Frequently Asked Questions

updated 9/28/2005

1. Q - I have $1000 to invest, will your system work with this amount of money?

A - No, it will not work. The commission rate will eat up your profits. It is very important to get the lowest commission rate per trade as possible. However, if you have $10,000 and use the right discount broker you can profitably participate. Here are some guide lines:

Fidelity Commissions = $16 per trade ($8 in and $8 out)

Commission / Amount Invested in the Trade = Commission Rate per Trade

$16 / $1000 = 1.60% Won't Work
$16 / $2000 = .80% Won't Work
$16 / $4000 = .40% Borderline
$16 / $8000 = .20% OK
$16 / $15000 = .106% Good
$16 / $30000 = .053% Great

2. Q - I'm interested in your system, but don't fully understand the relationship of total dollars invested to commissions as shown in FAQ #1.

A - Lets look at this another way, the average profit per trade ("A" picks) is about .75%. Therefore, you absolutely have to position yourself so that your commissions are less than .75% per trade. Assume that you are only willing to give your broker .33% per trade, this leaves you .43% profit per trade (.75% - .33% = .43%).

Fidelity ($16.00/trade in and out): $16.00/.0033 = $4848, this is your minimum investment per position
Lombard ($29.90/trade in and out): $29.90/.0033 = $9060, this is your minimum investment per position

The minimum amount to invest in each position is dependant on how much your broker charges, get a cheap one!! You can trade this system with $10,000 in your account. You won't be able to invest in all the open positions, there will be so many that even millionaires can't afford to do this. Here is an email I got today (8/7/97) from a viewer:

You're on fire buddy...............I only have 11,000 dollars so I must be
selective with your picks........if I like it, I jump in. THANKS A LOT.
Wish I had more money but that's the life of the poor and infamous.

3. Q - I am trying to formulate my overall strategy to match your ROI. You have stated that to do this I need to:

have an equal amount of money invested in each position.
have a 100% invested portfolio - at times I will have to use margin.
pay zero commissions - it's possible for very large accounts.

Do I need to invest equally in each of your selections or can I invest different amounts at random depending on my available funds and still achieve the same ROI?

A - To match the ROI exactly you would have to put an equal amount in each position and buy all the positions. From a practical standpoint this isn't possible. What you want to do is develop a strategy to approximate the ROI. To a certain degree this depends on the amount of funds available to invest. If you can afford to have twenty positions at the same time you should be able to approximate our ROI. If you can't afford this many positions it doesn't mean you won't be able to achieve the ROI, it means that chances are you will deviate from it by a greater degree. This may be higher or lower depending on if you are lucky enough to pick mainly the winners.

The whole thing is a balancing act. Each individual has to work out their own strategy of which issues to buy, how much to invest in each position and when to use margin.

4. Q - Was your AI stock forecasting intelligent enough to recognize that Hurricane Katrina would hit New Orleans and cause the shut down of oil imports and refineries, and therefore increase the price of Energy stocks?

A - The neural network can't forecast specific events... hurricanes, fires, floods, earnings drops, corporate fraud, etc. It doesn't even know that these type things exists. Instead it looks at everyday market movement that relate to these events. In theory it takes all events into consideration in an indirect manner. Lets say a stock falls a great deal in one day on heavy volume and the S&P 500 is up on the same day, maybe the company got sued. The neural network 'looks' at past history of similar movement. It doesn't know the company got sued but history shows that stocks that 'act' the same way will usually continue down. It then gives the potential trade a bad score, probable resulting in an "E" rating.

Think of the neural network as a really good baseball manager. A good manager might know that his next batter doesn't do well against left handed pitchers during day games. The bases are loaded and he needs a hit bad so he sends in a pinch hitter to better his odds. The neural network knows the odds in every case and acts accordingly. It isn't right all the time but given many chances it's right a lot more often than it's wrong.

5. Q - How would the system do if a Bear market develops?

A - In a Bear market the nets will probably lose money, it depends on how it develops. Any system that goes long will probably lose money in a Bear market. Everybody wants to know when a Bear will start, I don't think it is possible to tell when one starts or stops until it is over. From our standpoint, the best thing that could happen would be a big crash as opposed to a long slow decline. Under this scenario it is highly possible that the nets could be 100% in cash at the time of the drop. On the other hand the system will most probably do very well in a situation where the market is range bound and moves sideways.

There will be infrequent times that the system will lose money even in a Bull market. See Trades #101 thru #125, this also tells you that just because we may hit an abnormal dip it doesn't mean a Bear has started and the system will continue to fail. It also means that there will be positive periods in a prolonged Bear where the system will perform very well.

New Input on 4/11/2001 concerning Bear markets: For the first five years of website history the above answer remained unchanged. However, during our 5th year the S&P 500 dropped 22.34%, an official Bear market. At the same time our system achieved a 64.42% gain.

6. Q - Why aren't A+, A and A- recommendations broken out in your History Table?

A - One of the improvements I've made 'recently', for paid subscribers, is grading the "A" picks... "A", "A+" and "A-" before the position is opened. I've been doing this for a few years now. When I first started doing this I made a decision not to track the historical ROI's of each category. If I did this it would have messed up my historical record. Since "A" picks weren't multi classified the first 7 years we were on-line the ROI of old "A" picks wouldn't have been directly comparable to that of new "A" picks. It's the old Apples to Apples and Oranges to Oranges type thing. Therefore "A+" and "A-" picks are noted as "A" picks in our Trading History after the position is closed.

7. Q - I have made my living trading stocks for 5 years but am not at all familiar with neural networks. What educational background do you have and how did you become interested in neural networks? Is this a technical model, fundamental or combination of both? I would like to do more research in the area but wonder if I'm qualified.

A - I have been interested in the stock market since high school. While in college I saw that many of the techniques I was learning could be applied to the market. We are talking mainly about computer programing, statistical analysis, multiple regression and correlation coefficients. I graduated in 1969 from Bradley University with a BS in Industrial Engineering. At that point in time I couldn't really make any practical use of this knowledge as it pertained to the market. I was headed into manufacturing not toward Wall Street.

The last few years have really made it possible for the individual to bring powerful analytical techniques into play. We have computers on every desk. We have data vendors that can feed fundamental and technical data to the machine every day. And most importantly, we have deep discount brokers where you can buy or sell $100,000 worth of stock for $2. This combination allows the individual to trade on a short term basis and do so profitably. This couldn't have been done in the recent past unless you had a seat on the exchange.

Neural networks are the easiest and most logical way to bring these "tricks of the trade" to bear on the problem. You can buy a commercially available program and create your own neural net. Before you attempt this you should be able to write your own trading systems with a program like Omega Super Charts. Be prepared to invest a great deal of time to create a good one.

To learn more about neural nets just search the WWW, there is more than enough to get you totally confused - we are talking Artificial Intelligence.

8. Q - I was viewing the results of your investments and I was wondering why you didn't have a Stop in place to automatically sell a stock like IRF (trade #48 Trading History). IRF fell from about $24 to $17 which resulted in about a 27% loss. If you consistently used a stop about a dollar below the purchase price wouldn't that protect your investment in a situation like that?

A - Placing Stops is a good way to limit your losses in case the particular trade heads south which is always a possibility. The problem with the Stop order is that it is a double edge sword. It will limit large losses on one trade but it will also cost you money on other trades. For example, a position might deteriorate by 12% from where you bought it and then rebound 6% at which point you sell it. This leaves you with a 6% loss on the trade. If you put a Stop in at the 10% loss level you lose 10% instead of 6%. This would happen frequently if you put a Stop on every position.

So, what's the best way to go? All the studies we have done strongly indicate that Stops will cost you money and we recommend that you not use them. However, some investors may choose to put 100% of their funds in a single position and they should evaluate their own tolerance for loss on such a position and judge what is best.

9. Q - When you say sell, does that mean that the stock can be shorted?

A - Do not short a stock if its recommended as a sell. There is a big difference between closing a position and going short. I have been trying to develop a neural net for short sales but the results haven't been good so far. More work is needed.

10. Q - Why don't you form a Mutual Fund? This would let the less sophisticated traders and many elderly profit from your system without the hassle of everyday trading.

A - There is a big problem in taking the STSS system to a wider audience. Say a hypothetical STSS based Mutual Fund had $50 million in assets. This means the manager would have to place approximately one million in each position. In order to fill million dollar buy and sell orders the price would move too much and destroy a good portion, if not all, of the system profitability. Keep in mind that we are dealing with an average gain of less than 2% per trade on "A" picks. STSS is a nice niche system but that's all it will ever be, it can't ever be traded extensively without ruining it.

Disclaimer