Replies to post #6565 on A.I.M. Users Bulletin Board (AIMUSERS)

[Undertakr]

It appears you have the 'buy now, sell now' outlook on stocks, which is fine, and I hope you have lots of success with that! However, if you look at AIM and understand AIM, you would see that if you choose a good stock for a good company, not a high risk stock, getting all your money out and taking all your shares off the table isn't always the best solution. It's nice that you get 100% of your investment out of a stock quickly, however, you are also cutting the amount of shares you have far greater than AIM does.

The stock doesn't dictate the amount of profit really, the price of the stock (meaning if it's a $120 stock or a $2 dollar stock) doesn't affect it, it's number of shares that creates your profit over time.

- Takr

[aptus]

Hi Nimbus,

I'm just getting caught up with the BB posts after a great Christmas.

I've responded to your note with some clarifications on what I meant by backtesting as well as a question and an observation. See below...

CLARIFICATION
What I meant by backtesting is to choose a setup based on some period of past data, then use that setup on future data and see how it performs. You want to perform blind tests as this is the only way to determine whether a system has a decent chance of working in the future. Unfortunately a few months of returns doesn't cut it.

For example, I might look at the period from Jan 21st to Jan 31st, 2002 and then create a setup based on that period. Then I'd run the test on data from Feb. 1st to Feb. 11th, 2002 and note the results. I'd repeat the process (creating my setup using data from Feb. 1st to Feb. 11th and running it on data from Feb. 12th to Feb. 22nd) until the current date was reached.

Then I'd repeat this process for other periods, such as 1969 to 1970, that had different market characteristics. I'd want to do this as many times as possible.

Once I had the backtested returns, I'd compare the returns to the B&H strategy's returns to see which strategy worked best.

QUESTION

"Most of these are safer stocks than I ever looked at to get AIM action"

How do you define "safer?" Is it a lower standard deviation? Or some other measure?

OBSERVATION
One observation I've made is that although x-dev is compared to AIM in a number of posts I've read, I don't believe it can be compared to AIM. AIM is a long-term investment strategy designed to increase returns and minimize risk while x-dev is a short-term trading strategy. Any logical comparison ends right there. It's like comparing a TA system to the Buy and Hold strategy or comparing apples to oranges. At the end of the day we're after the same results (i.e. increased returns/minimized risk), but the path we take to get there is completely different.

x-dev contains one AIM-like feature, that being the ability to size a trade, however I can't see any other similarity. The algorithm used to size the trade is completely different from AIM, x-dev doesn't keep track of the relative amount of money you've invested in your portfolio (i.e. it has no Portfolio Control), it will completely sell you out of a position, it's meant for grabbing short term gains, you must continually modify your setups over relatively short periods of time and the list goes on.

I view x-dev as a trading tool that may or may not work over the long-term (if anyone has supporting data please let me know). However I don't view it as an AIM derivative or even anything remotely resembling AIM.

Comparing x-dev to AIM is like comparing a system that uses fast and slow EMAs that sizes trades, based on, say, how many days the fast EMA took to cross the slow EMA, with AIM. Sure it has the trade sizing functionality, but it's based on a completely different algorithm AND it's a trading system, not an investment system.

Keep in mind that I'm not debating whether x-dev works for some people or not. Rather I'm stating my belief that x-dev is not in anyway like AIM and therefore cannot be compared to AIM in that regard.

If we want to compare the relative performances of the two systems, then that's a logical goal. However we'd have to setup blind tests that covered a long period of time and that covered a variety of market conditions. Then we'd have to measure both the returns and risks in order to acurately compare the two systems.

You can do this automatically with AI, but I'm not sure if you can do this automatically with x-dev (somebody correct me if I'm mistaken). If anybody wants to do this for x-dev I'd be interested in seeing the results.

Finally, it is possible to continually change the AIM parameters (much like x-dev continually changes its setup) to take advantage of short-term fluctuations, however that introduces emotion into the strategy and elminates the mechanical nature of AIM. Therefore I usually don't recommend this to anyone but the most advanced AIM users.

Regards,
Mark

http://www.automaticinvestor.com