITCHED to net long the S&P – and the current price is ABOVE it’s 20 DMA, switch to the long position (leveraged or non according to the rules above).

7. If you are in a short position (leveraged or non) and the COT report indicates that commercial traders SWITCHED to net long the S&P – and the current price CROSSES ABOVE it’s 20 DMA during the week, switch to a long position (leveraged or non according to the rules above).

8. On any day that the S&P trades 20% or more below its 200 day simple moving average, get out of the market by taking profits on short positions and re-establish leveraged short positions when the S&P moves to within 5% of its 200 day moving average.


one of the “flaws” the current system has is entry. either too early or, occasionally, flat out wrong.

the idea of using the 40 week MA as a confirming indicator got me thinking… we could use a shorter moving average (20 day?) to confirm entry too. this might prevent whipsaws, too early entry and large drawdowns.

the basic idea of THIS strategy is: use the COT to establish a directional bias – but use the PRICE direction (i.e. the 20 DMA) to dictate entry.

let me know what you think.

-ja

Ja,
did you test it? How did it do? If it does better over the last 17 years than strategies 1 through 4, does it matter what any of us think?
howard

i have NOT tested it. i want to hear any obvious flaws someone may see that i may have overlooked.

larry knows the COT data so well he can sometimes just look at a system and say "yea or nay".

-ja

Joe,

Yes, I had a chance to read your proposed rules and I can tell you that it would work very well. It is basically Strategy No. 4 with the 20 day moving average included. You will have more trades but you will have less in drawdowns. I can't tell you if it would outperform Strategy No. 4 without testing it, but I'm sure it would be close.

The only suggestion I would have is to use an expnonential moving average rather than a simple moving average with the 20 day part. The only reason I say that is I tested a 20 day EMA over quite a few years with very good results.

Larry

Joe how did the test of your system come out?

hey,

i was working all weekend on my house... i haven't done my testing yet.

-ja

(anyone want to test it 1st? be my guest!)