I too am Canadian and currently live in Toronto!

I've not been too active here of late ... it takes me too much time merely to read all of the posts.

Barry Savage

Thanks to everyone for their kind wishes.

As to where I am ... I am not more or less established in Toronto. I just bought a house and am working as a computer programmer (something I have not done for umpteen years) on a major actuarial modelling system (something with which I am more familiar)

As soon as I take possession and get settled (I hope all done by the new year) I should be visiting regularly again. 'Til then I'll drop by here from time to time ... It's almost impossible to keep up with all the posts.

Barry

Thanks Bernie for your excellent series of recent posts and the sanity that you inject!

I have not been posting much of late here. The reason being that I am still getting settled in my new digs and new situation ... and trying to get caught up with the posts ... whew ...

Anyway ... I'm still around (in case anyone wanted to know).

Barry

Actually ... Dr. Chrakrapani addresses this issue in The Money Spinner. His contention is that if everybody used The Money Spinner (i.e. essentially AIM) then volatility would not only not decrease but would probably increase.

I've never been sure about that point of view but at least it illustrates that the matter is not cut and dried.

Barry

Why then would a life insurance company insure against your death because they know properly invested your payments will over time more than cover the final payment.

The short answer is that Life insurers do it to make money! One of the ways has to do with interest spread.

If you pay regular premiums for whole life a large portion (not all) of that premium goes into a kind of savings fund on which the insurance company pays a guaranteed rate of interest. They in turn invest that money and earn a higher rate. A portion of your premium goes to pay for term insurance for the amount of your coverage. This charge goes up every year as you age. Eventually the premium that you pay is not enough to pay for this term insurance. At this point the insurance company starts taking money from this savings fund to make up the difference. The level premium you pay is calculated so that the fund will run out on your last day of life.

The existence of this fund has lead some advisors to say buy term insurance instead of whole life and invest the difference and you'll be better off. The situation is of course more complex than this but the above gives a brief description.

Barry

Hi Conrad,

you wrote:

their house could be destroyed right away, and it would ruin those people(having to pay the mortgage and no longer having a house).

This is not technically true. One can always walk away from a mortgage but one would lose the house because it has been pledged as security on the mortgage loan. In this case the house is already lost and so one would no doubt walk away from the mortgage.

On the similarities and differences ... I would say the following:

(1) GAMBLING, INSURANCE, and SPECULATION are all similar in that they all involve changing the money one holds based on a random (i.e. chance) event.

(2) GAMBLING and SPECULATION are similar to each other and different from INSURANCE in that there is no profit motive in INSURANCE. There is an existing risk and INSURANCE is used to mitigate the adverse financial effects of the risk. In fact to try to use INSURANCE for profit (e.g. intentionally burning down an insured house) is illegal. (By the way ... what is the difference between "illegal" and "unlawful" ... answer ... one is against the law and the other is a sick bird ;=)

(3) SPECULATION and INSURANCE are similar in that they each relate to a real risk which exist quite independently of wether the individual chooses to buy insurance or engage in speculation. By contrast in GAMBLING the very act of GAMBLING creates a risk which did not exist before. It could be argued that the only difference between GAMBLING and SPECULATION is the type of risk.

INVESTMENT is concerned with long term profit and ideally seeks to avoid risk. If risk is encountered (as is frequently the case) the investor must be adequately compensated for this risk. The investor may take on risk but this is not his goal. The speculator primarily seeks risk on which he can be dramatically compensated. In practice however there is a little speculator in every investor.

Barry

This lottery type question is an interesting one! Buying a lottery ticket may at first seem to be illogical. Why?

Well no matter how large the prize, the probability of winning it is so small that, on average, every time one buys a one dollar lottery ticket one is losing about 99 cents (possibly more). In other words one's mathematical expectation is (probability of winning times the prize value, say one million for example) minus (the probability of losing times the cost of the ticket, say one dollar for example). If this difference (i.e. the expectation) is negative it is arguably not logical to buy such a ticket. Yet people (including me) often do so. Why?

Let's look at another case which is in a certain sense the opposite. Suppose I own a house worth $100,000 and suppose that there is a probability one tenth of one percent that it will be completely destroyed during the coming year. Now suppose that I can buy insurance for $200 per year that will protect me against this loss. Should I buy this insurance?

In a sense buying the insurance is also illogical. Why? Well on average if don't buy the insurance I will lose $100 per year (i.e. the value of the house times the probability that it will be destroyed). On the other hand if I buy the insurance my loss per year is fixed at $200. Thus buying the insurance on average is more expensive. Phrased another way my expected loss per year is $100 without insurance and $200 if I do buy insurance. It is illogical, one could argue, to choose the more expensive alternative.

Why then would one buy the insurance as indeed most (if not all) of us would? The answer has to do not with the expected value of the actual cash outcome but with the expected value of the expected resulting utility. My current wealth has a certain "utility" (there is a whole mathematical theory about this). If I buy the insurance then it is certain that my new wealth will be the current wealth less $200. For most people
there would be very little difference in the utility of these two positions. If I remain uninsured the my expected utility is 99.9% times the utility of my current wealth plus 0.1% times the utility of my current wealth reduced by $100,000. For most of us this utility is significantly less than that of our current wealth. So much so in fact that the expected utility of buying insurance is much greater than that of not buying insurance.

This is why buying one lottery ticket may not be illogical. The reduction in the utility of our current wealth caused by buying a one dollar lottery ticket is insignificant compared to the dramatic increase in the utility owing to having an additional one million dollars say. Thus buying a lottery ticket has a positive expectation of utility and therefore can be considered logical. Of course buying too many lottery tickets could result in negative expected utility in some cases. For example a welfare family that spends money for lottery tickets rather than food.

Anyway, this digression on Utility theory was prompted by the lottery discussion and while it has little to do with the topic of this forum I hope it was interesting. Thus does lead to an interesting question which I have struggled with over the years. I ask it now of any readers. I ask it as a rhetorical question only to be answered for yourselves only.

What are the differences and similarities among: GAMBLING, INSURANCE, SPECULATION, and INVESTING?

Barry

Maybe you can explain how you use options.

Sorry if I wasn't clear ...

I was referring to one thing and one thing only ...

selling covered calls at a strike price equal to the price at which AIM would direct the next sale. This is really a substitute for a GTC sell order. One could perhaps call them GTE (good till expired) orders ;=)

PUTs are not used. I have thought about also using PUTs in some fashion (likely on the buy side) but as yet I have no strategy that I am happy with.

The question was raised (or implied) what would I do if an AIM buy were triggered while the call option was still outstanding?

In Theory ... I would calculate the new AIM directed sell (and buy) prices after this most recent buy. I would then sell a new call for this new lower sell price while at the same time cancelling the old call by buying it back. Of course this might involve fees which were not justified. (e.g. there is almost certainly no point in buying back a call option that will expire the next day) Therefore ...

In Practice ... I would study the actual situation and act accordingly. For example if the old call was very near its expiry date I might take no action until it expired or prices made a drastic upward move. I might or might not sell a call before the old had expired. My actions would depend both the costs (i.e. fees etc.) and risks.

Barry

There are a few places where options (used as an adjunct to AIM) is not gambling but provides some extra profit potential.

For example suppose that your next sell point would be "sell 100 shares at $10" ... you could sell a call option for 100 of the stock at a strike price of $10. What can happen now?

(1) The option expires without being exercised and you pocket the amount you received for selling the option.

(2) The option is exercised and you end up both pocketing the money you received for the option and you have effected an AIM directed sell.

The only downside to (2) [and this is primarily psychological] is that at the option was exercised the stock price was higher than $10. If you had had an AIM directed sell at this point you would have pocketed more from the sale. So what!!! How often will this happen vs the number of sold options that are not exercised?

I think selling call options on stocks you own at a strike price equal to your next AIM directed sale price is an excellent risk free method of increasing your profits.

Barry

Charlie S3 wrote:

I'm all in cash at the 90 degree peak

This is incorrect.

AIM will never have you sell all of your holdings no matter how high the price goes.

By contrast one can run out of cash on long price declines.
Thus AIM is not symmetric!

Also it could be observed that equal buy and sell resistances (e.g. 10%) result in Lichello bands around the portfolio control and these bands are not symmetric.

Barry

jibes wrote Or you could try my new AIM RE-Bal method

I have reviewed the AIM RE-Bal method web pages and have a few comments and observations.

Firstly, this method is not new at all. It is discussed by Lichello.

Secondly, Bernie Goldberg gave a presentation on this type of approach at AIM 2001.

Thirdly, this approach can out perform AIMing the stocks individually but it also can significantly under perform as well. This depends on the correlations of the stocks involved. AIM derives much of its profit from volatility. The more stocks you AIM in one "basket" in general the less volatility. (AIM RE-Bal advocates AIMing several stocks, say four, and periodically re-balancing them as well as following AIM directed transactions). Great care must be taken so that the stocks used are positively correlated so as to increase volatility otherwise there will be fewer AIM directed transactions than if the stocks were AIMed individually.

One complication to this approach is the question of frequency of re-balancing in relation to frequency of actual AIM transactions. For example, one will of course always re-balance when there is an AIM directed transaction ... but ... should one also re-balance even if there was no AIM directed transaction? Thus there are two different approaches and hence two different potential flavours of AIM RE-Bal.

In order to get a fair comparison one should compare the results of AIM RE-Bal with the overall results of AIMing the stocks involved individually using the same AIM parameters. In addition both flavours of AIM RE-Bal should be presented. As the web site is fleshed out I recommend including such a comparison.

Fourthly, there is an error at the Web Site. It is stated that First, you don't use SHARES and PRICE as usual in your AIM formula

This is not usual at all. It is Market Value of the basket of securities that is used in the AIM formula as is clearly shown in Lichello's book. Of course if the basket happens to contain only one stock then SHARES and PRICE together is a mathematically equivalent way of looking at market value but this notwithstanding, it is market value that is part of the AIM formula. Of course jibes correctly employs this. The statement is only a minor error and is in fact a pedagogically sound approach to use. I only added this comment because new comers to AIM who viewed this site might have been mislead.

Barry

Descartes, after a hard day at the office, stopped off at his favourite drinking spot (a lovely little Paris cafe) for a cognac.

When he had finished his cognac the waiter asked.

"Would you care for another cognac, Mr. Descarte?"

Descarte pondered for a moment and replied:

"I think not!"

And then puff ... disappeared

Barry

The only downside I see to using one stock would be if the stock went down and never recovered.

This is not necessarily correct!

A good AIM stock has two essential characteristics:

(1) It has high price volatility ... the higher the better in fact because this volatility is exactly what AIM captures and turns into profits which can be far in excess of buy and hold.

(2) It has sound fundamentals ... because AIM is a long term strategy.

If a stock that one is AIMing loses one or both of these characteristics then one should get out of that stock and start AIMing another. This should be done even if the stock price is at an all time high and is expected to go even higher.

By contrast, even if a stock has tumbled a great deal in price and remains forever hovering around that new much lower price ... one should still continue to AIM it if that stock still retains those two essentials.

Of course if a stock has fallen significantly that might be an indication that the fundamentals have deteriorated. Thus after such a fall one should check the underlying fundamentals anew ... but if they prove to remain sound then there is no reason to panic. Even if the price levels never recover ... if the volatility remains great the stock could continue to reap huge long term profits if it continues to be AIMed.

Barry

Hi Conrad,

I am originally from Ottawa, Ontario, Canada and while I spent a good part of my early life there, I also spent many formative years in Geneva (Swiss), Chicago, and Tokyo. More recently I lived and worked in Osaka, Seoul. Sao Paulo, and New York. I recently moved back to Canada and am living in Toronto.

My basic thinking has been significantly influenced by The Money Spinner. By that I mean the following:

AIM (and any related systems such as The Money Spinner or Vortex AIM) could be described either as a "Risk Management System" or as a "Volatility Capture System". It is in fact both but I tend to think of it primarily as the latter because this is the aspect that was emphasized in the Money Spinner. It seems that those who first read Lichello think of the system primarily as the former. Both sides are equally correct. It is merely a matter personal style I think.

I originally used the Money Spinner but never really trusted it because I did not understand it. Oh, I understood the mechanics well enough but not the underlying system. For this reason among others I stopped investing for a while.

When I began again I started with the Money Spinner but in an attempt to understand it I "reverse engineered" it to produce an alternative formulation of the money spinner. I then followed this reformulated system. It was shortly after this that I discovered Lichello's AIM which turned out to be almost identical to my reformulation of the Money Spinner. For this reason I claim that the Money Spinner is basically 99% AIM. The main difference is the minimum buy and sell intervals.

Best Regards,

Barry

Anyone out there familiar with it? Barry?

Sorry Steve,

I've never even heard of it.

Barry

While this topic is perhaps off topic I hope that coffee drinkers and others may find it of interest.

Conrad wrote: On Partial Fills it is so that in Canada and the States they keep filling up you coffee mug with partials fills ad infinitum, or at least as long as you do not object to this free service(usually the coffee is pretty weak, but that is a different story).

In my experience the coffee in Asia (at least in Corea and Japan) as well as in Europe and Latin America is strong and good. By contrast it is very hard to find decent coffee in the US or Canada. I have a theory about this.

Please note: In general I use the pre 1938 spelling (i.e. Corea) of Korea. This spelling was changed, I heard, because during the 1938 olympics the teams came out in English alphabetical order. Corea was then under the control of Japan and so Japan officially changed Corea's official spelling so that Japan would precede Korea rather than follow Corea. It should also be observed that French for example still uses C rather than K.

Canada was and still is more of a tea drinking society than is the US so I suspect that Canadian coffee (generally weak and insipid like its US counterpart) style was influenced by the US.

The US was originally a tea drinking nation but England's unfair taxation on tea (which also was a partial cause of the revolution) caused Americans to seek tea substitutes such as coffee. My theory is (and was) that because of this history, American's do not really drink coffee but rather they (and therefore Canadians too) drink coffee flavoured tea because they never really ceased being tea drinkers. They just changed the flavour of their tea! In saying all of this I must confess that while I do enjoy good coffee I much prefer tea. In my experience it is generally very hard to get a good cup of tea in both American and Canadian restaurants. By contrast in England and Scotland is is very difficult to not get a good cup of tea.

The above is background on my theory. The point of this message is to share with you an interesting fact that I discovered a few years ago while living in Brazil. The Brazilians it seems agree with my theory and in fact developed it independently. The portuguese word for tea is cha (pronounced SHAW) and coffee is CAFE (pronounced KAH FAY). Brazilians frequently refer to north american style coffee as ... SHAW FAY

Barry

I performed some ROCAR calculations on Tom's actual STKL activity history. Here is a table of the results with some discussion following the table.

1997/06/01 N/A N/A N/A
1997/06/11 _99.66% _99.66% 0.00%
1997/09/24 221.03% 208.63% 12.40%
1997/10/20 32.02% _0.41% 32.43%
1997/12/03 _3.47% _26.46% 22.99%
1997/12/24 112.26% 66.01% 46.25%
1998/07/31 _5.11% _18.84% 13.73%
1998/08/21 _7.26% _20.06% 12.80%
1998/09/14 _11.91% _23.09% 11.18%
1998/11/23 _22.56% _29.65% 7.09%
1999/01/06 _36.75% _37.17% 0.42%
1999/01/22 25.67% _3.02% 28.69%
1999/05/11 _5.08% _18.91% 13.83%
1999/07/27 7.69% _9.41% 17.10%
1999/08/02 11.16% _6.93% 18.09%
1999/08/17 17.34% _2.25% 19.59%
1999/08/25 20.02% _0.07% 20.09%
1999/10/22 _0.67% _15.65% 14.98%
1999/11/18 _3.05% _17.09% 14.04%
1999/12/07 _7.93% _19.98% 12.05%
2000/01/19 17.25% _3.78% 21.03%
2000/02/09 22.07% _0.06% 22.13%
2000/02/10 24.41% 1.68% 22.73%
2000/02/11 28.69% 5.01% 23.68%
2000/02/15 34.32% 9.65% 24.67%
2000/02/16 37.55% 12.57% 24.98%
2000/02/17 42.63% 17.38% 25.25%
2000/02/22 48.20% 23.00% 25.20%
2000/07/19 23.40% _1.60% 25.00%
2000/07/26 21.89% _3.18% 25.07%
2000/08/09 20.16% _4.77% 24.93%
2000/09/05 32.49% 7.95% 24.54%
2000/09/06 34.33% 10.91% 23.42%
2000/10/18 21.10% _2.96% 24.06%
2000/11/16 33.79% 10.81% 22.98%
2000/12/20 22.40% _1.41% 23.81%
2000/12/26 21.04% _2.81% 23.85%
2001/05/07 31.86% 10.26% 21.60%
2001/06/05 35.35% 14.34% 21.01%
2001/06/06 36.99% 16.18% 20.81%
2001/08/10 25.89% 3.20% 22.69%
2002/01/22 31.85% 12.27% 19.58%
2002/03/22 32.47% 13.64% 18.83%

The above table displays messed up (I don't know how to fix it) although it is still readable. After each date (other than the first) are three percentage rates. The first is the ROCAR of all AIM trades up to that point. The second is the return on Buy and Hold. The third is the AIM ROCAR minus the Buy and Hold return.

All of these rates are effective annual rates of interest. It is interesting to note that up to 1997/06/11 both returns are the same as expected since up to this point AIM has in effect been Buy and Hold. After that AIM is always ahead of Buy and Hold as measure by rates of return.

I think this result in very interesting and speaks a lot about AIM's ability to capture volatility and turn it into return.

Barry

Not Yet,

I have just returned from a short offline time. I'm trying to catch up with the posts here. Then I'll finish my simulation runs. Part 4 should be out in 7 to 10 days.

Barry

Are you heading in the general direction of Tom's Vealie?

No not really.

A Vealie is only implemented at the time of an AIM directed sell. It is a way of avoiding increasing cash but is never used to decrease cash.

What I am talking about would take place at specific periodic intervals whether or not there were an AIM directed trade at that point. Also it could be used sometimes to increase cash and sometimes to decrease cash.

Again, I am not necessarily promoting such a technique. I am only seeking people's reactions to the idea. Thanks for yours.

Barry

Status Update

Hi everyone. I just wanted to let everyone know that I will be away from the computer for five days. Please keep the number of posts down during that period? I don't want to return to find more than two or three thousand posts to catch up with.

Bye For Now,

Barry

A General Question

I am curious. What do members of this board think about and/or do about the AIM variation that I will describe below. I'll first describe it and then offer a few of my own comments. Please let me know your thoughts?

Description:

Whenever we start a new AIM account we decide how much of our total investment will go into the stock or mutual fund and how much will go into cash. Lichello's various editions set various splits from 50/50 to 80/20 in the latest edition. Tom Veale's Idiot Wave (Thank you Tom) in effect provides a weekly split which could be in some way the optimal starting split at that time point. OK ... FINE ... BUT!

Suppose I started an AIM account with some stock (ACME WIDGETS say) one year ago with a 50/50 split following the original Lichello settings. Its now one year later and my total portfolio value is now $20,000 instead of the original $10,000 that I invested. But let's say that the current stock value is $15,000 and cash is $5,000.

Now let's say that you have $20,000 with which you want to start an AIM account also with ACME WIDGETS. You follow Lichello's standard settings just as I did and so you use a 50/50 split.

The situation is now that each of us has an AIM account worth $20,000 at the same time but my split is 75/25 and your split is 50/50. In answer to the questions ... which one of us is right? at first glance it seems to me that the answer must be at least one of us is wrong.

Assuming that you and Lichello are correct then perhaps I should make some adjustment outside of the normal AIM transactions and rebalance my stock to cash ratio.

Finally my question ...

Should one periodically rebalance one's AIM account and if so ... what technique should be used and how frequently should the rebalancing be performed?

My Comments:

If rebalancing is appropriate then the frequency of rebalancing should depend upon one's normal AIM update frequency. For example if I normally update monthly and if therefore I rebalance yearly then you should probably rebalance more frequently if your update period is weekly.

One important aspect of the ACME WIDGETS example above is that even though the total portfolio values are the same, the portfolio control values will be different. Therefore in my opinion any rebalancing should probably also result in an adjustment to portfolio control.

One final note:

In his AIM book "I Guarantee You Will Buy Low Sell High and Make Money", Jeffrey Weber advocates monthly updating with a yearly rebalancing. If this topic is of interest to you then you may find Jeffrey's book interesting.

Barry

RE: ROCAR

Hi Tom,

I have been meaning to do a thorough analysis of your ROCAR formula. I promise to do this soon although I may have to bother you with some questions about it.

I intuitively feel that your formula will give a good approximation to my formula but I need to test this.

It seems to me that your E (i.e. equity value at risk in each period) is based on the actual market value of the stock held. My calculation never uses the market value except at the very end and the very beginning of the period. All other equity values are implicitly the accumulated value of all cash flows into and out of the stock with the ROCAR rate being the rate of accumulation.

More important, my calculation takes into account the exact timing and amount of each cash flow. The actual amount of cash plays no role in the calculation of my ROCAR. This is why I envision another measure of the average amount at risk but this would be determined after the ROCAR has been calculated.

For example suppose that at the beginning of one period (a month say) I start with $1,000 of stock. This could either have been the initial purchase of my AIM stock or it could simply be the current market value of that stock at the start of this period. It doesn't matter. I have simply decided to measure ROCAR over a period beginning at this point. Thus in theory I sell all of the stock and then rebuy at the current market value (with no commissions). This then becomes my initial investment. Now let's say that next month the market value has risen enough so that AIM directs me to sell $200 at then current market price. The month after that AIM directs me to buy $300 of stock at the then current market price. Finally next month the the total market value of my holdings happens to be $1,389.

I would calculate ROCAR this way. First I would construct the following cash flow.

At time zero I deposited $1,000 into an account paying a constant but unknown rate of interest which could be zero.

At time one I withdrew $200.

At time two I deposited another $300.

Finally at time three I discover that the current balance in the account is $1,389. Obviously my net deposit into the account was $1,100 (i.e. = 1,000 - 200 + 300). Thus a total of $289 interest was credited. But what is the rate of interest credited?

I calculate this as follows:

I assume there is some monthly rate of interest called i.

At time three the 1,000 deposited at time zero has increased to 1000 * (1+i)^3

At time three the $300 deposited at time two has increased to 300 * (1+i)

Finally at time three the $200 removed from the account at time one would have accumulated to 200 * (1+i)^2

If we add up these three pieces (actual we subtract the last) we get the total

1000*(1+i)^3 + 300*(1+i) - 200*(1+i)^2

but we know that this total is equal to $1,389 so I just equate the two and solve for i.

In this case we get a value of i of 10%. I then convert this to an annual rate giving

1.1^12 - 1 = 213.84%

If these transactions had been weekly then the annual rate would have been

1.1^52 - 1 = 14,104.29%

If these transactions had been quarterly then the annual rate would have been

1.1^4 - 1 = 46.41%

If these transactions had been semi-annual then the annual rate would have been

1.1^2 - 1 = 21.00%

Finally if these transactions had been annual then the annual rate is in fact 10%

In practice there can be very many cash flows at a variety of fractional times over a year. The calculation would be the same as above but more complicated. In particular if there are n separate cash flows then we have created an n degree polynomial in the unknown i. If n is 5 or more (as will be the case with any reasonable AIM period) then it is a mathematical fact that the roots of the polynomial cannot be solved directly. Thus we need to use an iterative process to find the roots.

Each root of the polynomial is a value of ROCAR. We hope that the polynomial in question will have exactly one real (as opposed to complex) root. If there is none ROCAR does not exist. If there is more than one we have multiple ROCAR's. Neither of these situations is tolerable.

I believe, though I have not yet been able to prove (but I'm sure I will some day), that any such polynomial arising from AIM transactions does have one root only.

As part of my calculations I examine the resulting polynomial to determine its root structure. In theory the IRR routine in EXCEL should be able to calculate ROCAR in this way. In practice EXCEL's IRR routine doesn't check for multiple roots. I have seen it (in non AIM contexts) calculate incorrectly because it gets the wrong root. That's one reason, among many, why I never use spreadsheets.

Well, this has gone on too long. I hope it answers your question.

Take Care,

Barry

Hi Conrad,

I ran some simulations on your test data using the assumptions I mentioned in my other post. I ran three simulations. The three were almost identical in that they all used 10% buy and sell safes and added 50% of the amount purchased to portfolio control (all standard AIM). I also started each simulation by purchasing exactly 100 shares.

The three systems were SA, HA, and TA. Under SA I set minimum transaction sizes of 5% of the then current total market value. Under HA and TA I followed standard AIM with no regard to minimum size transactions. I then adjusted the SA derived transaction to the size specified by HA or TA. I then applied the same minimum size transaction to the result (i.e. 5% of the then current market value).

Now the results:

SA:
total cash = 9,308.77
number of shares = 150
total investment value = 24,308.77
total interest received = 608.56
total trading costs = 594.79
total number of transactions = 20 (11 buys / 9 sells)
final portfolio control = 15,792.50
annual ROI = 12.84%
annual ROCAR = 20.33%

HA:
total cash = 9,149.83
number of shares = 150
total investment value = 24,149.83
total interest received = 601.69
total trading costs = 636.86
total number of transactions = 22 (13 buys / 9 sells)
final portfolio control = 15,906.50
annual ROI = 12.38%
annual ROCAR = 19.75%

TA:
total cash = 7,766.75
number of shares = 166
total investment value = 24,366.75
total interest received = 505.43
total trading costs = 638.68
total number of transactions = 21 (10 buys / 11 sells)
final portfolio control = 17,136.00
annual ROI = 13.00%
annual ROCAR = 19.49%

The above results are very interesting. Despite what I think are superior esthetic features of HA and especially TA over SA it is clear that SA has performed the best in this case.

WHY?

Because of both its superior ROCAR and the fact that maintains a higher cash position than the other two systems. It is true that TA has a slightly higher ROI (i.e. 0.16% or less than 2 basis points higher) but TA's ROCAR is 0.84% or more than 8 basis points lower.In addition TA has more cash at risk.

Of course with other parameters TA might have outperformed SA but not in this case.

Barry

AIM Testing / Some Questions And Comments

Hi Conrad,

I am about to run several simulations using your proposed test case. However there is some ambiguity in the definition of the test which I need to clear up. In addition I have a major observation about the possible results. I'll deal with all of these in numbered points.

(1) Trading Cost --- are the following calculations correct?

If one buys 100 shares at 100 per share then the total trading cost is 20 + 100*100*0.006 = 80

If one buys 99 shares at 100 per share then the total trading cost is 20 + 99*100*0.006 = 79.40

(2) Whenever shares are purchased or sold can fractional shares be involved? For example if my formula directed me to sell 100 worth of stock whose current price per share were 8 then this would indicate a sale of 12 and one half shares.

I assume that only a whole number of shares can be traded at any one time. Please specify if you want some other assumption.

(3) What exactly do you mean by Start-out cash at 50% exactly?

Assuming that the trading cost calculation above is correct and that only whole numbers of shares can be used then it is mathematically impossible to start out with 50% cash!

Either I buy 100 shares and thus end up with 10,000 in stock and 9,920 in cash (after trading cost) or I buy 99 shares and thus end up with 9,900 in stock and 10,020.60 (again after trading cost). In neither case is cash at 50%. I assume that you mean the first purchase is for 100 shares and that the 50% cash is measured before the payment of initial trading cost. Please confirm this?

(4) How should rounding be handled? For example each week interest needs to be calculated and credited to cash. Almost always the interest amount will involve a fraction of a cent. Should the interest amount calculated always be truncated, always rounded to the next higher penny, or rounded in some way and if so how? In addition the trading cost calculation might end up with a fraction of a cent. What rounding should be done in this case?

Also using the share example above there is a question of how the number of shares should be rounded. However I don't think you need to address this question. I view the rounding of shares method to be part of the AIM method being tested.

(5) You seem to imply that ROI (Return On Investment) will be the sole criteria for comparing two methods. I believe that ROCAR is also of fundamental importance and should be included in any comparison of the efficacy of two methods.

For example, if my TA produced an overall return of 20% and my HA only returned 10% one might at first think that TA was better than HA. But if at the same time TA's ROCAR were 23.5% and HA's ROCAR were 30% what would imply? First it implies that HA was much better at capturing the stock's underlying volatility. Secondly it means that on average TA had 80% of the total invested in stocks whereas HA on average only had 20% invested in stocks. Thus TA is four times more risky than HA. Therefore in this case it's not clear which method is better.

One must ask one's self the question ... is the higher overall gain sufficient payment for the extra risk taken? Even if both of TA's overall return and ROCAR were better than HA's it is still not clear that TA is better (although it is much more likely). We really need to consider a third item and that is the average percentage at risk (i.e. in the stock) of the total portfolio value (i.e. cash plus stock).

I am currently working on producing a simple yet accurate formula to approximate this average percentage. My current formula, which is exact, perforce uses integral calculus and the previously calculated ROCAR (which in turn uses advanced numerical analytic techniques) in its determination. Therefore I am not at this point able to provide a simple formula.

In any event I suggest that we use total return on investment plus ROCAR to compare test results since using total return alone does not take into account the whole risk/return relationship.

Take Care,

Barry

Tom wrote:

as someone mentioned ROCAR is really ROACAR. Return On Average Capital At Risk. It's the average of each period's percent at risk.

I don't agree!

The way I calculate ROCAR there is no average amount used. The capital at risk is always the total actual amount at risk at each microscopic instance because I use calculus to determine the value.

Your formula, which is a very simple and good approximation, does indeed use an avarage of sorts but the true ROCAR which it is approximating is actual not average.

Therefore if ROACAR were to mean "return on actual capital at risk" ... then I would claim it to be correct but redundant ... however "return on average capital at risk" ... is incorrect IMHO.

Take Care,

Barry

Conrad,

Thanks for your message. There are a few points that I want to respond to:

(1) I question if it is not simpler to only use the SAFE only as a Filter to decide: To Buy or Not to Buy

One of the "problems" with Standard AIM is that it uses two things (i.e. portfolio control and buy safe) to determine how much to buy and it uses three things (i.e. portfolio control, buy safe, and update frequency) to determine when to buy. This of and by itself is not really the problem. The problem lies in the fact that two of these things are used for both purposes meaning that these two functions are not really independent of one another.

(2) Now for the Buy Mode in your Hybrid AIM ... This is, if I understand it correctly, identical to the Vortex Method

I have not yet fully explored Vortex AIM so I cannot be certain ... but I don't think this is correct. On the buy side TA is identical to VA (i.e. Vortex AIM) if one sets the buy safe (i.e. B) to zero. On the other hand HA is identical to SA on the buy side if B is zero.

VA, as I understand it, buys enough so that the new market value is equal to the new portfolio control. This is what TA does if B is zero. HA buys enough only to make market value equal to the portfolio control before the addition of one half the purchase amount (again assuming B is zero).

(3)Test Case

Good Idea. I will include simulations from your Test Case in part 4 of my Tangent AIM series.

Best Regards,

Barry

Conrad wrote:

So, we can conclude that for Selling there is essentially no difference between the Tangent Method and the Lichello Method?

I think the answer here is NO .

Let's use the same example.

Portfolio Control (F) is $9,000
Starting Market Value (M) is $12,000
Sell Safe (S) is 10%

Now under normal Lichello AIM (assuming no minimum transaction size) we first note that M is greater than F so we have a sell advice of 12,000-9,000 = $3,000.

Now the SAFE in this case is 10% of $12,000 or $1,200. Thus the Lichello sell order would be 3,000-1,200 = $1,800.

After this sale F is still $9,000 and M is now 12,000-1,800 = $10,200.

Now normally Lichello AIM would do nothing more at this time point. One month later (or one week, one day, one hour, depending upon one's normal update frequency) let's assume that the market value has not changed. Then we would go through the same procedure as above.

Because M is greater than F we have a sell advice of 10,200-9,000 = $1,200. Here SAFE is 10% of 10,200 or $1,020. Thus the sell order is 1,200-1,040 = $160. Now this sale would probably not be made because $160 is probably less than our normal minimum transaction size but because we have set no minimum we end up with M = 10,200-160 = $10,040.

Now we wait another month and again assume that the market has not changed. Standard Lichello AIM (I'll ignore the calculations) will issue a sale order of 36. Each month thereafter a new sell order will be generated, each one smaller than the last. Assuming that we could actually perform transactions for fractions of a penny this process will never end. If one totals up this series of payments (there are an infinite number but this series does converge) we would get a total of $2,000.

Now under the above same conditions (i.e. monthly updates together with the market never changing) Tangent AIM would sell the same total amount as Lichello AIM but it would sell that total amount at the first checkup point and then never sell again.

Thus there are several differences:

(1) TA would generate fewer commissions than SA

(2) TA would have more money in cash than SA earning interest for longer

(3) If market conditions should in fact change between checkup points (as they almost surely will) then there is no comparison.

Note that points (1) and (2) imply that TA is better (in fact much better) than SA. However, point (3) is to me clear evidence that we cannot say which one is better. In practice I only use strict AIM by the book. I do not use TA. I am presenting it here only for academic interest.

So my answer to your query is ... Despite the fact that TA and SA could ultimately generate the same total sell amount (ignoring minimum transaction sizes) they are otherwise very different indeed.

Best Regards,

Barry

TANGENT AIM ... part 3 of 4

First let's recall from part 2 that I am using:

SA to represent Standard AIM

TA to represent Tangent AIM

and I will also introduce Hybrid AIM and use

HA to represent Hybrid AIM.

We will use the following symbols in this part:

M = the total market value of our stock before the purchase

F = the portfolio control before the purchase

X = the amount of market value purchased based on SA

Y = the amount of market value purchased based on HA

Z = the amount of market value purchased based on TA

N = the market value of the stock after the purchase

G = the portfolio control after the purchase

B = the buy resistance or BUY SAFE

A = the adjustment factor to the portfolio control

Now looking at SA first we have that M is less than F/(1+B) because this is a purchase.

SA says:

X = F - M*(1+B)

and

G = F + A*X

where normally A is equal to one half (i.e. 0.5)

Let's look at an example.

Let F=$9,240 and M=$7,260 and B=0.1 (i.e. 10%) and A=1/2

Then F/(1+B) is $8,400 and M is clearly less than this so we have a purchase.

SA would have us purchase X = 9240 - 7260*1.1 = $1,254

This would result in N = M + X = 7,260 + 1,254 = $8,514

and G = F + X/2 = 9,240 + 1,254/2 = $8,970

Let's analyse this. First of all we see that N is greater than $8,400 (i.e. F/(1+B)) thus is a sense we purchased more than we had to when we consider F. Of course G/(1+B) is $8,970 which is greater than N and so in a sense we have not purchased enough when we consider G. We can prove that this kind of result will always be true although we shall not do so here.

We will now try to calculate a new purchase amount Y so that N will be "tangent" to the lower Lichello band of F (i.e. F/(1+B)). As in part 2 for the new sell amount we can set up the equation

M+Y=F/(1+B)

and solve for Y which gives

Y=X/(1+B)

which is less than X (assuming that B is positive).

Returning to the above example we have

Y = 1,254 * 10/11 = $1,140

N = 7,260 + 1,140 = $8,400

and clearly N = F/(1+B)

and

G = F + Y/2 = 9,240 + 1,140/2 = $9,810

and G/(1+B) = $8,918.18

so clearly although Y is perfect from the F point of view, we still need to buy more from the G point of view. Let's refer to this variation as HA. That is HA is SA on the sell side but on the buy side HA buys at the same time as SA but only buys an amount of 10/11 times the SA amount (assuming a standard 10% buy safe).

Now let's look at another variation.

We wish to purchase Z so that the new market value N is tangent to the lower Lichello band of G. We can calculate this by writing the following equation:

N=G/(1+B)

or

M+Z=(F+A*Z)/(1+B)

and solve for Z. When we perform the algebraic manipulations we find

Z=X/(1+B-A)

Thus using SA settings of B=10% and A=50% we get

Z=X*5/3

Returning to our example we get:

Z = 1,254 * 5/3 = $2,090

N = M + Z = 7,260 + 2,090 = $9,350

G = F * Z/2 = 9,240 + 2,090/2 = $10,285

and G/(1+B) = 10,285 * 10/11 = $9,350 which is exactly N

Thus a TA buy takes place at the same time as an SA buy the purchase amount is 5/3 times the SA amount.

In part 4 we will summarize all of this and present some simulation results.

Barry

TANGENT AIM ... part 2 of 4

We will use the following symbols in this part:

M = the total market value of our stock before the sale

F = the portfolio control before the sale

X = the amount of market value sold based on Standard AIM

N = the market value of the stock after the sale

G = the portfolio control after the sale

S = the sell resistance or SELL SAFE

Y = the amount of market value sold based on Tangent AIM

First we note that M is greater than F/(1-S) because this is a sale.

We also note that:

N = M - X

G = F

and finally according to Standard AIM

X = (M*(1-S)) - F

where * indicates multiplication

clearly X is greater than zero because M is greater than F/(1-S)

now if we calculate N we get

N = F + M*S

and since M is greater than F/(1-S) we get that

N = F + M*S is greater than F + (F*S)/(1-S) = F/(1-S)

Thus since N is greater than F/(1-S) we in theory have another Standard AIM sell. Of course Standard AIM would have you wait one month (or some other period depending upon your update frequency) before making this next sale (assuming nothing happened in the interim). After this next sell the resulting market value will still be greater than F/(1-S). Thus this implies an infinite number of sells each of which is smaller than the last. This series of sales could be added up because, as Conrad quite correctly observed in his development of Vortex AIM, this series does converge. However we can arrive at the sum in a more direct manner.

Under Tangent AIM I want to sell an amount Y such that the new resulting market value will be exactly equal to F/(1-S) so that no further sale is generated. Therefore let us set

(M-Y)=F/(1-S) and solve for Y giving

Y = (M*(1-S)-F)/(1-S) which is simply

Y = X/(1-S)

Thus the first difference between Standard AIM (from now on SA) and Tangent AIM (from now on TA ... not Technical Analysis) is that even though SA and TA produce a sell at the same time, TA will produce a larger amount of sale. For example if S is 10% then Y is about 111% of X.

Let's look at a specific numerical example. Suppose that the current market value were $12,000 and that the portfolio control were $9,000. Also assume that our sell resistance is 10%. Thus we have:

M = 12,000
F = 9,000
S = 0.1

First we see that F/(1-S) gives 10,000 and M is greater so that we do indeed have a sale.

Either using the above formula or the SA approach (i.e. sell advice plus safe adjustment) we find that

X = 1,800

N = M - X = 12,000 - 1,800 = 10,200 which implies another sale

But 1/(1-S) = 10/9 in SA ... thus

Y = X * 10/9 = 2,000

and

M - Y = 12,000 - 2,000 = 10,000 which implies no sale.

In summary then we can say that to use TA on the sell side simply calculate the sell amount using SA but don't consider minimum transaction size. Now increase this sell amount by multiplying it by 1/(1-S) [or alternatively by dividing by (1-S)]. If you are using standard safe settings of 10% you simply multiply by 10/9. Now of course you can consider minimum size. Of course if the SA amount was already large enough then so will the TA amount be but ... even if the SA amount did not meet your minimum requirement, the TA amount still could.

Let me conclude this section with two observations.

First, I based everything above on F. I could have however used G instead of F. Since G=F however, the results are the same. As we will see in part 3, this is not true of buys. In this case F and G are not the same. Of course we could further modify TA by adding to F a certain multiple of the sale amount (this multiple could be negative) so that F and G were different. Then I would find a third sell amount Z which would be determined in relationship to G. I won't do this here but will do it in part 3 when we consider buys.

Second, in geometry when two objects just touch but do not intersect we say that they are tangent to each other. The sell amount in TA is determined so that the resulting market value is just equal to the top Lichello Band or in other words is tangent to the top Lichello Band. Similarly after a buy TA sets the resulting market value tangent to the bottom Lichello Band. It is for this reason I use the term "Tangent AIM".

Barry

Conrad,

I just had the same kind of problem. In my most recent post "Tangent AIM part 1" I used both > and < signs.

In the preview I notices that the > sign kept disappearing. Perhaps this was because the expression (1-S) kept showing up as a spelling error whereas (1+B) did not. Each time I returned to the data entry screen the > sign was gone. Finally after I posted the message the sign disappeared again. I went in to edit it within the alloted 15 minutes and was able to add it again,

Even in this message all of the > signs vanished. I had to add them after this message was published.

Very strange!!!
Barry

TANGENT AIM ... part 1 of 4

I remember how, at AIM 2000, I had a conversation with Bill McKinley, (and at least one other person whom I do not recall at the moment), in which I discussed a particular AIM variation that I had been working on. Bill seemed interested in the concept but I never really did much more work on the concept. These memories were again brought to mind by Conrad's so called Lichello Flaw and with Vortex AIM. My purpose with this series of posts (this is the first of four on this topic) is to present the fundamentals of this AIM variation. For the purposes of discussion let us call this particular variation TANGENT AIM.

Please note that I am not suggesting that Tangent AIM is necessarily any better than AIM by the book (in fact we will see an example in part 4 that proves that it is not better at least some of the time). Also I am not comparing Tangent AIM with Vortex AIM except to note that, like Vortex AIM, Tangent AIM gets rid of the so called Lichello Flaw.

Let's begin by describing standard AIM by the book:

We begin by looking at two current values:

M = the total market value of our stocks

F = the portfolio control value (i.e. before the current transaction if any)

Then armed with two other values:

S = the SELL SAFE or sell resistance

B = the BUY SAFE of buy resistance

we determine what if any action we take:

if M > F/(1-S) then we sell a portion of our holdings

if M < F/(1+B) then we buy some additional stock to increase our holdings

if M is any other value then we simply hold on to our position until the next checkup date.

These bands around F are what Tom has referred to as the Lichello Bands (with appropriate provisions to reflect trade size minimums of course). Notice that even if S equals B equals 10% the bands about F are not symmetrical. The range above F is 11.1% and below F is 9.1%. The total range is 20.2% of F. This suggests another possible AIM variation. Specifically, could one not set the Lichello Bands directly based on F rather than M. Moreover could they not be set to be symmetrical? This lack of symmetry about F found in standard AIM is still another "flaw". Tangent AIM does not touch on this issue however. I merely mention it as another point of interest.

The amounts that we buy and sell will be discussed in subsequent parts 3 and 2 respectively.

There are two other aspects to standard AIM. First, we do not actually buy or sell unless the amount calculated exceeds a certain minimum amount. For the purposes of this discussion we will assume that this minimum is zero. This is not to suggest that minimums are not important. Rather it is to simplify the presentation in order to make clear the essential aspects of Tangent AIM. While some mention of minimums will be made in subsequent parts it is clear that Tangent AIM, at least as presented here, is not a complete or final system. I present it here to you all merely as a matter of academic interest. Second, standard AIM has a provision for adjusting the portfolio control after certain (i.e. buy) transactions. This will be discussed in subsequent parts.

Part 2 will discuss sells while part 3 will discuss buys. Finally part 4 will contain the results of some simulations comparing standard AIM to Tangent AIM. By the way ... the reason why I am using the name Tangent will be explained in part 2.

Finally before concluding part 1, I want to once again stress that I do not necessarily advocate using Tangent AIM instead of Standard AIM nor any other AIM variant (e.g. Vortex AIM). I merely thought that this exposition might be of academic interest to some of the participants of this forum.

Barry

Thus we see that Risk Management is simply the recognition that different securities have different returns based on their probabilities (as perceived in the market place) of failure to meet obligations. There is (normally) in the market place an added bonus for assuming this risk. One can improve one's overall return (on average) by investing in these securities. But one should also diversify over the various risk levels (e.g. only put some of your money in my riskier GICs).

Up to this point I have portrayed Tom as similar to government T-bills (i.e. so-called risk free) but this is not in fact true for government T-bills or Tom (is that T a coincidence?). Tom and the government ... like I do ... each have an URN with one black marble. Of course the government's urn has millions and perhaps billions of white marbles but it still contains one black marble. There are no truly risk free investments but for practical purposes we can act as though T-bills are risk free.

Now for one final and interesting point. Once I have issued my GICs there could be speculation as to just how many marbles are in my urn. This will affect the price of my GICs in the secondary market. In fact suppose that I have three different urns: one as previously described, one with four black marbles and one white, on one with only one black marble but 10 white marbles. Further suppose that I change the urn that I use to pay off GICs and that I change the urns in an irregular basis. This will give my GICs a lot of fluctuation in the secondary market and will probably contribute to them being very good AIM candidates.

I hope this series of posts has cast some light on how risk is reflected in the market place.

Barry

As I mentioned in the previous posts, I believe that all investments are actually a combination of pure investment and pure gambling. In practice we only choose investments whose gambling portion is of type (c), at least as we perceive it. Of course we might purchase an investment whose gambling portion were of type (b) or even type (a) but we would do so only if we mistakingly believed them to be of type (c).

Let's imagine for a minute that Tom Veale was offering one year GICs with a return of 10% and par value of $100. Thus just invest $100 with him today and one year from now he will return to you $110. Sounds like a pretty good deal to me! Now let's also say that Tom is risk free so there is no doubt that Tom will make good on his financial obligations. We are certain that $100 today will be paid back at $110 in one year.

Now let's say that I am equally trust worthy as Tom and that I too am offering similar GICs except for two differences. First I am offering a rate of 20% rather than 10% (sounds good so far but ... ) and second I own a beautiful Grecian Urn (what's a Grecian Earn? ... I don't know ... about three drachma an hour? in which I store five beautiful marbles four of which are white and one of which is black. Each time one of my GIC's comes due I extract one marble. If I extract a white marble I will pay the full amount of $100 plus the full year's interest which in this case is $20. However if I extract the black marble I will pay you nothing. You have lost your total investment.

A careful analysis of this above deal will so that this is in fact a combination pure investment and a pure gamble of type (a). I won't go through the analysis here but one can do it for himself following the analysis that I will go through for my next example.

My second example is that I have failed to sell any of the above 20% GICs so I raise the return to 37.5%. Now who will buy? Well again nobody in their right mind. Why ... because this is a pure investment together with a pure gamble of type (b). If you were to buy 5 of these GICs form me then you would find:

0.032% of the time you would lose everything
0.640% of the time only one of these GICs would pay off
5.120% of the time exactly two of these GICs would pay off
20.480% of the time exactly three of these GICs would pay off
40.960% of the time exactly four of these GICs would pay off
32.768% of the time all five of these GICs would pay off

Thus if you invest in this 37.5% GIC you will do better than if you invest with Tom almost 33% of the time and in addition you will do just as well as with Tom almost 41% of the time. Thus you will do as well or better with me almost 74% of the time. BUY NOW!!!

However almost 26% of the time you will do worst than with Tom and on average you will do just as well as with Tom. Now since there is the possibility of a loss, including complete loss, with me why would you invest with me at all. Of course if Tom suddenly stopped issuing GICs mine might look pretty good since on average you have a solid 10% investment return.

But unfortunately Tom keeps selling his risk free GICs so mine aren't selling. Well that's easy ... I am now offering 40%. Thus we have now have a pure investment plus a pure gamble of type (c). In fact if you buy my new improved GICs you will on average earn a return of 12% which is 2% more than with Tom. Now my GICs begin to look attractive.

But should you put all of you money into my GICs? NO! ... because there still exists the possibility of you losing all of you money in my GICs. But why not put 10% of your investments in my GICs? This way you are managing your risk while taking advantage of the higher returns (on average) offered by my GICs.

I will conclude this topic in one more post.

A Numerical Example of My Take on Risk/Return

IMHO pure investment is risk free and I invest only to receive a pure investment return on my money. Please note that I am not necessarily implying that there are any pure investments out there.

pure gambling involves entering into a contract where there is a chance of walking away with more than I put in but for which there is also a chance that I will walk away with less. Here are three types of pure gambling:

(a) I pay you one dollar and you throw a die. If it comes up one you give me back two dollars otherwise you give me nothing. Thus on average one out of six times I will be ahead by one dollar and five out of six times I will be behind one dollar.

(b) Same as (a) but this time you pay me two dollars if the die comes up even otherwise you pay me nothing. Thus on average half the time I will be ahead one dollar and half the time I will be behind one dollar.

(c) Same as (a) but this time if the die comes up one you give nothing otherwise you give me two dollars. Thus on average I will be behind one dollar one out of six times and ahead one dollar five out of six times.

Let's examine each of these three cases in a little more detail.

(a) In Las Vegas people enter into these kinds of deals all the time. Of course the odds aren't quite this bad nor are they so obvious. Why would anybody enter into this kind of arrangement? Simple ... either for the thrill or because their expected utility is improved. For example the utility of one dollar is virtually nil. By contrast the utility of ten million dollars is great at least for me (and I expect for most people). If I buy a lottery ticket for one dollar my chances of winning as so low that on average my loss is 99 cents or possibly more (though less than a loss of a full dollar). On the other hand the resulting utility on average will be positive. Of course this has little directly to do with investing.

(b) One might enter into this kind of deal in a friendly low stakes game among friends. As will be seen in my next post this too has little to do with investing.

(c) Almost anybody would take this kind of bet. In particular most investments are a combination pure investment plus this type of gambling. In fact only in so-called risk free investments is there no taint of pure gambling mixed in. Let's look at an illustration which I will place in my next post.

Barry

Thanks Conrad,

There is still one point I don't get.

Let's say that s is Lichello's safe (i.e. 10%) and let S be your value (i.e. S = 1 + s )

Then Lichello's actual buy order is

pc - y * (1+s) = pc - S*y

but you assert that Lichello's buy order is of the form

S * (pc-y)

I don't see how this is possible. Of course there may be something obvious which I am not seeing.

Could you provide a mathematical demonstration of how the first expression can be transformed into the second. (Of course the two S factors would be different.)

Thanks,

Barry

Hi Conrad,

I just visited your pages on the vortex method. I found them very interesting and I have the following comments:

(1) There are a number of "TYPOS" most of which are unimportant but some of which might be misleading to the casual reader. For example: In the section "The Multifunctional Case" just after you speak of the Lichello Flaw you give the example ...

y1=5000 y2=3000 pc1=5000
pc1-y2=5000-3000=2000

but then you say

Test1 --- Advice= Buy 3000 ...

Clearly this 3000 is incorrect and merely a typo because you subsequently correctly use 2000.

(2)Perhaps I am rehashing old ground here because I gather that this may have been discussed over at TMF. I plan to read all of the post over there but these days it's hard just keeping up over here. In essence what you term the Lichello Flaw is something that I first noticed with AIM and it bothered me terribly but mostly for esthetic reasons.

I think it is unfair to describe it as a flaw ... I think of it as a feature. Of course if one uses GTC orders then it can be considered a flaw but if one uses periodic updating (e.g. monthly) which is the context in which Lichello was operating, then it can be a blessing in a rapidly decreasing market.

(3) Your answer to the flaw is in effect to buy an amount greater than what standard AIM would suggest and that amount would be such that the revised portfolio control would equal the total value of the stock held after the buy. I played with this idea more than two years ago. I ran countless simulations and found that this revised method sometimes outperformed standard AIM and sometimes did not.

(4) Your idea of employing adjustments to portfolio control both with buys and sells (and not necessarily the same adjustments) together with the idea of possibly using negative values of "f" for example ... is a very interesting idea. I will perform some simulations in this area. One could of course claim that Lichello used two separate factors with AIM. One adds to the portfolio control 50% of the amount purchased and 0% of the amount sold.

(5) There is a significant error in at least one place. You use "S" to represent Lichello's SAFE-factor and you then state that Lichello's buy-advice is S*(pc1-y2). There are in fact two errors here. First Lichello's "buy advice" is actually (pc1-y2) regardless of the value of S which is 10%. The "buy order" is this "buy advice" modified by S. In other words your term "buy-advice" seems to correspond to Lichello's "buy order" rather than Lichello's "buy advice". OK ... OK this is not really an error but more a question of semantics. This might however confuse AIMers who were reading casually. Let's use your term of "buy-advice"

You state that Lichello's buy-advice would be

S*(pc1-y2)

but this is incorrect!

Lichello's buy-advice (i.e. buy order in Lichello terminology) is

pc1-y2*(1+S)

Your fourth summary point (i.e. By adjusting the f-actor downwards by a small amount this has an identical effect as using a SAFE-factor, as is done in the Lichello AIM Model. The use of the f-factor eliminates the need for a SAFE altogether as the investor can tune the investment aggression exactly to his needs. ) seems to be based on this incorrect equation. This is something however, that I will investigate further.

I hope the above comments are helpful. Thanks for provide a very interesting colour variation to AIM.

Barry

Hello All,

I just finished reading every single post of the 745 posts so far since January 29th. That's an average of more than 31 posts per day. WOW!

I have responded to a few of the posts directly. This post however is a kind of general overall response to several posts.

(1) I am extremely pleased to see so many new voices. This bodes well for the future of this board.

(2) AIM 2002 ... I vote YES ... we should have it. I will attend for sure even if it is held in Vegas ... but ... I would much prefer some other venue. For example ... a country hotel/resort near Tom's digs had been proposed for last year.

(3) Compared with SI, IHUB seems to have both some advantages and some disadvantages. On balance however, I am quite pleased with it. In short ... I like our new home! One improvement that I would like to see is the ability to remain logged on.

(4) There is a tremendous amount of knowledge and wisdom already shared on this board both from former frequent SI posters as well as the newer voices. In particular, Tom continues to throw out statements that at first glance seem to be mildly interesting but when subsequently thought about are found to be profound. A case in point ... in one message (I forget which one as I have more than 700 currently swimming in front of my eyes) Tom talks about using the individual stock IW cash reserve as a guide to cash levels for each stock individually but using the IW mutual fund cash reserve for the aggregate of all AIMed stocks because they form a kind of mutual fund themselves. I believe there is a lot of meat in there for subsequent study and simulation testing. Thanks Tom, for this brilliant insight.

(5) The discussion on risk was most interesting. I am formulating a post on this subject which I hope to share in the near future.

Thanks for listening,

Barry

Isn't it interesting how such a simple algorithm can invoke such discussions? Thanks Mr. L

AMEN!!!!!


Barry

Hello Everyone,

Even though I came over to IHUB from SI because of AIM, I find this board to be very interesting and useful. I shall certainly read every post and contribute whenever I feel I have something valuable to say. The purpose of this post however is to ask three questions. After stating these questions I will provide some background on how and why I have these questions. I hope some of you might be able to either answer the questions or direct me to possible sources of answers. The questions are:

(1) What is "Double Dollar Cost Averaging" ?

(2) What is "Dollar Value Selling"?

(3) Who is E. S. Emory and in what publication did he (or she) introduce the concept of "Dollar Value Selling"?

The first edition of Lichello's "How To Make $1,000,000 ..." was published in 1977 and I gather that Synchrovest (i.e. Power Investing) was published earlier than that.

In 1980 Chuck Chakrapani published "The Money Spinner". This system (which what I first used and continue to be influenced by)is very similar to AIM. Basically The Money Spinner (TMS) uses standard AIM settings but rather than updating periodically its asks the question at what prices per share would I buy or sell 100 shares using standard AIM settings? One then issues GTC buy and sell orders not unlike what many modern AIMers do.I won't go into any more detail here except to observe that TMS looks very similar to AIM and in fact is clearly (in my opinion) a clear variation on AIM. My guess is that TMS was based on AIM. At this point let's look a direct quotation from the TMS book.

Is the Money Spinner Concept new? The answer to this question is a qualified YES. The concept of "dollar cost averaging" has been known for years. An improvement on this system known as "double dollar cost averaging" was introduced somewhat later which increased the profitability of the earlier system.

Subsequent to those developments, E. S. Emory introduced another system known as "dollar value selling" which is, in fact, the forerunner of the Money Spinner.

The dollar value selling system was adopted with varying degrees of success by different investors. Among this genre of techniques, perhaps the system developed by R. Lichello comes closest to the Money Spinner. Like the Money Spinner, Lichello's system IMPLICITLY uses the dollar cost averaging and dollar value selling principles.

I have to ask myself ... what is the system of R. Lichello which is referred to in this quote? As I observed above AIM and TMS are so similar to one another that AIM could be the system in question. Moreover, the publishing dates of AIM (1977) and TMS (1980) certainly make this possible.

Now here comes the interesting part.

Chuck Chakrapani also published an excellent investment book called "Financial Freedom On $5 A Day". I have two copies of book. One copy is the 6th edition from 1995 and the other is the first edition from 1983. The book explores and explains several investment techniques. The 1983 edition briefly describes TMS but the 1995 edition does not. I don't know in which edition reference to TMS disappeared but it is sufficient here to quote from the first edition.

STRATEGY 3: FORMULA PLANS ...

Another system developed by R. Lichello, is called the Synchrovest (Superpower Investing, published by New American Library, New York). This system recommends that you invest a fixed amount every month. Part of this amount should be invested in the stock market and the rest deposited in a savings account. The amount to be invested in the stock market each month will be decided on the stock's performance the previous month. The leftover money will go into the savings account. His book describes the system in detail.

More recently I developed a system called the Money Spinner. This system is based on several earlier systems including Lichello's system. The Money Spinner system differs from earlier systems in many subtle ways. Simply, this is how the system works: ...

Are we being told three years later that TMS is based on Synchrovest? (Rhetorical Question)

In spite of this quote it still seems clear to me that TMS was based on AIM but who knows.

In any event I would appreciate any information on the three questions posed at the start of posting.

Thanks,

Barry Savage

Hi Tom,

It looks like I moved out of NYC just in time. If were still there I might have had to entertain you. That would've been OK with me but I'm sure you wouldn't want to hear me perform the AIM song again

Take Care,

Barry