Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Hi again Matt,
Good questions. I can imagine there was a time when IBM might have experienced 20% daily swings, but that would have been a long time ago.
As a company grows, it goes through a number of phases. Each phase it enters causes its stock to behave in a different manner. So perhaps (and I haven't checked the historical data on this so I'm only guessing) IBM did behave like ICGE at some point, but it has since left that phase and moved into another one.
I certainly didn't mean to imply that a stock couldn't start out in one group and then change groups based on the business phase it's in. However if my time horizon is, say, 5 years, then I'd feel quite confident in saying that IBM is in group A (as an example) and ICGE is in group D (for example) and they should be treated differently based on their current characteristics.
Now if one of ICGE's company's suddenly discovered a breakthrough technology with high barriers to entry, and this propelled ICGE into another phase such that its stock no longer behaved like a group D stock, then I'd say you'd have to change the AI model you're using with it.
However these changes will certainly be noticed by the investor; I don't think a company jumps from one phase to another without anyone noticing.
Regarding your fourth group of stocks (with no volatility fingerprint), I wouldn't rule that out. However I'd suggest that if we're going to use AI, we should search out stocks with characteristics that are good for AI.
However your points are well taken. I'm not saying it will be easy to group stocks, however I think it's quite possible with lot of hard work and a systematic approach.
Regards,
Mark.
http://www.automaticinvestor.com
Hello Tom,
Welcome back. Sounds like you had a great time over at LH's place.
I was chatting with him a while back and he mentioned he was involved in a housing development. That guy sure keeps busy.
I'm glad to see the IW back in circulation. I've been waiting to buy more NT (AI has been recommending a very LARGE purchase for some time now, but I've decided to wait until the IW is in the low risk area).
Regards,
Mark.
http://www.automaticinvestor.com
Hello Matt,
Yes, you're right to say that something like the DOW 30 wouldn't make a good group. I mean these stocks were picked arbitrarily because they're supposed to be the biggest and the best of US companies.
However when speaking about groups, I mean groups of stocks that have similar characteristics (such as volatility, price pattern and the like).
As an example, let's use IBM and ICGE. I'd say a 20% one day swing in IBM would be a rare occurrence. However 20% in one day for ICGE would not be as rare.
Therefore I'd subjectively say that these two stocks aren't in the same group. I could further refine this by adding additional characteristics, such as Standard Deviation, that define my groups.
Once I have a complete list of these definitions, I can then objectively test each stock to see which group it falls into (based on the definitive characteristics I've chosen).
The trick is to choose the right characterisitics. As you can guess, I don't have a complete list of groups (or even a definitive set of characteristics that define a group) yet -- but I am working on it and hope that one day I'll have most of the groups categorized.
Once that's done it will be a simple matter to choose a model based on a chosen stock. However I don't expect this to be completed in the near-term -- unless a few dozen people jump in and help ;0)
I'm looking forward to your ideas on W%R (and other predictive indicators).
Regards,
Mark.
http://www.automaticinvestor.com
Hello Tom,
That's a good question. I hadn't given any thought as to when to rebalance. The common wisdom is to rebalance every year or two at the beginning of the year. I lean towards rebalancing every two years (perhaps even every three years depending on your holdings).
But it would probably be better to rebalance at some point other than the start of each year (when everyone and their dog is rebalancing and concentrating on their portfolios).
However I don't have any data to support this, but would be interested in hearing if there is data available.
I decided to create this board so that I wouldn't be cluttering up Tom Veale's main board with cutting edge Automatic Investor stuff (which some AIMers might not consider to be AIM and others might consider all of the references to a specific software product, i.e. AI, to be blatant advertising).
As you continue to learn about optimization and Models, please feel free to post your ideas here.
Regards,
Mark.
http://www.automaticinvestor.com
Hello Don, yes this site can be used as a forum to influence AI improvements, but I also hope it will be much more.
I'd like to see some really cutting edge ideas and theories here. Even if none of them end up in AI, I think we might learn something we never thought about. And this germ of an idea, when combined with other germs, might lead to an idea that will end up in AI and improve everyone's returns as a result.
My view on multiple stocks in one portfolio is that there's no need to do it. I think you'd be better off with one stock per portfolio. The multi-stock approach, in effect, creates a new stock (or an index or whatever you want to call it) with its own unique characteristics.
However it becomes increasingly difficult to optimize this index because you've introduced new variables (i.e. each stock in the portfolio contributes something to the behaviour of the index).
The interactions between these stocks increase exponentially as each new stock is added to the index. The search space is already large for one stock (in a portfolio). If you exponentially increase this, it becomes clear that the time necessary to find the optimium settings will be too great (even with a time-saving tweak such as a genetic algorithm).
Optimizing one stock with the 6 standard optimization parameters set to their maximum ranges is already too much for a brute-force search. (See the AI 2.0 User's Guide for an example.)
In addition, I think that an index's past performance will not as accurately reflect its future performance as much as a single stock's past performance will reflect its future performance.
I do think that your suggestion would be a good one, but more as a research tool to indirectly gather insights, rather than a direct way to optimize multiple stocks in one portfolio.
Unfortunately there are a number of "need to add" enhancements, lined up like planes on an Atlanta runway during a thunder storm, that I have to address before I can even consider putting this one in.
Jack Park (http://www.thinkalong.com ) did some interesting stuff many years ago that you might find interesting.
Regards,
Mark.
http://www.automaticinvestor.com
Hello Conrad,
Welcome to the board. I'm looking forward to hearing your ideas.
"This index will behave quite different than any one of the stocks and has also lower volatility than any one of the stocks."
The index will never have a lower volatility than all of the stocks (i.e. some stocks will have a lower volatility than the index and some will have a higher volatility), assuming all stocks in the index don't have identical volatilities.
I agree with you that AIMing each stock separately is the way to go. I don't see any advantage to AIMing a basket of stocks in one AI portfolio. The only reason you would do this is to diversify, however you can diversify by AIMing multiple stocks each in its own portfolio.
You can even go one step further and share the cash reserve amongst all of these stocks (i.e. AI thinks each stock has its own cash reserve, but the underlying brokerage account lumps all of the cash into one big pot). The danger here is that you might not be able to act on BUY recommendations in one portfolio if another portfolio has used up the cash.
I also have high hopes for using periodic rebalancing with AI. In the future, AI will include tools that will make this much easier. My current favorite approach is to use MPT (although I'm not out of the testing phase with this yet), but I'm keeping a eye on what Jibes comes up with (AIM-Rebal).
I'd also like to hear what you and others think about optimizing.
Regards,
Mark.
http://www.automaticinvestor.com
Hi Don,
Yes, it will be as you describe.
My thought at the moment is to add a checkbox to the Configuration window that will allow you to turn the MattMod on or off. If turned on, then the Buy and Sell Resistance parameters (which can be set to any value from 0 to 100 inclusive) will control the percentage used by the MattMod.
You'll also be able to use it with the Historical Analyzer and the Optimizer.
As always, thanks for your fine suggestions.
Mark.
http://www.automaticinvestor.com
Hello Matt, Yes, I believe you are correct when you say that people use optimization (or tuning) incorrectly.
I agree with everything you state regarding optimization except for how you would go about finding the optimum parameters.
I believe that stocks generally fall into a relatively small set of groups. Therefore I don't think that an optimization should be done over a universal set of stocks. Rather a number of optimizations should be done over a statistically significant set of stocks in each of these groups.
You'll then come up with parameters tuned for stocks that behave like the ones in that group. So you would have a set of optimum parameters for Group A, another for Group B, still another for Group C, and so on and so on and so on...
This is the main purpose for my adding Models to AI. My intent was not to have a different model for each individual stock (although I'm sure some users will do this), but rather to have a Model for a group of stocks that share similar characteristics.
Then once you choose a stock, you'd determine its group (by backtesting it in AI's historical analyzer with the various Models) and select the appropriate Model.
This can be a very objective exercise, once the Models have been created and tested. And this is where I hope all of the Automatic Investor 2.0 users will jump in. I'm hoping to see a variety of Models appear. These Models can then be tested by other AI 2.0 users to either confirm or dispute them. Once a Model has been confirmed, it can then be made available to everyone.
Regarding W%R, you are correct that AIM doesn't take time into account. This is something that I also feel needs to be addressed. I'll be working on this at some point in the near-future and will post my findings here.
If you have any specific suggestions or comments, please let me know.
Thanks,
Mark.
http://www.automaticinvestor.com
Hi banjanxed, thanks for the endorsement and I'm glad you like the current version.
But wait until you see what's just around the corner...
More charts (I know, I know, you're probably thinking, "here he goes with the charts again...", but they're really coming this time -- very soon).
An Enhanced optimizer incorporating a genetic algorithm (makes for faster optimizations).
Performance improvements (possibly including the MattMod as well as other stuff the elves at the Automatic Investor Factory are currently working on).
And they'll all be classified as minor upgrades (translation: FREE for registered Automatic Investor 2.0 users).
I noticed that in the past you had some ideas for AIM tweaks. If you ever get the urge, feel free to post them here.
Regards,
Mark.
http://www.automaticinvestor.com
For those AIM tweakers among you (you know who you are), I've created a new board called the Advanced Automatic Investor board.
You can access it here: http://www.investorshub.com/boards/board.asp?board_id=1172
Hop on over if you'd like to discuss the latest and greatest Automatic Investor ideas and tweaks. This is your chance to make suggestions that just might make it into a future version of AI.
WARNING: this board might contain unsuitable language -- words like optimize, tuning and self-adjusting-SAFE for example, viewer discretion is strongly advised ;0)
Regards,
Mark.
http://www.automaticinvestor.com
Some do.
If you backtest over a sufficiently long time period you'll find that using Margin can significantly increase returns, but (and there's always a "but" isn't there?) it can also kill you pretty quickly if you choose a lower-quality stock.
Many times you'll find AIM will dip into margin a bit and then quickly pay it off -- this is the ideal case. If your backtesting reveals that AIM is constantly borrowing on margin (especially if it's always hovering down around your maximum margin limit) over long, consecutive periods of time, then margin should probably not be used with those types of stocks.
Here's some advice from the Automatic Investor Web site...
"Use margin sparingly. Margin can help you take advantage of unique opportunities. But always remember you are taking out a loan. One that will have to be paid back. In addition, if your holdings go down, you may be required to deposit additional funds - or risk being forced to sell your stocks at a poor price. If you do decide to use margin, buy only high-quality stocks, and pay it off as soon as possible."
Regards,
Mark.
http://www.automaticinvestor.com
One of the more promising modifications comes to us from Matt Crider.
He proposed to set the buy and sell resistance values to a percentage of the entire portfolio value (i.e. Equity Value + Cash Reserve) rather than simply Equity Value. The percentage he recommended is 5%, however other values can also be used.
So far this simple modification appears to consistently improve results over standard AIM. I'm currently in the midst of testing it further, however it's been christened the "MattMod."
When used with real stocks (e.g. the Nasdaq 100), I've found using 10/10 resistance settings to be better than Matt's recommended 5/5. However the 5/5 works better on the theoretical Lichello example.
Don Carlson has found different settings work better in his testing.
While I'm not adverse to tuning parameters based on an equity's particular characteristics, I do feel that a standard setting (or perhaps two or three standard settings) are superior to multiple (i.e. 8 or 9 different) settings. So there needs to be more work done in that area.
Regards,
Mark.
Hello Jibes, I'll have to check that out. I became interested in W%R because it seemed to do a good job of predicting ups and downs just before the up or down.
Even though AI is reactive, rather than predicitve, I think there might be some value in using a predictive indicator as follows:
Let AI continue to react to price (and optionally volume) changes in order to issue recommendations. But, once AI issues a recommendation, modify it according to some predictive indicator. This might allow you to allocate your trades more efficiently -- of course I haven't tested this particular theory yet (one of the drawbacks to running a software company is that you actually have to produce software -- which really cuts into your research and testing time
The big problem, I see, is that although W%R appears to predict prices nicely, many times it will predict a price move but the actual move won't take place for a while. This might cause a premature modification of an AI recommendation.
What do you think about that line of thought?
I've also briefly looked at your AIM-REBAL effort. I think this is a good idea (as a matter of fact I gave a presentation at the 2001 AIM User's meeting, in Las Vegas, about rebalancing and AIM -- although I suggested using Modern Portfolio Theory to handle the rebalancing, see http://www.automaticinvestor.com/articles/mpt.html for more details).
I'll be following along and reading your results. If, during the course of your research, you come across anything interesting, feel free to post it here.
I noticed you didn't get good results with the first 4 DOW stocks. That mirrors my findings when testing the DOW -- it usually underperforms other indexes when used with AIM-like strategies. I'd suggest trying the Nasdaq 100. There are a number of good, volatile issues in there.
I'd also be interested in how AIM-Rebal works over various time periods (e.g. 1969 to 1979) other than those periods ending with this year.
Regards,
Mark.
http://www.automaticinvestor.com
Thanks ES, that answered my question.
Mark.
http://www.automaticinvestor.com
Can someone tell me how you put pictures (i.e. gifs and jpegs) into a post?
Thanks,
Mark.
http://www.automaticinvestor.com
Hello Myst,
I've heard a great deal about X_DEV but haven't had the time to look at it yet.
Lots of people saying lots of good things about it. When I get a chance I expect I'll delve into it.
If you have anything you'd like to discuss on improving the AIM algorithm, please feel free to post it here.
From time to time I'll be throwing out ideas for discussion -- well while I'm at it, why not start now...
Any thoughts on using the W%R to modify AIM recommendations?
Regards,
Mark.
http://www.automaticinvestor.com
Hello Charlie,
The main reason that AIM is not symmetrical is because you add to the portfolio control on each purchase, but you don't adjust the portfolio control for a sale.
Therefore as you are selling on the way up, the portfolio control isn't changed. However when you begin to buy, the portfolio control is increased. Since buy and sell recommendations depend on the portfolio control, the recommendations will be different (i.e. not symmetrical).
In addition, under standard AIM, you'll never be 100% in cash. However you can be 100% in equities.
Regards,
Mark
http://www.automaticinvestor.com
Hello Jibe,
Yes, there are at least 5 other software packages (including online services) available that allow you to implement AIM.
Mine is called Automatic Investor. You can download a free 30-day trial at http://www.automaticinvestor.com/trial.html
Please let me know if you have any other questions.
Thanks,
Mark.
http://www.automaticinvestor.com
Hello Matt, I didn't run the 8% test against the standard lichello algorithm, however the 5% and 10% runs did beat standard lichello when I ran them a few days ago (the dates were slightly different, but the results should be similar).
I'll keep you updated as I continue looking at this.
Regards,
Mark.
http://www.automaticinvestor.com
Hello Jack,
To get around the help problem, just copy the user's guide
(http://www.automaticinvestor.com/romeo/AIHelp20.pdf ) to the folder where you installed Automatic Investor.
If you downloaded the user's guide prior to June 4th, you'll need to rename it to "AIHelp20.pdf" (before June 4th it was called "AI20Help.pdf"). All downloads on or after June 4th have the correct name.
The user's guide has an explanation of the Volumizer (you can simply search on "Volumizer").
In general, it performs well on stocks that have an average volume of more than a few million shares a day.
To determine if it makes sense to use it, you should run a historical analysis on the stock/fund in which you're interested, both with and without the Volumizer enabled. Then look at the return and the % Capital at risk values.
It also works best over longer periods (i.e. 4+ years) rather than shorter periods, however in some cases it works very well over short periods (e.g. with the Internet stock sector).
If you have any other questions, please let me know.
Regards,
Mark.
http://www.automaticinvestor.com
Hello Matt,
I'm not sure what parameter settings Tom and you used, however I ran some 5 year (going back 5 years from yesterday) tests on the Nasdaq 100.
Basically I used an initial cash reserve of 0% (i.e. 100% invested to start), 5% minimum buy and sell, the Matt modification and then varied the buy and sell resistance parameters. I ran each configuration over all 100 stocks in the index. I also re-ran all the tests with the Volumizer enabled.
I used 5%, 8% and 10% values for buy and sell resistance. The 8% comes from Don Carlson's suggestion that he found 7.5% to be optimal in his tests.
There were a total of 600 individual results forming 6 groups (each group consists of one run of the 100 Nasdaq stocks). I've summarized the results below.
Group 1: 5% Buy Resistance/ 5% Sell Resistance
AUTOMATIC INVESTOR ANALYSIS RESULTS
Total Portfolio Value: $2,903,644.40
Total Investment: $1,000,000.00
Simple Return: 190.36%
Group 2: 8% Buy Resistance/ 8% Sell Resistance
AUTOMATIC INVESTOR ANALYSIS RESULTS
Total Portfolio Value: $3,289,723.32
Total Investment: $1,000,000.00
Simple Return: 228.97%
Group 3: 10% Buy Resistance/ 10% Sell Resistance
AUTOMATIC INVESTOR ANALYSIS RESULTS
Total Portfolio Value: $3,360,400.90
Total Investment: $1,000,000.00
Simple Return: 236.04%
Group 4: 5% Buy Resistance/ 5% Sell Resistance with Volumizer
AUTOMATIC INVESTOR ANALYSIS RESULTS
Total Portfolio Value: $3,295,647.64
Total Investment: $1,000,000.00
Simple Return: 229.56%
Buy and Hold Total Value: $3,164,894.04
Group 5: 8% Buy Resistance/ 8% Sell Resistance with Volumizer
AUTOMATIC INVESTOR ANALYSIS RESULTS
Total Portfolio Value: $3,608,397.96
Total Investment: $1,000,000.00
Simple Return: 260.84%
Group 6: 10% Buy Resistance/ 10% Sell Resistance with Volumizer
AUTOMATIC INVESTOR ANALYSIS RESULTS
Total Portfolio Value: $3,875,191.77
Total Investment: $1,000,000.00
Simple Return: 287.52%
As you can see, as the buy resistance and sell resistance parameters increased in value, so did the results. I'd like to stress, however, that this is not a complete test. I only used one period (june 97 to june 02) and only one timeframe (5 years).
However these tests do use real-world data on a major index that I feel is quite suitable to use with the AIM method.
I plan to run some more detailed tests over the next few weeks.
Regards,
Mark.
http://www.automaticinvestor.com
Hello John,
Well, although Matt's modification is not in the Public Domain, he and I have reached an agreement on its use in AI. So, assuming the modification passes testing, you should be able to use Matt's modification with AI 2.0 shortly.
It would be added as an optional switch that can be turned on or off as desired.
I'll keep you updated via this board.
Regards,
Mark.
http://www.automaticinvestor.com
Hello Tom,
That is surprising. I ran two tests over the last 5 years, one on the DOW and the other on the Naz 100, and both performed better with the 10% setting than with the 5% setting.
Of course without systematically testing the modification it's difficult to determine whether those two tests are anomalies or whether the theoretical example doesn't closely enough reflect the real world.
Regards,
Mark.
http://www.automaticinvestor.com
Hello Tom,
I was in the process of analyzing Matt's modifications myself to determine why it worked.
My first thought was that it worked well on the SELL side because as the cash portion grew, the rise in price needed to be greater (than standard lichello) in order for a sell recommendation to be triggered. This basically delayed sales if there was a relatively large amount of cash (similar to your Vealie technique).
On the BUY side, as more cash accumulated in the reserve, the price would have to fall more before a buy recommendation was triggered. This didn't intuitively make sense to me because my thinking was that the less cash you had, the more you wanted the price to fall before you bought -- not the other way around, i.e. the more cash you have the less you want the price to fall before you buy.
However in a few non-scientific tests I tried, the modification appeared to work well on the BUY side too (i.e. reducing the Buy side resistance caused performance to drop).
My guess is it's because of the fact that you can't just look at buy recommendations independently of the entire AIM system (i.e. Buys increase portfolio control which then affects the next buy and sell, etc., etc., etc.). Add to this the fact that perhaps the test period I used was such that stocks actually did go down to those deeper levels, and thus the modified algorithm was able to purchase at a deeper discount. If another period was used, perhaps prices would not have fallen as much and the modified algorithm would have missed some purchases -- it's hard to tell at this point.
However since Matt's modification isn't currently in the public domain and I've got a number of other things to work on, I've put this testing on the back burner. But from the little I've seen, it sure looks like it has potential.
Regards,
Mark.
http://www.automaticinvestor.com
Hello John,
When I posted the results last night, I had intended to set up a full test of Matt's method (basically the same testing methods I use to validate AI models) with an eye towards including it in AI 2.0 if it passed the tests.
But Matt sent me an email this morning stating that his modification is not in the public domain. So although it can be used by individuals, it can't simply be added to a commercial product.
So unless Matt removes this restriction, at some point in the future, or we come to an arrangement, AI 2.0 won't implement Matt's modification and I won't be testing it any further.
However AI 2.0 will include a number of other enhancements that should improve performance. It already includes the Volumizer, which can improve results significantly, and other improvements are just around the corner.
Let me know if you have any other questions.
Thanks,
Mark.
http://www.automaticinvestor.com
Hello Karel,
Yes I changed AI so that the results of Matt's version uses 5% of Shares Value + Cash while standard Lichello uses 10% of Shares value (no cash).
I've also run a few tests on the DOW and Naz100 using Matt's version with 10% (of Shares value + cash) and the results appear better than simply using 5%.
However if we increase to 15% and then 20% the results aren't as good. The 10% value seems to be an excellent choice.
Similarly with the Volumizer, 10% seems to perform best. Whatever Lichello did to come up with 10% seems to carry forward quite nicely.
Regards,
Mark.
http://www.automaticinvestor.com
Hello Matt,
Well if Luke and John (in that order) can add significantly to AIM I sure hope they show up soon ;0)
Regards,
Mark.
Matt's Modification Part 3...
And finally I used the Automatic Investor Aggressive model with Matt's modification (the Aggressive model starts with a 0% initial cash reserve and enables the Volumizer).
The results beat standard lichello by $1,383,610 over the last 5 years.
I also tested the AI Aggressive model without Matt's modification. Matt's modification beat standard lichello (with both using the AI Aggressive model) by $110,463. So although further testing is required, I think Matt's onto something here.
Matt's modification with AI Aggressive model:
AUTOMATIC INVESTOR ANALYSIS RESULTS
Total Portfolio Value: $3,476,787.78
Total Investment: $1,000,000.00
Simple Return: 247.68%
======= CONFIGURATION =======
Buy Resistance: 5%
Sell Resistance: 5%
Initial Cash Reserve: 0%
Margin is Disabled
Minimum Purchase: $0.00
Minimum % per Purchase: 5%
Minimum Sale: $0.00
Minimum % per Sale: 5%
Control Increment: 50%
Maximum Cash: 100%
Vealie: 50%
Volumizer is Enabled
Standard Lichello with AI Aggressive model:
AUTOMATIC INVESTOR ANALYSIS RESULTS
Total Portfolio Value: $3,366,324.45
Total Investment: $1,000,000.00
Simple Return: 236.63%
======= CONFIGURATION =======
Buy Resistance: 10%
Sell Resistance: 10%
Initial Cash Reserve: 0%
Margin is Disabled
Minimum Purchase: $0.00
Minimum % per Purchase: 5%
Minimum Sale: $0.00
Minimum % per Sale: 5%
Control Increment: 50%
Maximum Cash: 100%
Vealie: 50%
Volumizer is Enabled
regards,
Mark.
http://www.automaticinvestor.com
Matt's Modification Part 2...
As far as I can gather, Matt's modification doesn't take anything away from non-volatile stocks, however it apparently increases the return on volatile stocks. Such a simple idea, such profound results -- I'm wondering why nobody came up with this before now (good work Matt!)
Since the 11 stocks in my last post did so well with Matt's method, I decided to see how it would do on the Naz100 stocks over a 5 year period (note that this is for your information only as I didn't have the time to run a complete analysis over various market periods and timeframes).
For this test I invested $10,000 in each of the 100 Naz stocks and assumed a $10 commission per trade.
It performed very well, beating standard lichello by $220,637.50 over the last 5 years. I also used Matt's method with a 0% starting cash reserve and it returned $842,245 more than standard lichello.
Here are the Naz100 results...
Lichello settings:
AUTOMATIC INVESTOR ANALYSIS RESULTS
Total Portfolio Value: $2,093,177.89
Total Investment: $1,000,000.00
Simple Return: 109.32%
======= CONFIGURATION =======
Buy Resistance: 10%
Sell Resistance: 10%
Initial Cash Reserve: 50%
Margin is Disabled
Minimum Purchase: $0.00
Minimum % per Purchase: 5%
Minimum Sale: $0.00
Minimum % per Sale: 5%
Control Increment: 50%
Maximum Cash: 100%
Vealie: 50%
Volumizer is Disabled
Matt's modification settings:
AUTOMATIC INVESTOR ANALYSIS RESULTS
Total Portfolio Value: $2,313,815.36
Total Investment: $1,000,000.00
Simple Return: 131.38%
======= CONFIGURATION =======
Buy Resistance: 5%
Sell Resistance: 5%
Initial Cash Reserve: 50%
Margin is Disabled
Minimum Purchase: $0.00
Minimum % per Purchase: 5%
Minimum Sale: $0.00
Minimum % per Sale: 5%
Control Increment: 50%
Maximum Cash: 100%
Vealie: 50%
Volumizer is Disabled
Matt's modification settings with 0% initial cash reserve:
AUTOMATIC INVESTOR ANALYSIS RESULTS
Total Portfolio Value: $2,935,422.83
Total Investment: $1,000,000.00
Simple Return: 193.54%
======= CONFIGURATION =======
Buy Resistance: 5%
Sell Resistance: 5%
Initial Cash Reserve: 0%
Margin is Disabled
Minimum Purchase: $0.00
Minimum % per Purchase: 5%
Minimum Sale: $0.00
Minimum % per Sale: 5%
Control Increment: 50%
Maximum Cash: 100%
Vealie: 50%
Volumizer is Disabled
regards,
mark.
http://www.automaticinvestor.com
Regarding Matt's modification...
I found it very interesting and wanted to see how it would perform in the real world. So I modified the AI 2.0 software to use Matt's modification and ran a few tests. When used with DOW stocks over a 5 year period there was very little difference in results.
However when used with some tech stocks over a 5 year period there was a significant difference.
The stocks are:
MSFT, IBM, AMZN, YHOO, EBAY, AOL, MOT, HWP, ORCL, SUNW and AAPL (the most volatile stocks beat the lichello settings significantly, the least volatile had approximately the same returns).
I've included a summary below (I can post the full reports if there's interest).
Standard Lichello settings:
AUTOMATIC INVESTOR ANALYSIS RESULTS
Total Portfolio Value: $181,331.70
Total Investment: $110,000.00
Simple Return: 64.85%
======= CONFIGURATION =======
Buy Resistance: 10%
Sell Resistance: 10%
Initial Cash Reserve: 50%
Margin is Disabled
Minimum Purchase: $0.00
Minimum % per Purchase: 5%
Minimum Sale: $0.00
Minimum % per Sale: 5%
Control Increment: 50%
Maximum Cash: 100%
Vealie: 50%
Volumizer is Disabled
Matt's Modification settings:
AUTOMATIC INVESTOR ANALYSIS RESULTS
Total Portfolio Value: $241,419.10
Total Investment: $110,000.00
Simple Return: 119.47%
======= CONFIGURATION =======
Buy Resistance: 5%
Sell Resistance: 5%
Initial Cash Reserve: 50%
Margin is Disabled
Minimum Purchase: $0.00
Minimum % per Purchase: 5%
Minimum Sale: $0.00
Minimum % per Sale: 5%
Control Increment: 50%
Maximum Cash: 100%
Vealie: 50%
Volumizer is Disabled
regards,
Mark
http://www.automaticinvestor.com
Automatic Investor 2.0 is now available!
30-day free trials can be downloaded from the AI website at http://www.automaticinvestor.com
Please let me know if you have any questions.
Thanks,
Mark.
http://www.automaticinvestor.com
Hello Everyone,
Just a reminder, today's the last day to purchase Automatic Investor in order to qualify for the low upgrade price to AI 2.0.
You can order your copy here -- http://www.automaticinvestor.com/order.html
Just a few hours left
Regards,
Mark.
http://www.automaticinvestor.com
Hello ES,
Might be a good AIMer
I'll say. SHRP appears to be an excellent AIM stock. I ran a 5 year historical analysis in Automatic Investor using Lichello's latest and greatest settings (called the "Bullish" Model in AI). It beats buy and hold 2 to 1. Here are the results...
AUTOMATIC INVESTOR HISTORICAL ANALYSIS FOR SHRP
======= PERFORMANCE =======
Current Portfolio Value: $111,612.42 (3184 shares owned)
Profit or (Loss): $101,612.42
Simple Return: 1016%
Annualized Return: 62%
Buy/Hold Portfolio Value: $56,911.94 (2857 shares owned)
Buy/Hold Profit or (Loss): $46,911.94
Buy/Hold Simple Return: 469%
Buy/Hold Annualized Return: 42%
Return on Capital at Risk: 1491.16%
Average % Capital at Risk: 68.14%
Analysis run on 30-May-02
Daily Price Data Chosen
Filters Used:
Include Buy Advice
Include Sell Advice
Exclude Hold Advice
Average Commission: $10.00
Start of Period: 30-May-97
End of Period: 30-May-02
High for Period: $22.8500 on 3-May-02
Low for Period: $2.6600 on 8-Oct-98
Initial Number of Shares: 2,286
Original Share Price: $3.5000
Original Share Price Date: 02-Jun-97
Original Investment: $10,000.00
Last Share Price: $19.9200
Last Share Price Date: 29-May-02
Average Cost Per Share: $8.54
Current Cash Reserve: $48,187.14
Current Shares Owned: 3,184
Current Security Value: $63,425.28
Number of Purchases: 37
Number of Sales: 23
======= CONFIGURATION =======
Buy Resistance: 10%
Sell Resistance: 10%
Initial Cash Reserve: 20%
Margin is Disabled
Minimum Purchase: $0.00
Minimum % per Purchase: 10%
Minimum Sale: $0.00
Minimum % per Sale: 10%
Control Increment: 50%
Maximum Cash: 100%
Vealie: 50%
Volumizer is Disabled
Portfolio Control: 64,064
regards,
Mark.
http://www.automaticinvestor.com
Hello John,
My advice would be to follow the book. After putting the recommended portion of your cash into your mutual fund you should still have some cash left over to fund some additional AIM recommended purchases (unless your Initial Cash ratio is set to 100% equities and 0% cash).
Running out of cash is a fact of life in a down market. There are a number of techniques you can use to compensate for this (such as the half-way-to-the-wall technique or updating over longer periods -- which you are already doing), however most AIMers have probably run out of cash in the past year.
So unless you have a good reason not to follow the book, you'd do well to stick with it.
Regards,
Mark.
http://www.automaticinvestor.com
Hello Tom, the book's author purchased AI 1.0 and sent me a copy of the book to review. I found it quite well written.
Basically it explains what AIM is and adds a few other concepts. It also has a number of specific examples.
Thanks for reminding me about the links from your site. Next week when AI 2.0 is released I'll send you some updated information.
Regards,
Mark.
http://www.automaticinvestor.com
Hello Everyone,
*OT* (sort of).
Well it's finally here... Automatic Investor 2.0 will be released June 1st, 2002.
For those who've been emailing me asking when AI 2.0 will really be released, this is it.
If you liked AI 1.0, you'll love AI 2.0. All of the good stuff in AI 1.0 is still there, but a number of powerful new enhancements have been added.
If you'd like to see some detailed information on what's included in AI 2.0, feel free to download the AI 2.0 user's guide.
You can download it here -- http://www.automaticinvestor.com/romeo/AI20Help.pdf (approximately 1MB in size).
You'll need Acrobat Reader 4.0 or greater. It's a free download available at the Adobe site -- http://www.adobe.com/products/acrobat/readstep2.html (if you can already view PDF documents then you probably don't need to download Acrobat Reader).
On another note. AI 1.0 will not be available for sale or download after May 31st, 2002. So if you'd like to purchase it (and take advantage of the low upgrade price to AI 2.0), please ensure you *purchase* a license (http://www.automaticinvestor.com/order.html ) by May 31st, 2002.
AI 2.0 upgrade prices will only apply to *registered* AI 1.0 users (not to those who have just downloaded the software before the May 31st deadline). So if you haven't purchased an AI 1.0 license by May 31st, you'll need to pay the full retail price for AI 2.0 (sorry, no exceptions). However I think even the full price will be well worth it (look at the functionality described in the AI 2.0 user's guide to see what I mean), but if you can save a few dollars by upgrading, then why not?
FYI here are the prices.
AI 1.0: $99 US (only available until May 31, 2002).
AI 2.0 upgrade (for registered AI 1.0 users): $20 US.
AI 2.0 Full version (for users who have not registered AI 1.0): $197 US.
AI 2.0 (upgrade and full versions) will be available June 1st, 2002. (And when you purchase your AI 2.0 upgrade or full version, you'll also receive a new eBook, "Lichello's Golden Little Secret" by David Gressett, FREE of charge -- it usually sells for $14.95 US at http://www.adeptprime.com )
Finally I'd like to thank everyone who provided feedback on AI 1.0 -- many excellent suggestions have been incorporated into AI 2.0.
I'd also like to thank everyone who helped test AI 2.0. The feedback I received was great. But I'd like to single out one tester, Don Carlson, who went far beyond what I expected of a Beta tester. He had so many excellent suggestions that I now have a list of enhancements that will probably take me through to AI 5.0 Thanks Don!
If you have any questions please don't hesitate to let me know (mhing@automaticinvestor.com).
Regards,
Mark.
http://www.automaticinvestor.com
Hello Jibe,
Yes, using one stock usually means more volatility. However there are cases when using multiple stocks (all with similar characteristics) can bring more volatility.
This can happen because given a group of stocks with similar characteristics, you'd have to choose one for your single-stock portfolio. However the odds of that stock being the most volatile of the bunch will diminish as the number of stocks from which you can choose increase. Therefore if you choose a group of stocks (from the total set of stocks), you have a better chance of picking some with higher volatility and that MIGHT bring your average volatility up to where it's higher than a single stock's volatility. Basically it's diversifying on volatility rather than price.
However this can only happen if the stocks you choose are highly correlated. If they're uncorrelated, you'll see a significant reduction in volatility. And of course, regardless of correlation, it might not happen at all -- your one chosen stock's volatility might be higher than the average volatility of the multi-stock portfolio.
Regarding stocks that go down and never recover, that's a chance you take with most types of investing methods (not just AIM). That's where good fundamental analysis comes in.
There were also discussions on "deep divers" a year or so ago on the SI board. You might want to look at those posts (search for "deep diver").
My recommendation is to choose a basket of stocks, but AIM them individually. On average, you get the best performance that way, yet you're diversified in case one of them tanks.
Regards,
Mark.
http://www.automaticinvestor.com
Automatic Investor 2.0 includes a fairly comprehensive User's Guide that might be of interest to some.
Even if you don't plan on being an AI 2.0 user, you might find some interesting and useful information that you can apply to your own AIM program. Section One and Section Three might be of interest.
On the other hand, if you do plan on owning your very own brand-spanking-new copy of AI 2.0, then Section Two is definitely for you.
You can view the AI 2.0 User's Guide here --> http://www.automaticinvestor.com/romeo/AI20Help.pdf (approximately 1MB in size).
You'll need Acrobat Reader 4.0 or greater. It's a free download available at the Adobe site --> http://www.adobe.com/products/acrobat/readstep2.html (if you can already view PDF documents then you probably don't need to download Acrobat Reader).
Please let me know if you:
1. Have any suggestions.
2. Have any comments.
3. Have any questions.
Thanks,
Mark.
http://www.automaticinvestor.com
Welcome back Tom,
Chavez is back. Looks like you left Aruba too soon ;0)
Regards,
Mark.
http://www.automaticinvestor.com
Hello Barry,
Logically I would think (all other things being equal) that there should be no difference between people starting an AIM account today and others starting one year ago (i.e. they should have the same PC and Cash/Equity ratio, assuming their portfolio values are equal).
However this opens a very large can of worms. Why only rebalance once a year? Why not once a month? Or once a week? Or once a day? Or once a second? Everytime the stock value changes we'd have to look at potentially rebalancing (as if we were starting with this new portfolio value at that exact instant in time).
And, of course, there would need to be a definitive set of criteria that allowed us to choose which method was "better." (We might arrive at these criteria using backtesting or some other method.)
This strikes me as being very similar to the deep diver discussion over at SI a year ago ( http://www.siliconinvestor.com/stocktalk/msg.gsp?msgid=15592221 )
That one was still left unresolved and I think your point will also be unresolved. Everyone will do what they think is best for their particular situation. However from a logical point of view, what people choose to do doesn't always make Spock-logic sense.
Regards,
Mark.
http://www.automaticinvestor.com